diff --git a/.gitignore b/.gitignore
index 714fee65..d24baf62 100644
--- a/.gitignore
+++ b/.gitignore
@@ -49,6 +49,9 @@ test_data/
 result/
 export_ptr_model/
 data/
+Data/
+Log/
+log/
 __pycache__/
 
 log/
diff --git a/.idea/workspace.xml b/.idea/workspace.xml
index e5e86293..553f693b 100644
--- a/.idea/workspace.xml
+++ b/.idea/workspace.xml
@@ -2,7 +2,8 @@
 <project version="4">
   <component name="ChangeListManager">
     <list default="true" id="69f12537-6c3d-43b1-a62a-ad07d70a14cd" name="Default" comment="">
-      <change type="NEW" beforePath="" afterPath="$PROJECT_DIR$/recommendation/Basic-DMF-Model/main.py" />
+      <change type="NEW" beforePath="" afterPath="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/FISM.py" />
+      <change type="MODIFICATION" beforePath="$PROJECT_DIR$/.gitignore" afterPath="$PROJECT_DIR$/.gitignore" />
       <change type="MODIFICATION" beforePath="$PROJECT_DIR$/.idea/workspace.xml" afterPath="$PROJECT_DIR$/.idea/workspace.xml" />
     </list>
     <option name="EXCLUDED_CONVERTED_TO_IGNORED" value="true" />
@@ -23,7 +24,7 @@
     <SUITE FILE_PATH="coverage/tensorflow1_2$multi_task_joint_training.coverage" NAME="multi-task-joint-training Coverage Results" MODIFIED="1533110252951" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/basic/multi-task-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$run_CartPole.coverage" NAME="run_CartPole Coverage Results" MODIFIED="1543383921818" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/RL/Basic-Actor-Critic" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$BPR.coverage" NAME="BPR Coverage Results" MODIFIED="1530202420711" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-BPR-Demo" />
-    <SUITE FILE_PATH="coverage/tensorflow1_2$main.coverage" NAME="main Coverage Results" MODIFIED="1550206332836" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-DMF-Model" />
+    <SUITE FILE_PATH="coverage/tensorflow1_2$main.coverage" NAME="main Coverage Results" MODIFIED="1550206452986" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-DMF-Model" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$tf_data_demo.coverage" NAME="tf_data_demo Coverage Results" MODIFIED="1528181619719" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-NCF-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$compareAC.coverage" NAME="compareAC Coverage Results" MODIFIED="1543484381532" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/RL" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$multi_task_alternate_training.coverage" NAME="multi-task-alternate-training Coverage Results" MODIFIED="1533109405226" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/basic/multi-task-Demo" />
@@ -34,6 +35,7 @@
     <SUITE FILE_PATH="coverage/tensorflow1_2$main__1_.coverage" NAME="main (1) Coverage Results" MODIFIED="1545377037092" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/RL/Basic-DisRL-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$preprocess__1_.coverage" NAME="preprocess (1) Coverage Results" MODIFIED="1542602954609" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-MKR-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$kg_preprocess.coverage" NAME="kg_preprocess Coverage Results" MODIFIED="1541604856562" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-DKN-Demo/kg" />
+    <SUITE FILE_PATH="coverage/tensorflow1_2$NAIS.coverage" NAME="NAIS Coverage Results" MODIFIED="1551431946415" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$plotResult.coverage" NAME="plotResult Coverage Results" MODIFIED="1529723400750" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-MLR-Demo" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$news_preprocess.coverage" NAME="news_preprocess Coverage Results" MODIFIED="1541602017949" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-DKN-Demo/news" />
     <SUITE FILE_PATH="coverage/tensorflow1_2$data.coverage" NAME="data Coverage Results" MODIFIED="1529689500332" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="true" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/recommendation/Basic-MLR-Demo" />
@@ -57,120 +59,86 @@
   </component>
   <component name="FileEditorManager">
     <leaf SIDE_TABS_SIZE_LIMIT_KEY="300">
-      <file leaf-file-name="run_ppo.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/run_ppo.py">
+      <file leaf-file-name="model.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-MKR-Demo/model.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="0">
-              <caret line="95" column="45" lean-forward="false" selection-start-line="95" selection-start-column="45" selection-end-line="95" selection-end-column="45" />
+            <state relative-caret-position="45">
+              <caret line="3" column="42" lean-forward="false" selection-start-line="3" selection-start-column="42" selection-end-line="3" selection-end-column="42" />
               <folding>
-                <element signature="e#0#15#0" expanded="true" />
+                <element signature="e#0#18#0" expanded="true" />
               </folding>
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="run_gail.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/run_gail.py">
+      <file leaf-file-name="NAIS.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/NAIS.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="15">
-              <caret line="15" column="0" lean-forward="false" selection-start-line="15" selection-start-column="0" selection-end-line="15" selection-end-column="0" />
+            <state relative-caret-position="172">
+              <caret line="182" column="43" lean-forward="false" selection-start-line="182" selection-start-column="43" selection-end-line="182" selection-end-column="43" />
               <folding>
-                <element signature="e#0#15#0" expanded="true" />
+                <element signature="e#0#23#0" expanded="true" />
               </folding>
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="rnn.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/rnn.py">
+      <file leaf-file-name="Batch_gen.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Batch_gen.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="-102">
-              <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+            <state relative-caret-position="192">
+              <caret line="109" column="54" lean-forward="false" selection-start-line="109" selection-start-column="54" selection-end-line="109" selection-end-column="54" />
               <folding>
-                <element signature="e#852#890#0" expanded="true" />
+                <element signature="e#0#22#0" expanded="true" />
               </folding>
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="vecAttGruCell.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/vecAttGruCell.py">
+      <file leaf-file-name=".gitignore" pinned="false" current-in-tab="true">
+        <entry file="file://$PROJECT_DIR$/.gitignore">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="0">
-              <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
-              <folding>
-                <element signature="e#0#23#0" expanded="true" />
-              </folding>
+            <state relative-caret-position="569">
+              <caret line="53" column="4" lean-forward="false" selection-start-line="53" selection-start-column="4" selection-end-line="53" selection-end-column="4" />
+              <folding />
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="README" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/data/README">
+      <file leaf-file-name="Evaluate.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Evaluate.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="0">
-              <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+            <state relative-caret-position="-213">
+              <caret line="68" column="19" lean-forward="false" selection-start-line="68" selection-start-column="19" selection-end-line="68" selection-end-column="19" />
               <folding />
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="main.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/main.py">
+      <file leaf-file-name="FISM.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/FISM.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="0">
-              <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+            <state relative-caret-position="2648">
+              <caret line="180" column="112" lean-forward="false" selection-start-line="180" selection-start-column="112" selection-end-line="180" selection-end-column="112" />
               <folding>
-                <element signature="e#2#25#0" expanded="true" />
+                <element signature="e#0#38#0" expanded="true" />
               </folding>
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="dataset.py" pinned="false" current-in-tab="true">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/dataset.py">
+      <file leaf-file-name="Dataset.py" pinned="false" current-in-tab="false">
+        <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Dataset.py">
           <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="299">
-              <caret line="55" column="42" lean-forward="true" selection-start-line="55" selection-start-column="42" selection-end-line="55" selection-end-column="42" />
+            <state relative-caret-position="195">
+              <caret line="13" column="63" lean-forward="false" selection-start-line="13" selection-start-column="63" selection-end-line="13" selection-end-column="63" />
               <folding>
-                <element signature="e#26#44#0" expanded="true" />
+                <element signature="e#0#25#0" expanded="true" />
               </folding>
             </state>
           </provider>
         </entry>
       </file>
-      <file leaf-file-name="README.md" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/README.md">
-          <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="90">
-              <caret line="6" column="0" lean-forward="false" selection-start-line="6" selection-start-column="0" selection-end-line="6" selection-end-column="0" />
-              <folding />
-            </state>
-          </provider>
-        </entry>
-      </file>
-      <file leaf-file-name="test_policy.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/test_policy.py">
-          <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="1155">
-              <caret line="77" column="0" lean-forward="false" selection-start-line="77" selection-start-column="0" selection-end-line="77" selection-end-column="0" />
-              <folding>
-                <element signature="e#0#10#0" expanded="true" />
-              </folding>
-            </state>
-          </provider>
-        </entry>
-      </file>
-      <file leaf-file-name="discriminator.py" pinned="false" current-in-tab="false">
-        <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/network_models/discriminator.py">
-          <provider selected="true" editor-type-id="text-editor">
-            <state relative-caret-position="945">
