Spaces:
Sleeping
Sleeping
File size: 18,954 Bytes
3a93fb9 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 | <!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>The Screening Room - Inside the Screening Room</title>
<link rel="stylesheet" href="style.css">
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700&family=Playfair+Display:ital,wght@0,600;0,700;1,600&family=Space+Mono&display=swap" rel="stylesheet">
</head>
<body>
<div class="marquee-border top-marquee">
<div class="lights"></div>
</div>
<nav class="nav-bar">
<div class="nav-container">
<div class="nav-left">
<a href="index.html" class="nav-logo" aria-label="Home">
<img src="film_reel.png" class="film-reel-icon" alt="The Screening Room Logo">
</a>
</div>
<div class="nav-links">
<a href="index.html" class="nav-link">Analyze</a>
<span class="nav-divider">|</span>
<a href="inside.html" class="nav-link active">Inside the Screening Room</a>
<span class="nav-divider">|</span>
<a href="https://github.com" class="nav-link" target="_blank">GitHub</a>
</div>
</div>
</nav>
<!-- SECTION 1: Hero -->
<section class="inside-hero">
<div class="hero-projector"></div>
<div class="hero-content">
<h1>Inside the Screening Room</h1>
<p>Discover how three independent AI models analyze the same movie review, explore the technology behind the platform, and learn how The Screening Room brings machine learning and modern software engineering together.</p>
</div>
</section>
<!-- SECTION 2: Meet the Critics -->
<section class="section-container">
<h2 class="section-title">Meet the Critics <span class="star-icon">⭐</span></h2>
<div class="critics-grid inside-critics">
<!-- The Statistician -->
<div class="critic-card" id="criticLr">
<div class="critic-header">
<h4>The Statistician</h4>
<span class="model-type">Logistic Regression + TF-IDF</span>
</div>
<div class="critic-body">
<p class="critic-desc">The Statistician evaluates a review by measuring how strongly individual words and phrases indicate positive or negative sentiment. It provides a fast, reliable baseline and is highly interpretable.</p>
<div class="critic-strengths">
<h5>Strengths</h5>
<ul>
<li>Extremely fast inference</li>
<li>Easy to interpret</li>
<li>Strong traditional machine learning baseline</li>
</ul>
</div>
</div>
</div>
<!-- The Contextualist -->
<div class="critic-card" id="criticBert">
<div class="critic-header">
<h4>The Contextualist</h4>
<span class="model-type">BERT</span>
</div>
<div class="critic-body">
<p class="critic-desc">The Contextualist reads every word in relation to every other word, allowing it to understand context, nuanced language, and complex sentence structures.</p>
<div class="critic-strengths">
<h5>Strengths</h5>
<ul>
<li>Context-aware</li>
<li>Handles nuanced language</li>
<li>Strong performance on complex reviews</li>
</ul>
</div>
</div>
</div>
<!-- The Sequentialist -->
<div class="critic-card" id="criticLstm">
<div class="critic-header">
<h4>The Sequentialist</h4>
<span class="model-type">Bi-LSTM</span>
</div>
<div class="critic-body">
<p class="critic-desc">The Sequentialist processes each review in order, learning how meaning develops across a sentence. It captures long-term dependencies and contextual flow within the text.</p>
<div class="critic-strengths">
<h5>Strengths</h5>
<ul>
<li>Understands sentence order</li>
<li>Learns contextual sequences</li>
<li>Captures long-range dependencies</li>
</ul>
</div>
</div>
</div>
</div>
</section>
<!-- SECTION 3: How It Works -->
<section class="section-container">
<h2 class="section-title">How It Works</h2>
<div class="pipeline-container">
<div class="pipeline-step">
<div class="step-title">User Review</div>
<div class="step-desc">User submits a movie review.</div>
</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">
<div class="step-title">Model-specific Preprocessing</div>
