Spaces:
Running
Running
File size: 9,310 Bytes
30c7f0c 9963876 30c7f0c 94a65e4 a355e59 94a65e4 207a2e4 5e58061 207a2e4 1a2b17f 30c7f0c 94a65e4 1a2b17f 94a65e4 5e58061 9963876 df7ee24 9963876 5e58061 9963876 5e58061 30c7f0c 94a65e4 5ac49d2 94a65e4 85e3b89 94a65e4 c422cb8 94a65e4 df7ee24 e9db226 df7ee24 94a65e4 df7ee24 21bc2b4 709d77e 15e4be7 94a65e4 c37cf7b 94a65e4 c37cf7b 94a65e4 df7ee24 94a65e4 21bc2b4 116a946 21bc2b4 94a65e4 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 |
import streamlit as st
import random
# Page configuration
st.set_page_config(
page_title="Nexus NLP News Classifier"
)
import pandas as pd
from final import *
from pydantic import BaseModel
import plotly.graph_objects as go
# Update the initialize_models function
@st.cache_resource
def initialize_models():
try:
nlp = spacy.load("en_core_web_sm")
except:
spacy.cli.download("en_core_web_sm")
nlp = spacy.load("en_core_web_sm")
model_path = "./results/checkpoint-753"
tokenizer = DebertaV2Tokenizer.from_pretrained('microsoft/deberta-v3-small')
model = AutoModelForSequenceClassification.from_pretrained(model_path)
model.eval()
knowledge_graph = load_knowledge_graph()
return nlp, tokenizer, model, knowledge_graph
class NewsInput(BaseModel):
text: str
def generate_knowledge_graph_viz(text, nlp, tokenizer, model):
kg_builder = KnowledgeGraphBuilder()
# Get prediction
prediction, _ = predict_with_model(text, tokenizer, model)
is_fake = prediction == "FAKE"
# Update knowledge graph
kg_builder.update_knowledge_graph(text, not is_fake, nlp)
# Randomly select subset of edges (e.g. 60% of edges)
edges = list(kg_builder.knowledge_graph.edges())
selected_edges = random.sample(edges, k=int(len(edges) * 0.3))
# Create a new graph with selected edges
selected_graph = nx.DiGraph()
selected_graph.add_nodes_from(kg_builder.knowledge_graph.nodes(data=True))
selected_graph.add_edges_from(selected_edges)
pos = nx.spring_layout(selected_graph)
edge_trace = go.Scatter(
x=[], y=[],
line=dict(
width=2,
color='rgba(255,0,0,0.7)' if is_fake else 'rgba(0,255,0,0.7)'
),
hoverinfo='none',
mode='lines'
)
# Create visualization
pos = nx.spring_layout(kg_builder.knowledge_graph)
edge_trace = go.Scatter(
x=[], y=[],
line=dict(
width=2,
color='rgba(255,0,0,0.7)' if is_fake else 'rgba(0,255,0,0.7)'
),
hoverinfo='none',
mode='lines'
)
node_trace = go.Scatter(
x=[], y=[],
mode='markers+text',
hoverinfo='text',
textposition='top center',
marker=dict(
size=15,
color='white',
line=dict(width=2, color='black')
),
text=[]
)
# Add edges
for edge in selected_graph.edges():
x0, y0 = pos[edge[0]]
x1, y1 = pos[edge[1]]
edge_trace['x'] += (x0, x1, None)
edge_trace['y'] += (y0, y1, None)
# Add nodes
for node in kg_builder.knowledge_graph.nodes():
x, y = pos[node]
node_trace['x'] += (x,)
node_trace['y'] += (y,)
node_trace['text'] += (node,)
fig = go.Figure(
data=[edge_trace, node_trace],
layout=go.Layout(
showlegend=False,
hovermode='closest',
margin=dict(b=0,l=0,r=0,t=0),
xaxis=dict(showgrid=False, zeroline=False, showticklabels=False),
yaxis=dict(showgrid=False, zeroline=False, showticklabels=False),
plot_bgcolor='rgba(0,0,0,0)',
paper_bgcolor='rgba(0,0,0,0)'
)
)
return fig
# Streamlit UI
def main():
st.title("Nexus NLP News Classifier")
st.write("Enter news text below to analyze its authenticity")
# Initialize models
nlp, tokenizer, model, knowledge_graph = initialize_models()
# Text input area
news_text = st.text_area("News Text", height=200)
if st.button("Analyze"):
if news_text:
with st.spinner("Analyzing..."):
# Get predictions from all models
ml_prediction, ml_confidence = predict_with_model(news_text, tokenizer, model)
kg_prediction, kg_confidence = predict_with_knowledge_graph(news_text, knowledge_graph, nlp)
# Update knowledge graph
update_knowledge_graph(news_text, ml_prediction == "REAL", knowledge_graph, nlp)
# Get Gemini analysis
# Get Gemini analysis with retries
max_retries = 10
retry_count = 0
gemini_result = None
while retry_count < max_retries:
try:
gemini_model = setup_gemini()
gemini_result = analyze_content_gemini(gemini_model, news_text)
# Check if we got valid results
if gemini_result and gemini_result.get('gemini_analysis'):
break
except Exception:
pass
retry_count += 1
# Use default values if all retries failed
if not gemini_result:
gemini_result = {
"gemini_analysis": {
"predicted_classification": "UNCERTAIN",
"confidence_score": "50",
"reasoning": ["Analysis temporarily unavailable"]
}
}
# Display metrics in columns
col1 = st.columns(1)[0]
with col1:
st.subheader("ML Model and Knowedge Graph Analysis")
st.metric("Prediction", ml_prediction)
st.metric("Confidence", f"{ml_confidence:.2f}%")
# with col2:
# st.subheader("Knowledge Graph Analysis")
# st.metric("Prediction", kg_prediction)
# st.metric("Confidence", f"{kg_confidence:.2f}%")
# with col3:
# st.subheader("Gemini Analysis")
# gemini_pred = gemini_result["gemini_analysis"]["predicted_classification"]
# gemini_conf = gemini_result["gemini_analysis"]["confidence_score"]
# st.metric("Prediction", gemini_pred)
# st.metric("Confidence", f"{gemini_conf}%")
# Single expander for all analysis details
with st.expander("Detailed Analysis"):
try:
# Text Classification
st.subheader("📝 Text Classification")
text_class = gemini_result.get('text_classification', {})
st.write(f"Category: {text_class.get('category', 'N/A')}")
st.write(f"Writing Style: {text_class.get('writing_style', 'N/A')}")
st.write(f"Target Audience: {text_class.get('target_audience', 'N/A')}")
st.write(f"Content Type: {text_class.get('content_type', 'N/A')}")
# Sentiment Analysis
st.subheader("🎭 Sentiment Analysis")
sentiment = gemini_result.get('sentiment_analysis', {})
st.write(f"Primary Emotion: {sentiment.get('primary_emotion', 'N/A')}")
st.write(f"Emotional Intensity: {sentiment.get('emotional_intensity', 'N/A')}/10")
st.write(f"Sensationalism Level: {sentiment.get('sensationalism_level', 'N/A')}")
st.write("Bias Indicators:", ", ".join(sentiment.get('bias_indicators', ['N/A'])))
# Entity Recognition
st.subheader("🔍 Entity Recognition")
entities = gemini_result.get('entity_recognition', {})
st.write(f"Source Credibility: {entities.get('source_credibility', 'N/A')}")
st.write("People:", ", ".join(entities.get('people', ['N/A'])))
st.write("Organizations:", ", ".join(entities.get('organizations', ['N/A'])))
# Named Entities from spaCy
st.subheader("🏷️ Named Entities")
entities = extract_entities(news_text, nlp)
df = pd.DataFrame(entities, columns=["Entity", "Type"])
st.dataframe(df)
# Knowledge Graph Visualization
st.subheader("🕸️ Knowledge Graph")
fig = generate_knowledge_graph_viz(news_text, nlp, tokenizer, model)
st.plotly_chart(fig, use_container_width=True)
# Analysis Reasoning
st.subheader("💭 Analysis Reasoning")
for point in gemini_result.get('gemini_analysis', {}).get('reasoning', ['N/A']):
st.write(f"• {point}")
except Exception as e:
st.error("Error processing analysis results")
else:
st.warning("Please enter some text to analyze")
if __name__ == "__main__":
main()
|