-              <caret line="63" column="43" lean-forward="false" selection-start-line="63" selection-start-column="43" selection-end-line="63" selection-end-column="43" />
-              <folding />
-            </state>
-          </provider>
-        </entry>
-      </file>
     </leaf>
   </component>
   <component name="FileTemplateManagerImpl">
@@ -195,7 +163,12 @@
       <find>one_hot</find>
       <find>print</find>
       <find>self.tfd</find>
+      <find>positional_encoding</find>
+      <find>self.num_idx</find>
     </findStrings>
+    <dirStrings>
+      <dir>$PROJECT_DIR$/recommendation/Basic-NAIS-Demo</dir>
+    </dirStrings>
   </component>
   <component name="Git.Settings">
     <option name="PUSH_AUTO_UPDATE" value="true" />
@@ -204,11 +177,6 @@
   <component name="IdeDocumentHistory">
     <option name="CHANGED_PATHS">
       <list>
-        <option value="$PROJECT_DIR$/RL/Basic-A2C-Demo/A2C.py" />
-        <option value="$PROJECT_DIR$/RL/Basic-AC-Demo/AC.py" />
-        <option value="$PROJECT_DIR$/RL/Basic-A3C-Demo/A3C.py" />
-        <option value="$PROJECT_DIR$/RL/compareAC.py" />
-        <option value="$PROJECT_DIR$/recommendation/recommendation-FFM-Demo/FFM_model.py" />
         <option value="$PROJECT_DIR$/basic/Basic-Transformer-Demo/hyperparams.py" />
         <option value="$PROJECT_DIR$/basic/Basic-Transformer-Demo/eval.py" />
         <option value="$PROJECT_DIR$/basic/Basic-Transformer-Demo/data_path" />
@@ -249,12 +217,17 @@
         <option value="$PROJECT_DIR$/RL/Basic-GAIL-Demo/network_models/discriminator.py" />
         <option value="$PROJECT_DIR$/RL/Basic-GAIL-Demo/run_gail.py" />
         <option value="$PROJECT_DIR$/RL/Basic-GAIL-Demo/test_policy.py" />
-        <option value="$PROJECT_DIR$/.gitignore" />
         <option value="$PROJECT_DIR$/recommendation/Basic-DMF-Demo.py" />
         <option value="$PROJECT_DIR$/recommendation/Basic-DMF-Model/data/main.py" />
         <option value="$PROJECT_DIR$/recommendation/Basic-DMF-Model/README.md" />
         <option value="$PROJECT_DIR$/recommendation/Basic-DMF-Model/main.py" />
         <option value="$PROJECT_DIR$/recommendation/Basic-DMF-Model/dataset.py" />
+        <option value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Dataset.py" />
+        <option value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/NAIS.py" />
+        <option value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Evaluate.py" />
+        <option value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/FISM.py" />
+        <option value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Batch_gen.py" />
+        <option value="$PROJECT_DIR$/.gitignore" />
       </list>
     </option>
   </component>
@@ -283,8 +256,6 @@
       <foldersAlwaysOnTop value="true" />
     </navigator>
     <panes>
-      <pane id="Scope" />
-      <pane id="Scratches" />
       <pane id="ProjectPane">
         <subPane>
           <expand>
@@ -295,31 +266,47 @@
             <path>
               <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
               <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
-              <item name="recommendation" type="462c0819:PsiDirectoryNode" />
+              <item name="CV" type="462c0819:PsiDirectoryNode" />
+            </path>
+            <path>
+              <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
+              <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
+              <item name="GAN" type="462c0819:PsiDirectoryNode" />
+            </path>
+            <path>
+              <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
+              <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
+              <item name="nlp" type="462c0819:PsiDirectoryNode" />
+            </path>
+            <path>
+              <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
+              <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
+              <item name="other" type="462c0819:PsiDirectoryNode" />
             </path>
             <path>
               <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
               <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
               <item name="recommendation" type="462c0819:PsiDirectoryNode" />
-              <item name="Basic-DIEN-Demo" type="462c0819:PsiDirectoryNode" />
             </path>
             <path>
               <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
               <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
               <item name="recommendation" type="462c0819:PsiDirectoryNode" />
-              <item name="Basic-DMF-Model" type="462c0819:PsiDirectoryNode" />
+              <item name="Basic-NAIS-Demo" type="462c0819:PsiDirectoryNode" />
             </path>
             <path>
               <item name="tensorflow1.2" type="b2602c69:ProjectViewProjectNode" />
               <item name="tensorflow1.2" type="462c0819:PsiDirectoryNode" />
               <item name="recommendation" type="462c0819:PsiDirectoryNode" />
-              <item name="Basic-DMF-Model" type="462c0819:PsiDirectoryNode" />
-              <item name="data" type="462c0819:PsiDirectoryNode" />
+              <item name="Basic-NAIS-Demo" type="462c0819:PsiDirectoryNode" />
+              <item name="Data" type="462c0819:PsiDirectoryNode" />
             </path>
           </expand>
           <select />
         </subPane>
       </pane>
+      <pane id="Scratches" />
+      <pane id="Scope" />
     </panes>
   </component>
   <component name="PropertiesComponent">
@@ -330,13 +317,6 @@
     <property name="SearchEverywhereHistoryKey" value="&#9;FILE&#9;file:///Users/meituan_sxw/PycharmProjects/tensorflow1.2/RL/Basic-GAIL-Demo/algo/ppo.py" />
   </component>
   <component name="RecentsManager">
-    <key name="CopyFile.RECENT_KEYS">
-      <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model" />
-      <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model/data" />
-      <recent name="$PROJECT_DIR$/RL/Basic-GAIL-Demo" />
-      <recent name="$PROJECT_DIR$/basic/Basic-Transformer-Demo/data/de-en" />
-      <recent name="$PROJECT_DIR$/recommendation/Basic-MKR-Demo/data" />
-    </key>
     <key name="MoveFile.RECENT_KEYS">
       <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model" />
       <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model/data" />
@@ -344,6 +324,13 @@
       <recent name="$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/data" />
       <recent name="$PROJECT_DIR$/CV" />
     </key>
+    <key name="CopyFile.RECENT_KEYS">
+      <recent name="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo" />
+      <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model" />
+      <recent name="$PROJECT_DIR$/recommendation/Basic-DMF-Model/data" />
+      <recent name="$PROJECT_DIR$/RL/Basic-GAIL-Demo" />
+      <recent name="$PROJECT_DIR$/basic/Basic-Transformer-Demo/data/de-en" />
+    </key>
   </component>
   <component name="RunDashboard">
     <option name="ruleStates">
@@ -357,21 +344,21 @@
       </list>
     </option>
   </component>
-  <component name="RunManager" selected="Python.main">
-    <configuration name="main (1)" type="PythonConfigurationType" factoryName="Python" temporary="true">
+  <component name="RunManager" selected="Python.NAIS">
+    <configuration name="NAIS" type="PythonConfigurationType" factoryName="Python" temporary="true">
       <option name="INTERPRETER_OPTIONS" value="" />
       <option name="PARENT_ENVS" value="true" />
       <envs>
         <env name="PYTHONUNBUFFERED" value="1" />
       </envs>
       <option name="SDK_HOME" value="" />
-      <option name="WORKING_DIRECTORY" value="$PROJECT_DIR$/RL/Basic-DisRL-Demo" />
+      <option name="WORKING_DIRECTORY" value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo" />
       <option name="IS_MODULE_SDK" value="true" />
       <option name="ADD_CONTENT_ROOTS" value="true" />
       <option name="ADD_SOURCE_ROOTS" value="true" />
       <module name="tensorflow1.2" />
       <EXTENSION ID="PythonCoverageRunConfigurationExtension" enabled="false" sample_coverage="true" runner="coverage.py" />
-      <option name="SCRIPT_NAME" value="$PROJECT_DIR$/RL/Basic-DisRL-Demo/main.py" />
+      <option name="SCRIPT_NAME" value="$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/NAIS.py" />
       <option name="PARAMETERS" value="" />
       <option name="SHOW_COMMAND_LINE" value="false" />
       <option name="EMULATE_TERMINAL" value="false" />
@@ -449,19 +436,19 @@
       <option name="EMULATE_TERMINAL" value="false" />
     </configuration>
     <list size="5">
-      <item index="0" class="java.lang.String" itemvalue="Python.main (1)" />
-      <item index="1" class="java.lang.String" itemvalue="Python.main (2)" />
-      <item index="2" class="java.lang.String" itemvalue="Python.run_ppo" />
-      <item index="3" class="java.lang.String" itemvalue="Python.run_gail" />
-      <item index="4" class="java.lang.String" itemvalue="Python.main" />
+      <item index="0" class="java.lang.String" itemvalue="Python.main (2)" />
+      <item index="1" class="java.lang.String" itemvalue="Python.run_ppo" />
+      <item index="2" class="java.lang.String" itemvalue="Python.run_gail" />
+      <item index="3" class="java.lang.String" itemvalue="Python.main" />
+      <item index="4" class="java.lang.String" itemvalue="Python.NAIS" />
     </list>
     <recent_temporary>
       <list size="5">
-        <item index="0" class="java.lang.String" itemvalue="Python.main" />
-        <item index="1" class="java.lang.String" itemvalue="Python.run_gail" />
-        <item index="2" class="java.lang.String" itemvalue="Python.run_ppo" />
-        <item index="3" class="java.lang.String" itemvalue="Python.main (2)" />
-        <item index="4" class="java.lang.String" itemvalue="Python.main (1)" />
+        <item index="0" class="java.lang.String" itemvalue="Python.NAIS" />
+        <item index="1" class="java.lang.String" itemvalue="Python.main" />
+        <item index="2" class="java.lang.String" itemvalue="Python.run_gail" />
+        <item index="3" class="java.lang.String" itemvalue="Python.run_ppo" />
+        <item index="4" class="java.lang.String" itemvalue="Python.main (2)" />
       </list>
     </recent_temporary>
   </component>
@@ -491,15 +478,16 @@
     <frame x="0" y="0" width="1440" height="900" extended-state="0" />
     <editor active="true" />
     <layout>
-      <window_info id="Project" active="false" anchor="left" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="true" show_stripe_button="true" weight="0.17167382" sideWeight="0.5" order="0" side_tool="false" content_ui="combo" />
       <window_info id="TODO" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.32889965" sideWeight="0.5" order="6" side_tool="false" content_ui="tabs" />
       <window_info id="Event Log" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="7" side_tool="true" content_ui="tabs" />
-      <window_info id="Database" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="3" side_tool="false" content_ui="tabs" />