<div class="step-desc">Each model receives text prepared according to its own preprocessing pipeline.</div>
</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">
<div class="step-title">Parallel Inference</div>
<div class="step-desc">Logistic Regression, Bi-LSTM, and BERT run independently.</div>
</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">
<div class="step-title">Prediction Standardization</div>
<div class="step-desc">Outputs from all models are converted into one consistent response format.</div>
</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">
<div class="step-title">Majority Decision</div>
<div class="step-desc">The application compares predictions and determines the overall verdict.</div>
</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">
<div class="step-title">Frontend Rendering</div>
<div class="step-desc">Predictions, confidence scores, and model insights are displayed simultaneously.</div>
</div>
</div>
</section>
<!-- SECTION 4: Project Details -->
<section class="section-container">
<h2 class="section-title">Project Details</h2>
<div class="accordion">
<!-- Performance -->
<div class="accordion-item active">
<button class="accordion-header" aria-expanded="true">
<span>Performance</span>
<span class="accordion-icon">▼</span>
</button>
<div class="accordion-content" style="max-height: 1000px;">
<div class="table-container">
<table class="tech-table">
<thead>
<tr>
<th>Model</th>
<th>Accuracy</th>
<th>Precision</th>
<th>Recall</th>
<th>F1</th>
<th>Inference</th>
</tr>
</thead>
<tbody id="metricsTableBody">
<!-- Populated by JS -->
</tbody>
</table>
<p class="table-caption">Metrics are reported using the final evaluation performed on the held-out IMDb test dataset.</p>
</div>
</div>
</div>
<!-- Dataset -->
<div class="accordion-item">
<button class="accordion-header" aria-expanded="false">
<span>Dataset</span>
<span class="accordion-icon">▶</span>
</button>
<div class="accordion-content">
<div class="dataset-grid">
<div class="dataset-stats">
<h3>IMDb Large Movie Review Dataset</h3>
<div class="code-block">
50,000 Reviews<br>
25,000 Positive<br>
25,000 Negative<br>
Balanced Dataset
</div>
<p>The models were trained and evaluated using the IMDb Large Movie Review Dataset, a widely used benchmark for binary sentiment classification introduced by Andrew L. Maas and colleagues.</p>
</div>
<div class="dataset-attribution">
<h3>Dataset Attribution</h3>
<p><strong>Dataset:</strong> IMDb Large Movie Review Dataset</p>
<p><strong>Authors:</strong> Andrew L. Maas et al.</p>
<p><strong>Institution:</strong> Stanford AI Lab</p>
<p><strong>Official Source:</strong> <a href="https://ai.stanford.edu/~amaas/data/sentiment/" target="_blank">https://ai.stanford.edu/~amaas/data/sentiment/</a></p>
</div>
</div>
</div>
</div>
<!-- Technology Stack -->
<div class="accordion-item">
<button class="accordion-header" aria-expanded="false">
<span>Technology Stack</span>
<span class="accordion-icon">▶</span>
</button>
<div class="accordion-content">
<div class="tech-grid">
<div class="tech-card">
<h4>Frontend</h4>
<ul>
<li>HTML</li>
<li>CSS</li>
<li>JavaScript</li>
</ul>
</div>
<div class="tech-card">
<h4>Backend</h4>
<ul>
<li>Express.js</li>
<li>FastAPI</li>
</ul>
</div>
<div class="tech-card">
<h4>Machine Learning</h4>
<ul>
<li>Scikit-learn</li>
<li>TensorFlow / Keras</li>
<li>Hugging Face Transformers</li>
</ul>
</div>
<div class="tech-card">
<h4>Deployment</h4>
<ul>
<li>Docker</li>
<li>Render</li>
<li>Neon PostgreSQL</li>
</ul>
</div>
</div>
</div>
</div>
<!-- Architecture -->
<div class="accordion-item">
<button class="accordion-header" aria-expanded="false">
<span>Architecture</span>
<span class="accordion-icon">▶</span>
</button>
<div class="accordion-content">
<div class="pipeline-container architecture-flow">
