-
Notifications
You must be signed in to change notification settings - Fork 0
/
poc_2_survey_feedback_analysis.py
194 lines (163 loc) · 6.75 KB
/
poc_2_survey_feedback_analysis.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
# -*- coding: utf-8 -*-
"""POC_2_Survey_Feedback_Analysis.ipynb
Automatically generated by Colab.
Original file is located at
https://colab.research.google.com/drive/1dcgsH60VzZW1d-9Cz8NTt88N2_fS_QwQ
"""
from google.colab import drive
drive.mount('/content/drive')
pip install groq
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler, LabelEncoder, FunctionTransformer
from sklearn.metrics import accuracy_score, classification_report
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
from nltk.stem import WordNetLemmatizer
from imblearn.over_sampling import SMOTE
from sklearn.compose import ColumnTransformer
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
import re
import nltk
import os
#from langdetect import detect
#from googletrans import Translator
# Initialize Google Translator
#translator = Translator()
# Download necessary NLTK data
nltk.download('stopwords')
nltk.download('punkt')
nltk.download('wordnet')
from groq import Groq
!export GROQ_API_KEY='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'
client = Groq(
# This is the default and can be omitted
api_key='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'
#api_key=os.environ.get("GROQ_API_KEY"),
)
# Text preprocessing function
# Translate text if not in English
#def translate_text_if_needed(text):
# try:
# language = detect(text)
# if language != 'en':
# translated = translator.translate(text, dest='en')
# return translated.text
# return text
# except Exception as e:
# return text
def preprocess_text(text):
text = text.lower()
text = re.sub(r'\d+', '', text)
text = re.sub(r'\s+', ' ', text)
text = re.sub(r'[^\w\s]', '', text)
tokens = word_tokenize(text)
stop_words = set(stopwords.words('english'))
lemmatizer = WordNetLemmatizer()
filtered_text = ' '.join([lemmatizer.lemmatize(word) for word in tokens if word not in stop_words])
return filtered_text
def prepare_data(dataset_url):
df = pd.read_csv(dataset_url)
# preprocess data accroding to survey form
#df.drop('Timestamp = A12 + A4', axis=1, inplace=True)
df = df.dropna(subset=["Timestamp = A12 + A4"]) # if Timestamp == null, that means empty row, i.e, the form has never been filled
# Join all textual columns into one
feedback_columns = [col for col in df.columns if 'Overall feedback' in col]
df['All_feedback'] = df[feedback_columns].apply(lambda row: ' '.join(row.values.astype(str)), axis=1)
#df['All_feedback'] = df['All_feedback'].apply(translate_text_if_needed)
df.drop(feedback_columns, axis=1, inplace=True)
# Preprocess textual feedback
df['All_feedback'] = df['All_feedback'].apply(preprocess_text)
return df
def train_test_model(data):
X = data.drop('Final Label', axis=1)
y = data['Final Label']
# Train/Test Split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, shuffle=True)
# Define ColumnTransformer to separate numerical and text data processing
preprocessor = ColumnTransformer(
transformers=[
('num', StandardScaler(), X.select_dtypes(include='number').columns.tolist()),
('text', TfidfVectorizer(), 'All_feedback')
],
remainder='drop'
)
# Define the pipeline
pipeline = Pipeline([
('preprocessor', preprocessor),
('classifier', LogisticRegression(random_state=42))
])
pipeline.fit(X_train, y_train)
# Predict on the test set
y_pred = pipeline.predict(X_test)
#print("Accuracy:", accuracy_score(y_test, y_pred))
#print("Classification Report:")
#print(classification_report(y_test, y_pred))
return pipeline
def predict_feedback_proba(pipeline, sample):
sample_df = pd.DataFrame([sample])
sample_df = sample.to_frame().T
y_prob = pipeline.predict_proba(sample_df)
prob_pos = y_prob[0][1] * 100 # Probability of being positive
prob_neg = y_prob[0][0] * 100 # Probability of being negative
return prob_pos, prob_neg
# Function to generate detailed feedback using OpenAI's GPT-4
def generate_detailed_feedback(numerical_data, textual_feedback, prob_pos, prob_neg):
# Convert numerical_data to a string representation
numerical_data_str = str(numerical_data)
prompt = f"Using the following numerical ratings and feedback:\n\nNumerical Ratings: {numerical_data_str}\n\nTextual Feedback: {textual_feedback}\n\nProbability of being positive: {prob_pos:.2f}%\n\nProbability of being negative: {prob_neg:.2f}%\n\nPlease provide a very short final feedback according to the client, i.e, either positive or negative. Not both.Give output in the form of either Positive, Mostly Positive, Negative or Mostly Negative with the probability percentage. And then Improvements or Matters to focus upon:"
chat_completion = client.chat.completions.create(
messages = [
{
"role": "system",
"content": "You are an intelligent and helpful assistant.",
},
{
"role": "user",
"content": prompt, # Use the formatted prompt here
},
],
model="llama3-8b-8192" # Add the model keyword argument here
)
return chat_completion.choices[0].message.content.strip() # Extract the content from the response
def generate_feedback(pipeline, sample):
# Example usage
#prob_pos, prob_neg = predict_feedback_proba(pipeline, sample)
#print(f"Sample:\n{sample}")
#print(f"Probability of being positive: {prob_pos:.2f}%")
#print(f"Probability of being negative: {prob_neg:.2f}%")
numerical_data = sample.drop('All_feedback').to_dict()
textual_feedback = sample['All_feedback']
prob_pos, prob_neg = predict_feedback_proba(pipeline, sample)
detailed_feedback = generate_detailed_feedback(numerical_data, textual_feedback, prob_pos, prob_neg)
return detailed_feedback
def get_model_ready():
dataset_url = input("Enter the dataset URL: ")
df = prepare_data(dataset_url)
pipeline = train_test_model(df)
return pipeline
def get_feedback(pipeline):
feedform_url = input("Enter the URL where feedback data for generating analysis is stored: ")
data = prepare_data(feedform_url)
df = pd.read_csv(feedform_url)
df['Result'] = ""
for i in range(len(data)):
sample = data.iloc[i]
print("\nFeedback: ", i+1)
print("Client Name: ", data['Client Name'].iloc[i])
feed = generate_feedback(pipeline, sample)
df.at[i, 'Result'] = feed
print(feed)
#data = data.drop('All_feedback', axis=1)
df.to_csv(feedform_url, index=False)
def final_execute():
pipeline = get_model_ready()
get_feedback(pipeline)
def main():
final_execute()
if __name__ == "__main__":
main()