Spaces:
Sleeping
Sleeping
File size: 13,653 Bytes
dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 9df3fc4 dcb2841 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 | import streamlit as st
import pandas as pd
import plotly.express as px
def main():
st.title("📚 Project Documentation")
# Custom CSS for better styling
st.markdown("""
<style>
.question-card {
background-color: #f8f9fa;
padding: 20px;
border-radius: 10px;
border-left: 5px solid #1f77b4;
margin: 20px 0;
}
.question {
color: #1f77b4;
font-size: 1.2em;
font-weight: bold;
margin-bottom: 15px;
}
.answer {
color: #2c3e50;
line-height: 1.6;
}
</style>
""", unsafe_allow_html=True)
# Q1: Development Timeline
st.markdown("""
<div class="question-card">
<div class="question">⏱️ Q1: How long did it take to solve the problem?</div>
<div class="answer">
The solution was developed in approximately <b>5 hours</b> (excluding data collection and model training phases).
</div>
</div>
""", unsafe_allow_html=True)
# Q2: Solution Explanation
# Q2: Solution Explanation
st.markdown("""
<div class="question-card">
<div class="question">🔍 Q2: Can you explain your solution approach?</div>
<div class="answer">
The solution implements a multi-stage document classification pipeline:
<br><br>
<b>1. Data Collection & Processing:</b>
<ul>
<li>Dataset: 2500+ training URLs and 250+ test URLs</li>
<li>Implemented ThreadPooling with 20 workers for parallel processing</li>
<li>Reduced download time to ~40 minutes (vs. 3+ hours sequential)</li>
<li>Used PDFPlumber for robust text extraction</li>
</ul>
<br>
<b>2. Model Development Pipeline:</b>
<ul>
<li><i>Baseline Approach:</i>
<ul>
<li>TF-IDF vectorization for text representation</li>
<li>Logistic Regression for initial classification</li>
<li>Quick inference and resource-efficient</li>
</ul>
</li>
<br>
<li><i>Advanced Approach:</i>
<ul>
<li>BERT-based architecture for deep learning</li>
<li>Fine-tuned on construction document dataset</li>
<li>Superior context understanding and accuracy</li>
</ul>
</li>
</ul>
<br>
<b>3. Evaluation Strategy:</b>
<ul>
<li>Comprehensive metric suite (Precision, Recall, F1)</li>
<li>Special consideration for class imbalance</li>
<li>Comparative analysis between baseline and BERT</li>
</ul>
<br>
<b>4. Deployment & Demo:</b>
<ul>
<li>Streamlit-based interactive web interface</li>
<li>Real-time document classification</li>
<li>Comprehensive project documentation</li>
<li>Performance visualization and analytics</li>
</ul>
<br>
<div style='
background-color: #e8f4f8;
padding: 15px;
border-radius: 5px;
border-left: 4px solid #1f77b4;
'>
<b>💡 Key implementation:</b> The parallel processing implementation significantly reduced data preparation time,
allowing for faster iteration and model experimentation. This, combined with the dual-model approach,
provides both efficiency and accuracy in document classification.
</div>
</div>
</div>
""", unsafe_allow_html=True)
# Q3: Model Selection
st.markdown("""
<div class="question-card">
<div class="question">🤖 Q3: Which models did you use and why?</div>
<div class="answer">
Implemented baseline using TF-IDF and Logistic Regression and then used BERT-based model:
<br><br>
<b>Baseline Model:</b>
<ul>
<li>TF-IDF + Logistic Regression</li>
<li>Quick inference time</li>
<li>Resource-efficient</li>
</ul>
<br>
<b>BERT Model:</b>
<ul>
<li>Fine-tuned on 1800 samples text</li>
<li>Better context understanding</li>
<li>Better handling of complex documents</li>
</ul>
</div>
</div>
""", unsafe_allow_html=True)
# Q4: Limitations and Improvements
st.markdown("""
<div class="question-card">
<div class="question">⚠️ Q4: What are the current limitations and potential improvements?</div>
<div class="answer">
<b>Current Implementation & Limitations:</b>
<ul>
<li>~25% of dataset URLs were inaccessible</li>
<li>Used Threadpooling for parallel downloading of train and test documents</li>
</ul>
<br>
<b>Proposed Improvements:</b>
<ul>
<li>Use latest LLMs like GPT-4o, Claude 3.5 Sonnet etc with few shot prompting to speed up the development process</li>
<li>Optimize inference pipeline for faster processing using distilled models like DistilBERT, or the last BERT based model - ModernBERT to compare the performance</li>
<li>Add support for more document formats</li>