<div class="pipeline-step">Browser</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">Express Backend</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">FastAPI ML Service</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">Prediction Engine</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">Three Models</div>
<div class="pipeline-arrow">↓</div>
<div class="pipeline-step">Standardized Response</div>
</div>
</div>
</div>
</div>
</section>
<!-- SECTION 5: Behind the Build -->
<section class="section-container">
<h2 class="section-title">Behind the Build ⭐</h2>
<div class="accordion">
<div class="accordion-item">
<button class="accordion-header" aria-expanded="false">
<span>Engineering Decisions</span>
<span class="accordion-icon">▶</span>
</button>
<div class="accordion-content">
<div class="engineering-decision">
<h4>Why FastAPI?</h4>
<p>FastAPI powers the machine learning inference service because it provides high-performance APIs while integrating naturally with Python-based ML libraries.</p>
</div>
<div class="engineering-decision">
<h4>Why Express?</h4>
<p>Express separates application logic from machine learning inference by handling validation, routing, persistence, and API orchestration.</p>
</div>
<div class="engineering-decision">
<h4>Why PostgreSQL?</h4>
<p>PostgreSQL stores prediction history and model metadata using a relational schema to demonstrate production-style persistence.</p>
</div>
<div class="engineering-decision">
<h4>Why Three Models?</h4>
<p>The goal is not only to predict sentiment but also to compare how fundamentally different NLP approaches interpret the same review.</p>
</div>
</div>
</div>
<div class="accordion-item">
<button class="accordion-header" aria-expanded="false">
<span>Challenges</span>
<span class="accordion-icon">▶</span>
</button>
<div class="accordion-content">
<div class="table-container">
<table class="tech-table challenges-table">
<thead>
<tr>
<th>Challenge</th>
<th>Solution</th>
<th>Result</th>
</tr>
</thead>
<tbody>
<tr>
<td>Different preprocessing requirements</td>
<td>Model-specific preprocessing pipeline</td>
<td>Consistent and reliable predictions</td>
</tr>
<tr>
<td>Three incompatible prediction outputs</td>
<td>Unified response schema</td>
<td>Common API contract across all models</td>
</tr>
<tr>
<td>Large BERT model</td>
<td>Singleton loading during application startup</td>
<td>Reduced inference overhead after initialization</td>
</tr>
<tr>
<td>Different confidence formats</td>
<td>Standardized prediction response</td>
<td>Consistent frontend rendering</td>
</tr>
</tbody>
</table>
</div>
</div>
</div>
</div>
</section>
<!-- SECTION 6: Future Improvements -->
<section class="section-container">
<h2 class="section-title">Future Improvements</h2>
<div class="future-grid">
<div class="future-card">
<div class="future-header">
<h4>Multilingual Reviews</h4>
<span class="badge planned">Planned</span>
</div>
<p>Support additional languages beyond English.</p>
</div>
<div class="future-card">
<div class="future-header">
<h4>Batch Analysis</h4>
<span class="badge planned">Planned</span>
</div>
<p>Analyze multiple reviews simultaneously.</p>
</div>
<div class="future-card">
<div class="future-header">
<h4>Additional Transformer Models</h4>
<span class="badge future">Future</span>
</div>
<p>Compare more state-of-the-art NLP models.</p>
</div>
<div class="future-card">
<div class="future-header">
<h4>Mobile Experience</h4>
<span class="badge future">Future</span>
</div>
<p>Responsive mobile-first interface.</p>
</div>
<div class="future-card">
<div class="future-header">
<h4>Advanced Model Interpretation</h4>
<span class="badge research">Research</span>
</div>
<p>Expand interpretation techniques for deeper insight into model behavior.</p>
</div>
</div>
</section>
<footer class="site-footer">
<div class="footer-content">
<h4>The Screening Room</h4>
<p>Three Critics.<br>One Review.<br>Three Perspectives.</p>
</div>
</footer>
<script src="inside.js"></script>
</body>
</html>
|