Hermes is Lab41's foray into recommender systems. It explores how to choose a recommender system for a new application by analyzing the performance of multiple recommender system algorithms on a variety of datasets.
It also explores how recommender systems may assist a software developer of data scientist find new data, tools, and computer programs.
The Wiki associated with this project has details on many references that we utilized when implementing this framework. It also details the datasets used in this base framework, as well as some resources to help you get started in recommender systems and Spark.
For tips on how to get started, see the wiki page: Running Hermes.
##Blog Overviews There are a number of blog articles that we produced during the course of this project. They include:
| Join the Hermes Running Club | March 2016 |
| Python2Vec: Word Embeddings for Source Code | March 2016 |
| TPS Report for Recommender Systems? Traditional Performance Metrics | March 2016 |
| Recommender Systems - It's Not All About the Accuracy | January 2016 |
| The Nine Must-Have Datasets for Investigating Recommender Systems | February 2016 |
| Recommending Recommendation Systems (project intro) | December 2015 |
We are trying varied tools and concepts to visualize the results of this project.
conda install bokeh- from top-level hermes folder
$bokeh serve src/results/hermes_run_view.py - view in browser at
http://localhost:5006/hermes_run_view
easy_install web.py- from viz folder
$python app.py - view in browser from location:port displayed in terminal