</ul>
</div>
</div>
""", unsafe_allow_html=True)
# Q5: Model Performance
st.markdown("""
<div class="question-card">
<div class="question">📊 Q5: What is the model's performance on test data?</div>
<div class="answer">
<b>BERT Model Performance:</b>
<br><br>
<div style="overflow-x: auto;">
<table style="
width: 100%;
border-collapse: collapse;
margin: 20px 0;
font-size: 0.9em;
font-family: sans-serif;
box-shadow: 0 0 20px rgba(0, 0, 0, 0.15);
border-radius: 5px;
">
<thead>
<tr style="
background-color: #1f77b4;
color: white;
text-align: left;
">
<th style="padding: 12px 15px;">Category</th>
<th style="padding: 12px 15px;">Precision</th>
<th style="padding: 12px 15px;">Recall</th>
<th style="padding: 12px 15px;">F1-Score</th>
<th style="padding: 12px 15px;">Support</th>
</tr>
</thead>
<tbody>
<tr style="border-bottom: 1px solid #dddddd;">
<td style="padding: 12px 15px;"><b>Cable</b></td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">92</td>
</tr>
<tr style="border-bottom: 1px solid #dddddd; background-color: #f3f3f3;">
<td style="padding: 12px 15px;"><b>Fuses</b></td>
<td style="padding: 12px 15px;">0.95</td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">42</td>
</tr>
<tr style="border-bottom: 1px solid #dddddd;">
<td style="padding: 12px 15px;"><b>Lighting</b></td>
<td style="padding: 12px 15px;">0.94</td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">0.97</td>
<td style="padding: 12px 15px;">74</td>
</tr>
<tr style="border-bottom: 1px solid #dddddd; background-color: #f3f3f3;">
<td style="padding: 12px 15px;"><b>Others</b></td>
<td style="padding: 12px 15px;">1.00</td>
<td style="padding: 12px 15px;">0.92</td>
<td style="padding: 12px 15px;">0.96</td>
<td style="padding: 12px 15px;">83</td>
</tr>
</tbody>
<tfoot>
<tr style="background-color: #f8f9fa; font-weight: bold; border-top: 2px solid #dddddd;">
<td style="padding: 12px 15px;">Accuracy</td>
<td style="padding: 12px 15px;" colspan="3">0.98</td>
<td style="padding: 12px 15px;">291</td>
</tr>
<tr style="background-color: #f8f9fa; color: #666;">
<td style="padding: 12px 15px;">Macro Avg</td>
<td style="padding: 12px 15px;">0.97</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">291</td>
</tr>
<tr style="background-color: #f8f9fa; color: #666;">
<td style="padding: 12px 15px;">Weighted Avg</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">0.98</td>
<td style="padding: 12px 15px;">291</td>
</tr>
</tfoot>
</table>
</div>
</div>
</div>
""", unsafe_allow_html=True)
st.markdown("""
<div style='
background-color: #f8f9fa;
padding: 20px;
border-radius: 10px;
border-left: 5px solid #1f77b4;
margin: 20px 0;
'>
✨ Perfect performance (1.00) for Cable category<br>
📈 High recall (1.00) across most categories<br>
🎯 Overall accuracy of 98%<br>
⚖️ Balanced performance across all metrics
</div>
""", unsafe_allow_html=True)
# Q6: Metric Selection
st.markdown("""
<div class="question-card">
<div class="question">📈 Q6: Why did you choose these particular metrics?</div>
<div class="answer">
Our metric selection was driven by the dataset characteristics:
<br><br>
<b>Key Considerations:</b>
<ul>
<li>Dataset has mild class imbalance (Imbalance Ratio: 2.36)</li>
<li>Need for balanced evaluation across all classes</li>
</ul>
<br>
<b>Selected Metrics:</b>
<ul>
<li><b>Precision:</b> Critical for minimizing false positives</li>
<li><b>Recall:</b> Important for catching all instances of each class</li>
<li><b>F1-Score:</b> Provides balanced evaluation of both metrics</li>
<li><b>Weighted Average:</b> Accounts for class imbalance</li>
</ul>
</div>
</div>
""", unsafe_allow_html=True)
# Performance Visualization
st.markdown("### 📊 Model Performance Comparison")
metrics = {
'Metric': ['Accuracy', 'Precision', 'Recall', 'F1-Score'],
'Baseline': [0.85, 0.83, 0.84, 0.83],
'BERT': [0.98, 0.97, 0.98, 0.98]
}
df = pd.DataFrame(metrics)
fig = px.bar(
df,
x='Metric',
y=['Baseline', 'BERT'],
barmode='group',
title='Model Performance Comparison',
color_discrete_sequence=['#2ecc71', '#3498db'],
template='plotly_white'
)
fig.update_layout(
title_x=0.5,
title_font_size=20,
legend_title_text='Model Type',
xaxis_title="Evaluation Metric",
yaxis_title="Score",
bargap=0.2,
height=500
)
st.plotly_chart(fig, use_container_width=True)
main() |