-      <window_info id="Run" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="true" show_stripe_button="true" weight="0.38331318" sideWeight="0.5" order="2" side_tool="false" content_ui="tabs" />
+      <window_info id="Run" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="true" show_stripe_button="true" weight="0.42321643" sideWeight="0.5" order="2" side_tool="false" content_ui="tabs" />
       <window_info id="Version Control" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.32889965" sideWeight="0.5" order="7" side_tool="false" content_ui="tabs" />
       <window_info id="Python Console" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.32889965" sideWeight="0.5" order="7" side_tool="false" content_ui="tabs" />
+      <window_info id="Terminal" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.2031439" sideWeight="0.5" order="7" side_tool="false" content_ui="tabs" />
+      <window_info id="Project" active="false" anchor="left" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="true" show_stripe_button="true" weight="0.18240343" sideWeight="0.5" order="0" side_tool="false" content_ui="combo" />
+      <window_info id="Database" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="3" side_tool="false" content_ui="tabs" />
+      <window_info id="Find" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="1" side_tool="false" content_ui="tabs" />
       <window_info id="Structure" active="false" anchor="left" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.25" sideWeight="0.5" order="1" side_tool="false" content_ui="tabs" />
-      <window_info id="Terminal" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.21886337" sideWeight="0.5" order="7" side_tool="false" content_ui="tabs" />
       <window_info id="Debug" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.39782345" sideWeight="0.5" order="3" side_tool="false" content_ui="tabs" />
       <window_info id="Favorites" active="false" anchor="left" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="2" side_tool="true" content_ui="tabs" />
       <window_info id="Data View" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="3" side_tool="false" content_ui="tabs" />
@@ -509,7 +497,6 @@
       <window_info id="Commander" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.4" sideWeight="0.5" order="0" side_tool="false" content_ui="tabs" />
       <window_info id="Inspection" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.4" sideWeight="0.5" order="5" side_tool="false" content_ui="tabs" />
       <window_info id="Hierarchy" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.25" sideWeight="0.5" order="2" side_tool="false" content_ui="combo" />
-      <window_info id="Find" active="false" anchor="bottom" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.33" sideWeight="0.5" order="1" side_tool="false" content_ui="tabs" />
       <window_info id="Ant Build" active="false" anchor="right" auto_hide="false" internal_type="DOCKED" type="DOCKED" visible="false" show_stripe_button="true" weight="0.25" sideWeight="0.5" order="1" side_tool="false" content_ui="tabs" />
     </layout>
   </component>
@@ -545,85 +532,6 @@
     <watches-manager />
   </component>
   <component name="editorHistoryManager">
-    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/eval.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="585">
-          <caret line="48" column="37" lean-forward="false" selection-start-line="48" selection-start-column="37" selection-end-line="48" selection-end-column="37" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/hyperparams.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="360">
-          <caret line="24" column="0" lean-forward="false" selection-start-line="24" selection-start-column="0" selection-end-line="24" selection-end-column="0" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/data_load.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="1200">
-          <caret line="84" column="10" lean-forward="false" selection-start-line="84" selection-start-column="10" selection-end-line="84" selection-end-column="10" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/modules.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="3645">
-          <caret line="244" column="0" lean-forward="false" selection-start-line="244" selection-start-column="0" selection-end-line="277" selection-end-column="18" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/train.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="15">
-          <caret line="131" column="78" lean-forward="false" selection-start-line="131" selection-start-column="78" selection-end-line="131" selection-end-column="78" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/RL/Basic-HER-Demo/readme">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="15">
-          <caret line="1" column="37" lean-forward="false" selection-start-line="1" selection-start-column="37" selection-end-line="1" selection-end-column="37" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/basic/some_numpy_func.py" />
-    <entry file="file://$PROJECT_DIR$/RL/Basic-HER-Demo/HindSightExperienceReplay.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="399">
-          <caret line="190" column="0" lean-forward="false" selection-start-line="188" selection-start-column="8" selection-end-line="190" selection-end-column="0" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/RL/DQN-demo/RL_brain.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="0">
-          <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$USER_HOME$/Anaconda/anaconda/envs/python3_5/lib/python3.5/site-packages/tensorflow/python/ops/state_ops.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="138">
-          <caret line="247" column="4" lean-forward="false" selection-start-line="247" selection-start-column="4" selection-end-line="247" selection-end-column="4" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/RL/Basic-NoisyNet-Demo/Config.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="285">
-          <caret line="19" column="26" lean-forward="false" selection-start-line="19" selection-start-column="26" selection-end-line="19" selection-end-column="36" />
-        </state>
-      </provider>
-    </entry>
-    <entry file="file://$PROJECT_DIR$/ssfsdfsdf.py" />
-    <entry file="file://$PROJECT_DIR$/RL/Basic-NoisyNet-Demo/utils.py">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="645">
-          <caret line="44" column="0" lean-forward="true" selection-start-line="13" selection-start-column="0" selection-end-line="44" selection-end-column="0" />
-        </state>
-      </provider>
-    </entry>
     <entry file="file://$PROJECT_DIR$/RL/Basic-NoisyNet-Demo/readme">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="15">
@@ -692,7 +600,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="161">
           <caret line="569" column="0" lean-forward="false" selection-start-line="569" selection-start-column="0" selection-end-line="569" selection-end-column="0" />
-          <folding />
         </state>
       </provider>
     </entry>
@@ -714,9 +621,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="75">
           <caret line="5" column="6" lean-forward="false" selection-start-line="5" selection-start-column="6" selection-end-line="5" selection-end-column="26" />
-          <folding>
-            <element signature="e#45#63#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -724,9 +628,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="2340">
           <caret line="156" column="30" lean-forward="false" selection-start-line="156" selection-start-column="30" selection-end-line="156" selection-end-column="30" />
-          <folding>
-            <element signature="e#0#28#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -734,9 +635,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="990">
           <caret line="66" column="39" lean-forward="false" selection-start-line="66" selection-start-column="39" selection-end-line="66" selection-end-column="39" />
-          <folding>
-            <element signature="e#0#18#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -744,7 +642,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="765">
           <caret line="51" column="18" lean-forward="true" selection-start-line="51" selection-start-column="18" selection-end-line="51" selection-end-column="18" />
-          <folding />
         </state>
       </provider>
     </entry>
@@ -752,7 +649,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="390">
           <caret line="26" column="0" lean-forward="false" selection-start-line="26" selection-start-column="0" selection-end-line="26" selection-end-column="0" />
-          <folding />
         </state>
       </provider>
     </entry>
@@ -760,9 +656,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="2295">
           <caret line="153" column="13" lean-forward="false" selection-start-line="153" selection-start-column="13" selection-end-line="153" selection-end-column="37" />
-          <folding>
-            <element signature="e#0#23#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -770,9 +663,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="375">
           <caret line="25" column="25" lean-forward="false" selection-start-line="25" selection-start-column="25" selection-end-line="25" selection-end-column="25" />
-          <folding>
-            <element signature="e#0#15#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -780,9 +670,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="569">
           <caret line="277" column="25" lean-forward="false" selection-start-line="277" selection-start-column="25" selection-end-line="277" selection-end-column="25" />
-          <folding>
-            <element signature="e#23#41#0" expanded="true" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -791,9 +678,6 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="1725">
           <caret line="115" column="0" lean-forward="false" selection-start-line="115" selection-start-column="0" selection-end-line="115" selection-end-column="0" />
-          <folding>
-            <element signature="e#0#23#0" expanded="false" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -811,39 +695,24 @@
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/.gitignore">
-      <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="855">
-          <caret line="57" column="9" lean-forward="false" selection-start-line="57" selection-start-column="9" selection-end-line="57" selection-end-column="9" />
-          <folding />
-        </state>
-      </provider>
-    </entry>
     <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/network_models/policy_net.py">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="660">
           <caret line="44" column="75" lean-forward="false" selection-start-line="44" selection-start-column="75" selection-end-line="44" selection-end-column="75" />
-          <folding />
         </state>
       </provider>
     </entry>
     <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/test_policy.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="1155">
+        <state relative-caret-position="1095">
