Update app.py
Browse files
app.py
CHANGED
@@ -2,25 +2,23 @@ import os
|
|
2 |
import numpy as np
|
3 |
import pandas as pd
|
4 |
import streamlit as st
|
5 |
-
from
|
|
|
6 |
from keybert import KeyBERT
|
|
|
7 |
|
8 |
-
# ───
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
trust_remote_code=True
|
14 |
)
|
15 |
|
16 |
@st.cache_resource
|
17 |
def load_sentiment_pipeline():
|
18 |
model_name = "mayf/amazon_reviews_bert_ft"
|
19 |
tok = AutoTokenizer.from_pretrained(model_name, use_auth_token=True)
|
20 |
-
mdl = AutoModelForSequenceClassification.from_pretrained(
|
21 |
-
model_name,
|
22 |
-
use_auth_token=True
|
23 |
-
)
|
24 |
return pipeline(
|
25 |
"sentiment-analysis",
|
26 |
model=mdl,
|
@@ -51,6 +49,7 @@ def main():
|
|
51 |
st.warning("Please enter a review to analyze.")
|
52 |
return
|
53 |
|
|
|
54 |
progress = st.progress(0)
|
55 |
|
56 |
# Load models
|
@@ -59,15 +58,14 @@ def main():
|
|
59 |
kw_model = load_keybert_model()
|
60 |
progress.progress(20)
|
61 |
|
62 |
-
#
|
63 |
progress.text("Analyzing sentiment...")
|
64 |
raw_scores = sentiment_pipeline(review)[0]
|
65 |
-
|
66 |
-
|
67 |
-
}
|
68 |
progress.progress(40)
|
69 |
|
70 |
-
#
|
71 |
progress.text("Extracting keywords...")
|
72 |
keywords = kw_model.extract_keywords(
|
73 |
review,
|
@@ -77,34 +75,29 @@ def main():
|
|
77 |
)
|
78 |
progress.progress(60)
|
79 |
|
80 |
-
# Display
|
81 |
col1, col2 = st.columns(2)
|
82 |
with col1:
|
83 |
st.subheader("Sentiment Scores")
|
84 |
st.json({k: round(v, 4) for k, v in sentiment_results.items()})
|
85 |
with col2:
|
86 |
-
st.subheader("Top Keywords")
|
87 |
for kw, score in keywords:
|
88 |
st.write(f"• {kw} ({score:.4f})")
|
89 |
|
90 |
# Bar chart
|
91 |
progress.text("Rendering chart...")
|
92 |
-
df_scores = pd.DataFrame.from_dict(
|
93 |
-
sentiment_results, orient='index', columns=['score']
|
94 |
-
)
|
95 |
df_scores.index.name = 'Sentiment'
|
96 |
st.bar_chart(df_scores)
|
97 |
progress.progress(80)
|
98 |
|
99 |
# Highlight highest sentiment
|
100 |
-
max_label, max_score = max(
|
101 |
-
sentiment_results.items(), key=lambda x: x[1]
|
102 |
-
)
|
103 |
st.markdown(f"**Highest Sentiment:** **{max_label}** ({max_score:.4f})")
|
104 |
|
105 |
# GPT-Driven Analysis & Suggestions
|
106 |
progress.text("Generating insights...")
|
107 |
-
# Build the prompt
|
108 |
prompt = f"""
|
109 |
You are an analytical amazon feedback expert.
|
110 |
Review: \"{review}\"
|
@@ -114,24 +107,23 @@ Tasks:
|
|
114 |
1. Analysis: Write a concise paragraph (3 sentences) interpreting customer sentiment by combining the scores and keywords.
|
115 |
2. Recommendations: Three separate paragraphs with actionable suggestions (max 30 words each).
|
116 |
"""
|
117 |
-
# Prepare chat messages
|
118 |
-
chat_input = [
|
119 |
-
{"role": "system", "content": "You are a product-feedback analyst."},
|
120 |
-
{"role": "user", "content": prompt}
|
121 |
-
]
|
122 |
-
# Flatten into a single text prompt
|
123 |
-
flat_prompt = "\n".join(
|
124 |
-
f"{msg['role'].upper()}: {msg['content']}" for msg in chat_input
|
125 |
-
)
|
126 |
-
# Generate
|
127 |
-
gen_output = pipe(flat_prompt, max_new_tokens=200)
|
128 |
-
gpt_reply = gen_output[0]['generated_text']
|
129 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
130 |
st.markdown(gpt_reply)
|
|
|
|
|
131 |
progress.progress(100)
|
132 |
progress.text("Done!")
