Update Readme.md

Readme.md

@@ -1,109 +1,11 @@
-title: Explainable Recommendation System
-emoji: 🤖
+---
+title: Explainable LLaMA Recommendation System
+emoji: 📚
 colorFrom: blue
-colorTo: green
+colorTo: purple
 sdk: gradio
-sdk_version: "4.36.0"
+sdk_version: "4.44.0"
 python_version: "3.10"
 app_file: app.py
 pinned: false
-Explainable Recommendation System Using Large Language Models
-Group 15 — CSAI 801
-MemberRoleBassant YehiaDeveloperMennaalla AhmedDeveloperMariem BastarousDeveloper
-Supervisor: Dr. Rahatara Ferdousi  |  TA: Eng. Salman Rakin
-
-Project Overview
-A content-based recommendation system that uses an LLM (DeepSeek-R1) to generate
-personalised recommendations with natural-language explanations, addressing:
-
-Cold-start problem — no user history needed; relies purely on semantic content
-Lack of transparency — every recommendation includes a structured WHY
-
-
-Architecture
-PDF Knowledge Base
-       │
-       ▼
-PyMuPDFLoader  →  RecursiveCharacterTextSplitter
-                          │
-                          ▼ (1000-char chunks, 200 overlap)
-              OllamaEmbeddings (nomic-embed-text)
-                          │  [parallel via ThreadPoolExecutor]
-                          ▼
-                    ChromaDB Vector Store
-                          │
-              ┌───────────┘
-              │  Semantic retrieval (Top-K chunks)
-              ▼
-   ChatPromptTemplate (System + Human)
-              │
-              ▼
-   ChatOllama (DeepSeek-R1, temp=0.5)
-              │
-              ▼
-   StrOutputParser → Explainable Recommendation
-              │
-       ┌──────┴──────┐
-       ▼             ▼
-  Gradio UI     FastAPI Endpoint
-
-Quick Start
-1. Prerequisites
-bash# Install Ollama  (https://ollama.ai)
-ollama pull nomic-embed-text
-ollama pull deepseek-r1
-
-# Install Python dependencies
-pip install -r requirements.txt
-2. Place your PDF
-Put your Foundational_LLM.pdf in the project root directory.
-3a. Run the Gradio UI
-bashpython app.py
-Opens at http://localhost:7860 (or the public share URL printed in the terminal).
-3b. Run the FastAPI server
-bashuvicorn api:app --host 0.0.0.0 --port 8000 --reload
-Interactive API docs at http://localhost:8000/docs
-
-API Endpoints
-MethodPathDescriptionGET/healthHealth checkGET/infoProject & system metadataPOST/recommendSingle recommendation queryPOST/recommend/batchBatch queries (up to 10)
-Example — Single Query
-bashcurl -X POST http://localhost:8000/recommend \
-  -H "Content-Type: application/json" \
-  -d '{"question": "What techniques improve LLM reasoning?", "top_k": 5}'
-Example — Batch Query
-bashcurl -X POST http://localhost:8000/recommend/batch \
-  -H "Content-Type: application/json" \
-  -d '{
-    "questions": [
-      "What is prompt engineering?",
-      "How does RAG work?"
-    ],
-    "top_k": 5
-  }'
-
-Features vs. Traditional Systems
-FeatureTraditional RSThis System (FM-based)Cold startWeakStrong (content-based)ExplainabilityLowHigh (structured WHY)ConversationalNoYesSetup speedSlowFastReasoningStatisticalSemantic
-
-Project Scope (30 Hours)
-Included ✅
-
-Content-based recommendations
-Embeddings + LLM explanations
-API-only implementation (FastAPI)
-Gradio prototype UI
-
-Excluded ❌
-
-Large-scale data
-Collaborative filtering
-Production frontend UI
-Production deployment
-
-
-File Structure
-recommendation_system/
-├── app.py            ← Gradio UI + full RAG pipeline
-├── api.py            ← FastAPI REST endpoint
-├── requirements.txt  ← Python dependencies
-├── README.md         ← This file
-└── Foundational_LLM.pdf   ← Knowledge base (add manually)
+---