           <caret line="77" column="0" lean-forward="false" selection-start-line="77" selection-start-column="0" selection-end-line="77" selection-end-column="0" />
-          <folding>
-            <element signature="e#0#10#0" expanded="true" />
-          </folding>
         </state>
       </provider>
     </entry>
     <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/run_ppo.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="0">
+        <state relative-caret-position="1350">
           <caret line="95" column="45" lean-forward="false" selection-start-line="95" selection-start-column="45" selection-end-line="95" selection-end-column="45" />
-          <folding>
-            <element signature="e#0#15#0" expanded="true" />
-          </folding>
         </state>
       </provider>
     </entry>
@@ -851,49 +720,99 @@
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="945">
           <caret line="63" column="43" lean-forward="false" selection-start-line="63" selection-start-column="43" selection-end-line="63" selection-end-column="43" />
-          <folding />
         </state>
       </provider>
     </entry>
     <entry file="file://$PROJECT_DIR$/RL/Basic-GAIL-Demo/run_gail.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="15">
+        <state relative-caret-position="135">
           <caret line="15" column="0" lean-forward="false" selection-start-line="15" selection-start-column="0" selection-end-line="15" selection-end-column="0" />
-          <folding>
-            <element signature="e#0#15#0" expanded="true" />
-          </folding>
         </state>
       </provider>
     </entry>
     <entry file="file://$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/rnn.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="-102">
+        <state relative-caret-position="0">
+          <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/vecAttGruCell.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="0">
           <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
           <folding>
-            <element signature="e#852#890#0" expanded="true" />
+            <element signature="e#0#23#0" expanded="false" />
           </folding>
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DIEN-Demo/vecAttGruCell.py">
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Demo.py" />
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/data/README">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="0">
           <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+          <folding />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/README.md">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="90">
+          <caret line="6" column="0" lean-forward="false" selection-start-line="6" selection-start-column="0" selection-end-line="6" selection-end-column="0" />
+          <folding />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/main.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="105">
+          <caret line="14" column="0" lean-forward="false" selection-start-line="14" selection-start-column="0" selection-end-line="14" selection-end-column="0" />
           <folding>
-            <element signature="e#0#23#0" expanded="true" />
+            <element signature="e#2#25#0" expanded="false" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/dataset.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="225">
+          <caret line="16" column="29" lean-forward="false" selection-start-line="16" selection-start-column="29" selection-end-line="16" selection-end-column="29" />
+          <folding>
+            <element signature="e#26#44#0" expanded="false" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/nlp/chat_bot_seq2seq_attention/model.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="1890">
+          <caret line="127" column="40" lean-forward="false" selection-start-line="127" selection-start-column="40" selection-end-line="127" selection-end-column="76" />
+          <folding>
+            <element signature="e#0#23#0" expanded="false" />
           </folding>
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Demo.py">
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DKN-Demo/dkn.py">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="0">
           <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+          <folding>
+            <element signature="e#0#18#0" expanded="false" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/train.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="1875">
+          <caret line="131" column="78" lean-forward="false" selection-start-line="131" selection-start-column="78" selection-end-line="131" selection-end-column="78" />
           <folding />
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/data/README">
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Data/ml-1m.test.rating">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="0">
           <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
@@ -901,33 +820,94 @@
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/README.md">
+    <entry file="file://$PROJECT_DIR$/basic/Basic-Transformer-Demo/modules.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="90">
-          <caret line="6" column="0" lean-forward="false" selection-start-line="6" selection-start-column="0" selection-end-line="6" selection-end-column="0" />
+        <state relative-caret-position="15">
+          <caret line="90" column="38" lean-forward="false" selection-start-line="90" selection-start-column="38" selection-end-line="90" selection-end-column="38" />
           <folding />
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/main.py">
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-MKR-Demo/model.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="45">
+          <caret line="3" column="42" lean-forward="false" selection-start-line="3" selection-start-column="42" selection-end-line="3" selection-end-column="42" />
+          <folding>
+            <element signature="e#0#18#0" expanded="true" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Data/ml-1m.train.rating">
+      <provider selected="true" editor-type-id="LargeFileEditor">
+        <state relative-caret-position="-6361">
+          <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Data/ml-1m.test.negative">
       <provider selected="true" editor-type-id="text-editor">
         <state relative-caret-position="0">
           <caret line="0" column="0" lean-forward="false" selection-start-line="0" selection-start-column="0" selection-end-line="0" selection-end-column="0" />
+          <folding />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Dataset.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="195">
+          <caret line="13" column="63" lean-forward="false" selection-start-line="13" selection-start-column="63" selection-end-line="13" selection-end-column="63" />
           <folding>
-            <element signature="e#2#25#0" expanded="true" />
+            <element signature="e#0#25#0" expanded="true" />
           </folding>
         </state>
       </provider>
     </entry>
-    <entry file="file://$PROJECT_DIR$/recommendation/Basic-DMF-Model/dataset.py">
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Evaluate.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="-213">
+          <caret line="68" column="19" lean-forward="false" selection-start-line="68" selection-start-column="19" selection-end-line="68" selection-end-column="19" />
+          <folding />
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/FISM.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="2648">
+          <caret line="180" column="112" lean-forward="false" selection-start-line="180" selection-start-column="112" selection-end-line="180" selection-end-column="112" />
+          <folding>
+            <element signature="e#0#38#0" expanded="true" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/NAIS.py">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="172">
+          <caret line="182" column="43" lean-forward="false" selection-start-line="182" selection-start-column="43" selection-end-line="182" selection-end-column="43" />
+          <folding>
+            <element signature="e#0#23#0" expanded="true" />
+          </folding>
+        </state>
+      </provider>
+    </entry>
+    <entry file="file://$PROJECT_DIR$/recommendation/Basic-NAIS-Demo/Batch_gen.py">
       <provider selected="true" editor-type-id="text-editor">
-        <state relative-caret-position="299">
-          <caret line="55" column="42" lean-forward="true" selection-start-line="55" selection-start-column="42" selection-end-line="55" selection-end-column="42" />
+        <state relative-caret-position="192">
+          <caret line="109" column="54" lean-forward="false" selection-start-line="109" selection-start-column="54" selection-end-line="109" selection-end-column="54" />
           <folding>
-            <element signature="e#26#44#0" expanded="true" />
+            <element signature="e#0#22#0" expanded="true" />
           </folding>
         </state>
       </provider>
     </entry>
+    <entry file="file://$PROJECT_DIR$/.gitignore">
+      <provider selected="true" editor-type-id="text-editor">
+        <state relative-caret-position="569">
+          <caret line="53" column="4" lean-forward="false" selection-start-line="53" selection-start-column="4" selection-end-line="53" selection-end-column="4" />
+          <folding />
+        </state>
+      </provider>
+    </entry>
   </component>
 </project>
\ No newline at end of file
diff --git a/recommendation/Basic-NAIS-Demo/Batch_gen.py b/recommendation/Basic-NAIS-Demo/Batch_gen.py
new file mode 100644
index 00000000..c040e4c8
--- /dev/null
+++ b/recommendation/Basic-NAIS-Demo/Batch_gen.py
@@ -0,0 +1,185 @@
+import multiprocessing
+import numpy as np
+
+_Dataset = None
+_batch_size = None
+_num_negatives = None
+_num_items = None
+_user_input = None
+_item_input = None
+_labels = None
+_index = None
+_num_batch = None
+_batch_length = None
+
+
+def shuffle(dataset, batch_choice, num_negatives):  # negative sampling and shuffle the data
+
+    global _Dataset
+    global _batch_size
+    global _num_negatives
+    global _num_items
+    global _user_input
+    global _item_input
+    global _labels
+    global _index
+    global _num_batch
+    global _batch_length
+    _Dataset = dataset
+    _num_negatives = num_negatives
+
+    if batch_choice == 'user':
+        _num_items, _user_input, _item_input, _labels, _batch_length = _get_train_data_user()
+        _num_batch = len(_batch_length)
+        return _preprocess(_get_train_batch_user)
+
+    else:
+        batch_choices = batch_choice.split(":")
+        if batch_choices[0] == 'fixed':
+            _batch_size = int(batch_choices[1])
+            _num_items, _user_input, _item_input, _labels = _get_train_data_fixed()
+            iterations = len(_user_input)
+            _index = np.arange(iterations)
+            _num_batch = iterations / _batch_size
+            return _preprocess(_get_train_batch_fixed)
+        else:
+            print("invalid batch size !")