|
133 |
|
134 |
if __name__ == "__main__":
|
135 |
main()
|
136 |
|
137 |
-
|
|
|
2 |
import numpy as np
|
3 |
import pandas as pd
|
4 |
import streamlit as st
|
5 |
+
from huggingface_hub import login
|
6 |
+
from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassification
|
7 |
from keybert import KeyBERT
|
8 |
+
from openai import AzureOpenAI # new
|
9 |
|
10 |
+
# ─── Azure OpenAI Client ─────────────────────────────────────────────────────
|
11 |
+
openai_client = AzureOpenAI(
|
12 |
+
api_key = "fbca46bfd8814334be46a2e5c323904c", # use your key here
|
13 |
+
api_version = "2023-05-15", # apparently HKUST uses a deprecated version
|
14 |
+
azure_endpoint = "https://hkust.azure-api.net" # per HKUST instructions
|
|
|
15 |
)
|
16 |
|
17 |
@st.cache_resource
|
18 |
def load_sentiment_pipeline():
|
19 |
model_name = "mayf/amazon_reviews_bert_ft"
|
20 |
tok = AutoTokenizer.from_pretrained(model_name, use_auth_token=True)
|
21 |
+
mdl = AutoModelForSequenceClassification.from_pretrained(model_name, use_auth_token=True)
|
|
|
|
|
|
|
22 |
return pipeline(
|
23 |
"sentiment-analysis",
|
24 |
model=mdl,
|
|
|
49 |
st.warning("Please enter a review to analyze.")
|
50 |
return
|
51 |
|
52 |
+
# Initialize progress bar
|
53 |
progress = st.progress(0)
|
54 |
|
55 |
# Load models
|
|
|
58 |
kw_model = load_keybert_model()
|
59 |
progress.progress(20)
|
60 |
|
61 |
+
# Run sentiment analysis
|
62 |
progress.text("Analyzing sentiment...")
|
63 |
raw_scores = sentiment_pipeline(review)[0]
|
64 |
+
# Map labels
|
65 |
+
sentiment_results = {LABEL_MAP[item['label']]: float(item['score']) for item in raw_scores}
|
|
|
66 |
progress.progress(40)
|
67 |
|
68 |
+
# Extract keywords
|
69 |
progress.text("Extracting keywords...")
|
70 |
keywords = kw_model.extract_keywords(
|
71 |
review,
|
|
|
75 |
)
|
76 |
progress.progress(60)
|
77 |
|
78 |
+
# Display scores and keywords side by side
|
79 |
col1, col2 = st.columns(2)
|
80 |
with col1:
|
81 |
st.subheader("Sentiment Scores")
|
82 |
st.json({k: round(v, 4) for k, v in sentiment_results.items()})
|
83 |
with col2:
|
84 |
+
st.subheader("Top 3 Keywords")
|
85 |
for kw, score in keywords:
|
86 |
st.write(f"• {kw} ({score:.4f})")
|
87 |
|
88 |
# Bar chart
|
89 |
progress.text("Rendering chart...")
|
90 |
+
df_scores = pd.DataFrame.from_dict(sentiment_results, orient='index', columns=['score'])
|
|
|
|
|
91 |
df_scores.index.name = 'Sentiment'
|
92 |
st.bar_chart(df_scores)
|
93 |
progress.progress(80)
|
94 |
|
95 |
# Highlight highest sentiment
|
96 |
+
max_label, max_score = max(sentiment_results.items(), key=lambda x: x[1])
|
|
|
|
|
97 |
st.markdown(f"**Highest Sentiment:** **{max_label}** ({max_score:.4f})")
|
98 |
|
99 |
# GPT-Driven Analysis & Suggestions
|
100 |
progress.text("Generating insights...")
|
|
|
101 |
prompt = f"""
|
102 |
You are an analytical amazon feedback expert.
|
103 |
Review: \"{review}\"
|
|
|
107 |
1. Analysis: Write a concise paragraph (3 sentences) interpreting customer sentiment by combining the scores and keywords.
|
108 |
2. Recommendations: Three separate paragraphs with actionable suggestions (max 30 words each).
|
109 |
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
|
111 |
+
response = openai_client.chat.completions.create(
|
112 |
+
model="gpt-35-turbo",
|
113 |
+
messages=[
|
114 |
+
{"role": "system", "content": "You are a product-feedback analyst."},
|
115 |
+
{"role": "user", "content": prompt}
|
116 |
+
],
|
117 |
+
temperature=0.7,
|
118 |
+
max_tokens=200
|
119 |
+
)
|
120 |
+
gpt_reply = response.choices[0].message.content.strip()
|
121 |
st.markdown(gpt_reply)
|
122 |
+
|
123 |
+
# Complete
|
124 |
progress.progress(100)
|
125 |
progress.text("Done!")
|
126 |
|
127 |
if __name__ == "__main__":
|
128 |
main()
|
129 |
|
|