+
+
+def batch_gen(batches, i):
+    return [(batches[r])[i] for r in range(4)]
+
+
+def _preprocess(get_train_batch):  # generate the masked batch list
+    user_input_list, num_idx_list, item_input_list, labels_list = [], [], [], []
+    cpu_count = multiprocessing.cpu_count()
+    if cpu_count == 1:
+        for i in range(_num_batch):
+            ui, ni, ii, l = get_train_batch(i)
+            user_input_list.append(ui)
+            num_idx_list.append(ni)
+            item_input_list.append(ii)
+            labels_list.append(l)
+    else:
+        pool = multiprocessing.Pool(cpu_count)
+        res = pool.map(get_train_batch, list(range(_num_batch)))
+        pool.close()
+        pool.join()
+        user_input_list = [r[0] for r in res]
+        num_idx_list = [r[1] for r in res]
+        item_input_list = [r[2] for r in res]
+        labels_list = [r[3] for r in res]
+    return (user_input_list, num_idx_list, item_input_list, labels_list)
+
+
+def _get_train_data_user():
+    user_input, item_input, labels, batch_length = [], [], [], []
+    train = _Dataset.trainMatrix
+    trainList = _Dataset.trainList
+    num_items = train.shape[1]
+    num_users = train.shape[0]
+    for u in range(num_users):
+        if u == 0:
+            batch_length.append((1 + _num_negatives) * len(trainList[u]))
+        else:
+            batch_length.append((1 + _num_negatives) * len(trainList[u]) + batch_length[u - 1])
+        for i in trainList[u]:
+            # positive instance
+            user_input.append(u)
+            item_input.append(i)
+            labels.append(1)
+            # negative instances
+            for t in range(_num_negatives):
+                j = np.random.randint(num_items)
+                while j in trainList[u]:
+                    j = np.random.randint(num_items)
+                user_input.append(u)
+                item_input.append(j)
+                labels.append(0)
+    return num_items, user_input, item_input, labels, batch_length
+
+
+def _get_train_batch_user(i):
+    # represent the feature of users via items rated by him/her
+    user_list, num_list, item_list, labels_list = [], [], [], []
+    trainList = _Dataset.trainList
+    if i == 0:
+        begin = 0
+    else:
+        begin = _batch_length[i - 1]
+    batch_index = list(range(begin, _batch_length[i]))
+    np.random.shuffle(batch_index)
+    for idx in batch_index:
+        user_idx = _user_input[idx]
+        item_idx = _item_input[idx]
+        nonzero_row = []
+        nonzero_row += trainList[user_idx]
+        num_list.append(_remove_item(_num_items, nonzero_row, item_idx))
+        user_list.append(nonzero_row)
+        item_list.append(item_idx)
+        labels_list.append(_labels[idx])
+    user_input = np.array(_add_mask(_num_items, user_list, max(num_list)))
+    num_idx = np.array(num_list)
+    item_input = np.array(item_list)
+    labels = np.array(labels_list)
+    return (user_input, num_idx, item_input, labels)
+
+
+def _get_train_data_fixed():
+    user_input, item_input, labels = [], [], []
+    train = _Dataset.trainMatrix
+    num_items = train.shape[1]
+    for (u, i) in train.keys():
+        # positive instance
+        user_items = []
+        user_input.append(u)
+        item_input.append(i)
+        labels.append(1)
+        # negative instances
+        for t in range(_num_negatives):
+            j = np.random.randint(num_items)
+            while train.has_key((u, j)):
+                j = np.random.randint(num_items)
+            user_input.append(u)
+            item_input.append(j)
+            labels.append(0)
+    return num_items, user_input, item_input, labels
+
+
+def _get_train_batch_fixed(i):
+    # represent the feature of users via items rated by him/her
+    user_list, num_list, item_list, labels_list = [], [], [], []
+    trainList = _Dataset.trainList
+    begin = i * _batch_size
+    for idx in range(begin, begin + _batch_size):
+        user_idx = _user_input[_index[idx]]
+        item_idx = _item_input[_index[idx]]
+        nonzero_row = []
+        nonzero_row += trainList[user_idx]
+        num_list.append(_remove_item(_num_items, nonzero_row, item_idx))
+        user_list.append(nonzero_row)
+        item_list.append(item_idx)
+        labels_list.append(_labels[_index[idx]])
+    user_input = np.array(_add_mask(_num_items, user_list, max(num_list)))
+    num_idx = np.array(num_list)
+    item_input = np.array(item_list)
+    labels = np.array(labels_list)
+    return (user_input, num_idx, item_input, labels)
+
+
+def _remove_item(feature_mask, users, item):
+    flag = 0
+    for i in range(len(users)):
+        if users[i] == item:
+            users[i] = users[-1]
+            users[-1] = feature_mask
+            flag = 1
+            break
+    return len(users) - flag
+
+
+def _add_mask(feature_mask, features, num_max):
+    # uniformalize the length of each batch
+    for i in range(len(features)):
+        features[i] = features[i] + [feature_mask] * (num_max + 1 - len(features[i]))
+    return features
diff --git a/recommendation/Basic-NAIS-Demo/Dataset.py b/recommendation/Basic-NAIS-Demo/Dataset.py
new file mode 100644
index 00000000..1c0f01bd
--- /dev/null
+++ b/recommendation/Basic-NAIS-Demo/Dataset.py
@@ -0,0 +1,96 @@
+import scipy.sparse as sp
+import numpy as np
+
+ITEM_CLIP = 300
+
+class Dataset(object):
+
+    def __init__(self,path):
+        self.trainMatrix = self.load_training_file_as_matrix(path + ".train.rating")
+        self.trainList = self.load_training_file_as_list(path + ".train.rating")
+        self.testRatings = self.load_rating_file_as_list(path + '.test.rating')
+        self.testNegatives = self.load_negative_file(path + ".test.negative")
+        assert len(self.testRatings) == len(self.testNegatives)
+        self.num_users, self.num_items = self.trainMatrix.shape
+
+
+
+    def load_negative_file(self,filename):
+        negativeList = []
+        with open(filename, "r") as f:
+            line = f.readline()
+            while line != None and line != "": # 一行是一个用户所有的neg
+                arr = line.split("\t")
+                negatives = []
+                for x in arr[1:]:
+                    negatives.append(int(x))
+                negativeList.append(negatives)
+                line = f.readline()
+        return negativeList
+
+    def load_rating_file_as_list(self,filename):
+        ratingList = []
+        with open(filename, "r") as f:
+            line = f.readline()
+            while line != None and line != "":
+                arr = line.split("\t")
+                user, item = int(arr[0]), int(arr[1])
+                ratingList.append([user, item])
+                line = f.readline()
+        return ratingList
+
+
+
+    def load_training_file_as_list(self,filename):
+        u_ = 0
+        lists, items = [], [] # 训练数据是按用户id排序过的
+        with open(filename, "r") as f:
+            line = f.readline()
+            index = 0
+            while line != None and line != "":
+                arr = line.split("\t")
+                u, i = int(arr[0]), int(arr[1])
+                if u_ < u:
+                    index = 0
+                    lists.append(items) # 每次的items是一个用户所有打过分的item
+                    items = []
+                    u_ += 1
+                index += 1
+                if index < ITEM_CLIP:
+                    items.append(i)
+                line = f.readline()
+        lists.append(items)
+        print("already load the trainList...")
+        return lists
+
+
+
+
+    def load_training_file_as_matrix(self,filename):
+
+        num_users,num_items = 0,0
+        with open(filename,"r") as f:
+            line = f.readline()
+            while line != None and line != "":
+                arr = line.split("\t")
+
+                u,i = int(arr[0]),int(arr[1])
+
+                num_users = max(num_users,u)
+                num_items = max(num_items,i)
+                line = f.readline()
+
+
+        mat = sp.dok_matrix((num_users+1,num_items+1),dtype=np.float32)
+        with open(filename,"r") as f:
+            line = f.readline()
+            while line != None and line != "":
+                arr = line.split("\t")
+                user,item,rating = int(arr[0]),int(arr[1]),float(arr[2])
+
+                if rating > 0:
+                    mat[user,item] = 1.0
+                line = f.readline()
+
+        print("already load the trainMatrix...")
+        return mat
\ No newline at end of file
diff --git a/recommendation/Basic-NAIS-Demo/Evaluate.py b/recommendation/Basic-NAIS-Demo/Evaluate.py
new file mode 100644
index 00000000..05dfc67f
--- /dev/null
+++ b/recommendation/Basic-NAIS-Demo/Evaluate.py
@@ -0,0 +1,117 @@
+'''
+Created on Apr 15, 2016
+Evaluate the performance of Top-K recommendation:
+    Protocol: leave-1-out evaluation
+    Measures: Hit Ratio and NDCG
+    (more details are in: Xiangnan He, et al. Fast Matrix Factorization for Online Recommendation with Implicit Feedback. SIGIR'16)
+@author: hexiangnan
+'''
+import math
+import heapq # for retrieval topK
+import multiprocessing
+import numpy as np
+from time import time
+#from numba import jit, autojit
+
+# Global variables that are shared across processes
+_model = None
+_testRatings = None
+_testNegatives = None
+_K = None
+_DictList = None
+_sess = None
+
+def init_evaluate_model(model, sess, testRatings, testNegatives, trainList):
+    """
+    Evaluate the performance (Hit_Ratio, NDCG) of top-K recommendation
+    Return: score of each test rating.
+    """
+    global _model
+    global _testRatings
+    global _testNegatives
+    global _trainList
+    global _DictList
+    global _sess
+    _sess = sess
+    _model = model
+    _testRatings = testRatings
+    _testNegatives = testNegatives
+    _trainList = trainList
+    return load_test_as_list()
+
+def eval(model, sess, testRatings, testNegatives, DictList):
+
+    global _model
+    global _testRatings
+    global _testNegatives
+    global _K
+    global _DictList
+    global _sess
+    _model = model
+    _testRatings = testRatings
+    _testNegatives = testNegatives
+    _DictList = DictList
+    _sess = sess
+    _K = 10
+
+    num_thread = 1#multiprocessing.cpu_count()
+    hits, ndcgs, losses = [],[],[]
+    if(num_thread > 1): # Multi-thread
+        pool = multiprocessing.Pool(num_thread)
+        res = pool.map(_eval_one_rating, range(len(_testRatings)))
+        pool.close()
+        pool.join()
+        hits = [r[0] for r in res]
+        ndcgs = [r[1] for r in res]
+        losses = [r[2] for r in res]
+    # Single thread
+    else:
+        for idx in range(len(_testRatings)):
+            (hr,ndcg, loss) = _eval_one_rating(idx)
+            hits.append(hr)
+            ndcgs.append(ndcg)
+            losses.append(loss)
+    return (hits, ndcgs, losses)
+
+def load_test_as_list():
+    DictList = []
+    for idx in range(len(_testRatings)):
+        rating = _testRatings[idx]
+        items = _testNegatives[idx]
+        user = _trainList[idx]
+        num_idx_ = len(user)
+        gtItem = rating[1]
+        items.append(gtItem)
+        # Get prediction scores
+        num_idx = np.full(len(items),num_idx_, dtype=np.int32 )[:,None]
+        user_input = []
+        for i in range(len(items)):
+            user_input.append(user)
+        user_input = np.array(user_input)
+        item_input = np.array(items)[:,None]
+        feed_dict = {_model.user_input: user_input, _model.num_idx: num_idx, _model.item_input: item_input}
+        DictList.append(feed_dict)
+    print("already load the evaluate model...")
+    return DictList
+
+def _eval_one_rating(idx):
+
+    map_item_score = {}
+    rating = _testRatings[idx]
+    items = _testNegatives[idx]
+    gtItem = rating[1]
+    labels = np.zeros(len(items))[:, None]
+    labels[-1] = 1
+    feed_dict = _DictList[idx]
+    feed_dict[_model.labels] = labels
+    predictions,loss = _sess.run([_model.output,_model.loss], feed_dict = feed_dict)
+
+    neg_predict, pos_predict = predictions[:-1], predictions[-1]
+    position = (neg_predict >= pos_predict).sum()
+
+    # calculate HR@10, NDCG@10, AUC
+    hr = position < _K
+    ndcg = math.log(2) / math.log(position+2) if hr else 0
+
+    return (hr, ndcg, loss)
+
diff --git a/recommendation/Basic-NAIS-Demo/FISM.py b/recommendation/Basic-NAIS-Demo/FISM.py
new file mode 100644
index 00000000..5fd4f462
--- /dev/null
+++ b/recommendation/Basic-NAIS-Demo/FISM.py
@@ -0,0 +1,226 @@
+from __future__ import absolute_import
+from __future__ import division
+
+import os
+
+os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
+
+import cProfile
+import tensorflow as tf
+import numpy as np
+import logging
+
+from time import time
+from time import strftime
+from time import localtime
+
+from Dataset import Dataset
+import Batch_gen as data
+import Evaluate as evaluate
+
+import argparse
+
+
+##################################
+# 1. change the evaluation ranking
+# 2. regularization changing
+#
+
+
+
+def parse_args():
+    parser = argparse.ArgumentParser(description="Run FISM.")
+    parser.add_argument('--path', nargs='?', default='Data/',
+                        help='Input data path.')
+    parser.add_argument('--dataset', nargs='?', default='pinterest-20',
+                        help='Choose a dataset.')
+    parser.add_argument('--epochs', type=int, default=500,
+                        help='Number of epochs.')
+    parser.add_argument('--verbose', type=int, default=1,
+                        help='Interval of evaluation.')
+    parser.add_argument('--batch_choice', nargs='?', default='user',
+                        help='user: generate batches by user, fixed:batch_size: generate batches by batch size')
+    parser.add_argument('--embed_size', type=int, default=16,
+                        help='Embedding size.')
+    parser.add_argument('--regs', nargs='?', default='[1e-7,1e-7]',
+                        help='Regularization for user and item embeddings.')
+    parser.add_argument('--alpha', type=float, default=0,
+                        help='Index of coefficient of embedding vector')
+    parser.add_argument('--train_loss', type=float, default=1,
+                        help='Caculate training loss or not')
+    parser.add_argument('--num_neg', type=int, default=4,
+                        help='Number of negative instances to pair with a positive instance.')
+    parser.add_argument('--lr', type=float, default=0.01,
+                        help='Learning rate.')
+    parser.add_argument('--pretrain', type=int, default=0,
+                        help='0: No pretrain, 1: Pretrain with updating FISM variables, 2:Pretrain with fixed FISM variables.')
+    return parser.parse_args()
+
+
+class FISM:
+    def __init__(self, num_items, args):
+        self.num_items = num_items
+        self.dataset_name = args.dataset
+        self.learning_rate = args.lr
+        self.embedding_size = args.embed_size
+        self.alpha = args.alpha
+        self.verbose = args.verbose
+        regs = eval(args.regs)
+        self.lambda_bilinear = regs[0]
+        self.gamma_bilinear = regs[1]
+        self.batch_choice = args.batch_choice
+        self.train_loss = args.train_loss
+
+    def _create_placeholders(self):
+        with tf.name_scope("input_data"):
+            self.user_input = tf.placeholder(tf.int32, shape=[None, None])  # the index of users
+            self.num_idx = tf.placeholder(tf.float32, shape=[None, 1])  # the number of items rated by users
+            self.item_input = tf.placeholder(tf.int32, shape=[None, 1])  # the index of items
+            self.labels = tf.placeholder(tf.float32, shape=[None, 1])  # the ground truth
+
+    def _create_variables(self):
+        with tf.name_scope("embedding"):  # The embedding initialization is unknown now
+            self.c1 = tf.Variable(
+                tf.truncated_normal(shape=[self.num_items, self.embedding_size], mean=0.0, stddev=0.01),
+                # why [0, 3707)?
+                name='c1', dtype=tf.float32)
+            self.c2 = tf.constant(0.0, tf.float32, [1, self.embedding_size], name='c2')
+            self.embedding_Q_ = tf.concat([self.c1, self.c2], 0, name='embedding_Q_')
+            self.embedding_Q = tf.Variable(
+                tf.truncated_normal(shape=[self.num_items, self.embedding_size], mean=0.0, stddev=0.01),
+                name='embedding_Q', dtype=tf.float32)
+            self.bias = tf.Variable(tf.zeros(self.num_items), name='bias')
+
+    def _create_inference(self):
+        with tf.name_scope("inference"):
+            self.embedding_p = tf.reduce_sum(tf.nn.embedding_lookup(self.embedding_Q_, self.user_input), 1)
+            self.embedding_q = tf.reduce_sum(tf.nn.embedding_lookup(self.embedding_Q, self.item_input), 1)
+            self.bias_i = tf.nn.embedding_lookup(self.bias, self.item_input)
+            self.coeff = tf.pow(self.num_idx, -tf.constant(self.alpha, tf.float32, [1]))
+            self.output = tf.sigmoid(
+                self.coeff * tf.expand_dims(tf.reduce_sum(self.embedding_p * self.embedding_q, 1), 1) + self.bias_i)
+
+    def _create_loss(self):
+        with tf.name_scope("loss"):
+            self.loss = tf.losses.log_loss(self.labels, self.output) + \
+                        self.lambda_bilinear * tf.reduce_sum(
+                            tf.square(self.embedding_Q)) + self.gamma_bilinear * tf.reduce_sum(
+                tf.square(self.embedding_Q_))
+            # self.loss = tf.nn.l2_loss(self.labels - self.output) + \
+            #             self.lambda_bilinear * tf.reduce_sum(tf.square(self.embedding_Q)) + self.gamma_bilinear * tf.reduce_sum(tf.square(self.embedding_Q_))
+
+    def _create_optimizer(self):
+        with tf.name_scope("optimizer"):
+            self.optimizer = tf.train.AdagradOptimizer(learning_rate=self.learning_rate,
+                                                       initial_accumulator_value=1e-8).minimize(self.loss)
+
+    def build_graph(self):
+        self._create_placeholders()
+        self._create_variables()
+        self._create_inference()
+        self._create_loss()
+        self._create_optimizer()
+        logging.info("already build the computing graph...")
+
+
+def training(flag, model, dataset, epochs, num_negatives):
+    saver = tf.train.Saver({'c1': model.c1, 'embedding_Q': model.embedding_Q, 'bias': model.bias})
+    weight_path = 'Pretraining/%s/alpha%.1f' % (model.dataset_name, model.alpha)
+
+    with tf.Session() as sess:
+        # pretrain nor not
+        if flag != 0:
+            ckpt = tf.train.get_checkpoint_state(os.path.dirname(weight_path + '/checkpoint'))
+            if ckpt and ckpt.model_checkpoint_path:
+                sess.run(tf.global_variables_initializer())
+                saver.restore(sess, ckpt.model_checkpoint_path)
+                logging.info("using pretrained variables")
+                print("using pretrained variables")
+        else:
+            sess.run(tf.global_variables_initializer())
+            logging.info("initialized")
+            print("initialized")
+
+        # initialize for training batches
+        batch_begin = time()
+        batches = data.shuffle(dataset, model.batch_choice, num_negatives)
+        batch_time = time() - batch_begin
+
+        num_batch = len(batches[1])
+        batch_index = range(num_batch)
+
+        # initialize the evaluation feed_dicts
+        testDict = evaluate.init_evaluate_model(model, sess, dataset.testRatings, dataset.testNegatives,
+                                                dataset.trainList)
+
+        # train by epoch
+        for epoch_count in range(epochs):
+
+            train_begin = time()
+            training_batch(batch_index, model, sess, batches)
+            train_time = time() - train_begin
+
+            if epoch_count % model.verbose == 0:
+
+                if model.train_loss:
+                    loss_begin = time()
+                    train_loss = training_loss(model, sess, batches)
+                    loss_time = time() - loss_begin
+                else:
+                    loss_time, train_loss = 0, 0
+
+                eval_begin = time()
+                (hits, ndcgs, losses) = evaluate.eval(model, sess, dataset.testRatings, dataset.testNegatives, testDict)
+                hr, ndcg, test_loss = np.array(hits).mean(), np.array(ndcgs).mean(), np.array(losses).mean()
+                eval_time = time() - eval_begin
+
+                logging.info(
+                    "Epoch %d [%.1fs + %.1fs]: HR = %.4f, NDCG = %.4f, loss = %.4f [%.1fs] train_loss = %.4f [%.1fs]" % (
+                        epoch_count, batch_time, train_time, hr, ndcg, test_loss, eval_time, train_loss, loss_time))
+                print("Epoch %d [%.1fs + %.1fs]: HR = %.4f, NDCG = %.4f, loss = %.4f [%.1fs] train_loss = %.4f [%.1fs]" % (
+                    epoch_count, batch_time, train_time, hr, ndcg, test_loss, eval_time, train_loss, loss_time))
+
+            batch_begin = time()
+            batches = data.shuffle(dataset, model.batch_choice, num_negatives)
+            np.random.shuffle(batch_index)
+            batch_time = time() - batch_begin
+
+
+def training_batch(batch_index, model, sess, batches):
+    for index in batch_index:
+        user_input, num_idx, item_input, labels = data.batch_gen(batches, index)
+        feed_dict = {model.user_input: user_input, model.num_idx: num_idx[:, None],
+                     model.item_input: item_input[:, None],
+                     model.labels: labels[:, None]}
+        sess.run(model.optimizer, feed_dict)
+
+
+def training_loss(model, sess, batches):
+    train_loss = 0.0
+    num_batch = len(batches[1])
+    for index in range(num_batch):
+        user_input, num_idx, item_input, labels = data.batch_gen(batches, index)
+        feed_dict = {model.user_input: user_input, model.num_idx: num_idx[:, None],
+                     model.item_input: item_input[:, None], model.labels: labels[:, None]}
+        train_loss += sess.run(model.loss, feed_dict)
+    return train_loss / num_batch
+
+
+if __name__ == '__main__':
+
+    args = parse_args()
+    regs = eval(args.regs)
+
+    log_dir = "Log/%s/" % args.dataset
+    if not os.path.exists(log_dir):
+        os.makedirs(log_dir)
+    logging.basicConfig(filename=os.path.join(log_dir, "log_pre%dembed_size%d_reg1%.7f_reg2%.7f_%s" % (
+    args.pretrain, args.embed_size, regs[0], regs[1], strftime('%Y-%m-%d%H:%M:%S', localtime()))), level=logging.INFO)
+
+    logging.info("begin training FISM model ......")
+    logging.info(args)
+
+    dataset = Dataset(args.path + args.dataset)
+    model = FISM(dataset.num_items, args)
+    model.build_graph()
+    training(args.pretrain, model, dataset, args.epochs, args.num_neg)
diff --git a/recommendation/Basic-NAIS-Demo/NAIS.py b/recommendation/Basic-NAIS-Demo/NAIS.py
new file mode 100644
index 00000000..5546f96f
--- /dev/null
+++ b/recommendation/Basic-NAIS-Demo/NAIS.py
@@ -0,0 +1,289 @@
+import tensorflow as tf
+import numpy as np
+
+import logging
+
+import os
+
+from time import time,strftime,localtime
+
+from Dataset import Dataset
+
+import Batch_gen as data
+import Evaluate as evaluate
+
+import argparse
+
+def parse_args():
+    parser = argparse.ArgumentParser(description="Run NAIS.")
+    parser.add_argument('--path', nargs='?', default='Data/',
+                        help='Input data path.')
+    parser.add_argument('--dataset', nargs='?', default='pinterest-20',
+                        help='Choose a dataset.')
+    parser.add_argument('--pretrain', type=int, default=1,
+                        help='0: No pretrain, 1: Pretrain with updating FISM variables, 2:Pretrain with fixed FISM variables.')
+    parser.add_argument('--verbose', type=int, default=1,
+                        help='Interval of evaluation.')
+    parser.add_argument('--batch_choice', nargs='?', default='user',
+                        help='user: generate batches by user, fixed:batch_size: generate batches by batch size')
+    parser.add_argument('--epochs', type=int, default=100,
+                        help='Number of epochs.')
+    parser.add_argument('--weight_size', type=int, default=16,
+                        help='weight size.')
+    parser.add_argument('--embed_size', type=int, default=16,
+                        help='Embedding size.')
+    parser.add_argument('--data_alpha', type=float, default=0,
+                        help='Index of coefficient of embedding vector')
+    parser.add_argument('--regs', nargs='?', default='[1e-7,1e-7,1e-5]',
+                        help='Regularization for user and item embeddings.')
+    parser.add_argument('--alpha', type=float, default=0,
+                        help='Index of coefficient of embedding vector')
+    parser.add_argument('--train_loss', type=float, default=1,
+                        help='Caculate training loss or nor')
+    parser.add_argument('--beta', type=float, default=0.5,
+                        help='Index of coefficient of sum of exp(A)')
+    parser.add_argument('--num_neg', type=int, default=4,
+                        help='Number of negative instances to pair with a positive instance.')
+    parser.add_argument('--lr', type=float, default=0.01,
+                        help='Learning rate.')
+    parser.add_argument('--activation', type=int, default=0,
+                        help='Activation for ReLU, sigmoid, tanh.')
+    parser.add_argument('--algorithm', type=int, default=0,
+                        help='0 for NAIS_prod, 1 for NAIS_concat')
+    return parser.parse_args()
+
+class NAIS(object):
+    def __init__(self,num_items,args):
+        self.pretrain = args.pretrain
+        self.num_items = num_items
+        self.dataset_name = args.dataset
+        self.learning_rate = args.lr
+        self.embedding_size = args.embed_size
+        self.weight_size = args.weight_size
+        self.alpha = args.alpha
+        self.beta = args.beta
+        self.data_alpha = args.data_alpha
+        self.verbose = args.verbose
+        self.activation = args.activation
+        self.algorithm = args.algorithm
+        self.batch_choice = args.batch_choice
+        regs = eval(args.regs)
+        self.lambda_bilinear = regs[0]
+        self.gamma_bilinear = regs[1]
+        self.eta_bilinear = regs[2]
+        self.train_loss = args.train_loss
+
+
+    def _create_placeholders(self):
+        with tf.name_scope("input_data"):
+            self.user_input = tf.placeholder(tf.int32, shape=[None, None])  # the index of users
+            self.num_idx = tf.placeholder(tf.float32, shape=[None, 1])  # the number of items rated by users
+            self.item_input = tf.placeholder(tf.int32, shape=[None, 1])  # the index of items
+            self.labels = tf.placeholder(tf.float32, shape=[None, 1])  # the ground truth
+
+    def _create_variables(self):
+        with tf.name_scope("embedding"):
+            trainable_flag = (self.pretrain != 2)
+            self.c1 = tf.Variable(
+                tf.truncated_normal(shape=[self.num_items, self.embedding_size], mean=0.0, stddev=0.01), name='c1',
+                dtype=tf.float32, trainable=trainable_flag)
+            self.c2 = tf.constant(0.0, tf.float32, [1, self.embedding_size], name='c2')
+            self.embedding_Q_ = tf.concat([self.c1, self.c2], 0, name='embedding_Q_')
+            self.embedding_Q = tf.Variable(
+                tf.truncated_normal(shape=[self.num_items, self.embedding_size], mean=0.0, stddev=0.01),
+                name='embedding_Q', dtype=tf.float32, trainable=trainable_flag)
+            self.bias = tf.Variable(tf.zeros(self.num_items), name='bias', trainable=trainable_flag)
+
+            # Variables for attention,两种计算attention的方式
+            if self.algorithm == 0:
+                self.W = tf.Variable(tf.truncated_normal(shape=[self.embedding_size, self.weight_size], mean=0.0,
+                                                         stddev=tf.sqrt(
+                                                             tf.div(2.0, self.weight_size + self.embedding_size))),
+                                     name='Weights_for_MLP', dtype=tf.float32, trainable=True)
+            else:
+                self.W = tf.Variable(tf.truncated_normal(shape=[2 * self.embedding_size, self.weight_size], mean=0.0,
+                                                         stddev=tf.sqrt(tf.div(2.0, self.weight_size + (
+                                                         2 * self.embedding_size)))), name='Weights_for_MLP',
+                                     dtype=tf.float32, trainable=True)
+            self.b = tf.Variable(tf.truncated_normal(shape=[1, self.weight_size], mean=0.0, stddev=tf.sqrt(
+                tf.div(2.0, self.weight_size + self.embedding_size))), name='Bias_for_MLP', dtype=tf.float32,
+                                 trainable=True)
+            self.h = tf.Variable(tf.ones([self.weight_size, 1]), name='H_for_MLP', dtype=tf.float32)
+
+    def _attention_MLP(self,q_):
+        with tf.name_scope("attention_MLP"):
+            b = tf.shape(q_)[0]
+            n = tf.shape(q_)[1]
+            r = (self.algorithm + 1) * self.embedding_size
+
+            # 得到attention 的值
+            MLP_output = tf.matmul(tf.reshape(q_, [-1, r]), self.W) + self.b  # (b*n, e or 2*e) * (e or 2*e, w) + (1, w)
+            if self.activation == 0:
+                MLP_output = tf.nn.relu(MLP_output)
+            elif self.activation == 1:
+                MLP_output = tf.nn.sigmoid(MLP_output)
+            elif self.activation == 2:
+                MLP_output = tf.nn.tanh(MLP_output)
+
+            A_ = tf.reshape(tf.matmul(MLP_output, self.h), [b, n])  # (b*n, w) * (w, 1) => (None, 1) => (b, n)
+
+            # softmax for not mask features
+            exp_A_ = tf.exp(A_)
+            num_idx = tf.reduce_sum(self.num_idx, 1) # 每个人有多少个评分
+            mask_mat = tf.sequence_mask(num_idx, maxlen=n, dtype=tf.float32)  # (b, n) 加入mask
+            exp_A_ = mask_mat * exp_A_ # 填充的位置变为0
+            exp_sum = tf.reduce_sum(exp_A_, 1, keep_dims=True)  # (b, 1)
+            exp_sum = tf.pow(exp_sum, tf.constant(self.beta, tf.float32, [1])) # 进行压缩
+
+            A = tf.expand_dims(tf.div(exp_A_, exp_sum), 2)  # (b, n, 1)
+
+            return tf.reduce_sum(A * self.embedding_q_, 1) # 得到
+
+    def _create_inference(self):
+        with tf.name_scope("inference"):
+            self.embedding_q_ = tf.nn.embedding_lookup(self.embedding_Q_, self.user_input)  # (b, n, e) # 历史评分
+            self.embedding_q = tf.nn.embedding_lookup(self.embedding_Q, self.item_input)  # (b, 1, e) # 目标
+
+            if self.algorithm == 0:
+                self.embedding_p = self._attention_MLP(self.embedding_q_ * self.embedding_q)
+            else:
+                n = tf.shape(self.user_input)[1]
+                self.embedding_p = self._attention_MLP(
+                    tf.concat([self.embedding_q_, tf.tile(self.embedding_q, tf.stack([1, n, 1]))], 2))
+
+            self.embedding_q = tf.reduce_sum(self.embedding_q, 1)
+            self.bias_i = tf.nn.embedding_lookup(self.bias, self.item_input)
+            self.coeff = tf.pow(self.num_idx, tf.constant(self.alpha, tf.float32, [1]))
+            self.output = tf.sigmoid(
+                self.coeff * tf.expand_dims(tf.reduce_sum(self.embedding_p * self.embedding_q, 1), 1) + self.bias_i)
+
+    def _create_loss(self):
+        with tf.name_scope("loss"):
+            self.loss = tf.losses.log_loss(self.labels, self.output) + \
+                        self.lambda_bilinear * tf.reduce_sum(tf.square(self.embedding_Q)) + \
+                        self.gamma_bilinear * tf.reduce_sum(tf.square(self.embedding_Q_)) + \
+                        self.eta_bilinear * tf.reduce_sum(tf.square(self.W)) # logloss + l2正则
+
+    def _create_optimizer(self):
+        with tf.name_scope("optimizer"):
+            self.optimizer = tf.train.AdagradOptimizer(learning_rate=self.learning_rate,
+                                                       initial_accumulator_value=1e-8).minimize(self.loss)
+
+    def build_graph(self):
+        self._create_placeholders()
+        self._create_variables()
+        self._create_inference()
+        self._create_loss()
+        self._create_optimizer()
+        logging.info("already build the computing graph...")
+
+
+def training(flag, model, dataset, epochs, num_negatives):
+    saver = tf.train.Saver({'c1': model.c1, 'embedding_Q': model.embedding_Q, 'bias': model.bias})
+    weight_path = 'Pretraining/%s/alpha%.1f' % (model.dataset_name, model.data_alpha)
+
+    with tf.Session() as sess:
+        # pretrain nor not
+        if flag != 0:
+            ckpt = tf.train.get_checkpoint_state(os.path.dirname(weight_path + '/checkpoint'))
+            if ckpt and ckpt.model_checkpoint_path:
+                sess.run(tf.global_variables_initializer())
+                saver.restore(sess, ckpt.model_checkpoint_path)
+                logging.info("using pretrained variables")
+                print("using pretrained variables")
+        else:
+            sess.run(tf.global_variables_initializer())
+            logging.info("initialized")
+            print("initialized")
+
+        # initialize for training batches
+        batch_begin = time()
+        batches = data.shuffle(dataset, model.batch_choice, num_negatives)
+        batch_time = time() - batch_begin
+
+        num_batch = len(batches[1])
+        batch_index = range(num_batch)
+
+        # initialize the evaluation feed_dicts
+        testDict = evaluate.init_evaluate_model(model, sess, dataset.testRatings, dataset.testNegatives,
+                                                dataset.trainList)
+
+        # train by epoch
+        for epoch_count in range(epochs):
+
+            train_begin = time()
+            training_batch(batch_index, model, sess, batches)
+            train_time = time() - train_begin
+
+            if epoch_count % model.verbose == 0:
+
+                if model.train_loss:
+                    loss_begin = time()
+                    train_loss = training_loss(model, sess, batches)
+                    loss_time = time() - loss_begin
+                else:
+                    loss_time, train_loss = 0, 0
+
+                eval_begin = time()
+                (hits, ndcgs, losses) = evaluate.eval(model, sess, dataset.testRatings, dataset.testNegatives, testDict)
+                hr, ndcg, test_loss = np.array(hits).mean(), np.array(ndcgs).mean(), np.array(losses).mean()
+                eval_time = time() - eval_begin
+
+                logging.info(
+                    "Epoch %d [%.1fs + %.1fs]: HR = %.4f, NDCG = %.4f, loss = %.4f [%.1fs] train_loss = %.4f [%.1fs]" % (
+                        epoch_count, batch_time, train_time, hr, ndcg, test_loss, eval_time, train_loss, loss_time))
+                print("Epoch %d [%.1fs + %.1fs]: HR = %.4f, NDCG = %.4f, loss = %.4f [%.1fs] train_loss = %.4f [%.1fs]" % (
+                    epoch_count, batch_time, train_time, hr, ndcg, test_loss, eval_time, train_loss, loss_time))
+
+            batch_begin = time()
+            batches = data.shuffle(dataset, model.batch_choice, num_negatives)
+            np.random.shuffle(batch_index)
+            batch_time = time() - batch_begin
+
+
+def training_batch(batch_index, model, sess, batches):
+    for index in batch_index:
+        user_input, num_idx, item_input, labels = data.batch_gen(batches, index)
+        feed_dict = {model.user_input: user_input, model.num_idx: num_idx[:, None],
+                     model.item_input: item_input[:, None],
+                     model.labels: labels[:, None]}
+        sess.run([model.loss, model.optimizer], feed_dict)
+        # Q_, Q, W,b,h = sess.run([model.embedding_Q_, model.embedding_Q, model.W, model.b, model.h], feed_dict)
+        # print "Q_: %.8f, Q: %.8f, \nW: %.8f, b: %.8f, h:%.8f" % (np.sum(np.square(Q_)), np.sum(np.square(Q)), np.sum(np.square(W)), np.sum(np.square(b)), np.sum(np.square(h)))
+
+
+def training_loss(model, sess, batches):
+    train_loss = 0.0
+    num_batch = len(batches[1])
+    for index in range(num_batch):
+        user_input, num_idx, item_input, labels = data.batch_gen(batches, index)
+        feed_dict = {model.user_input: user_input, model.num_idx: num_idx[:, None],
+                     model.item_input: item_input[:, None], model.labels: labels[:, None]}
+        train_loss += sess.run(model.loss, feed_dict)
+    return train_loss / num_batch
+
+
+if __name__ == '__main__':
+
+    args = parse_args()
+    regs = eval(args.regs)
+
+    algo = "NAIS_concat" if args.algorithm else "NAIS_prod"
+
+    log_dir = "Log/%s/" % args.dataset
+
+    if not os.path.exists(log_dir):
+        os.makedirs(log_dir)
+    logging.basicConfig(filename=os.path.join(log_dir, "log_%s_dataset_%s_lr%.2f_reg%.0e_%s" %
+                                              (algo, args.dataset, args.lr, regs[2],
+                                               strftime('%Y-%m-%d%H:%M:%S', localtime()))), level=logging.INFO)
+
+    logging.info("begin training %s model ......" % algo)
+
+    print(args)
+    logging.info(args)
+
+    dataset = Dataset(args.path + args.dataset)
+    model = NAIS(dataset.num_items, args)
+    model.build_graph()
+    training(args.pretrain, model, dataset, args.epochs, args.num_neg)