{"id":29913601,"url":"https://github.com/quantumcoderrr/adobe-india-hackathon25","last_synced_at":"2025-08-02T02:13:01.855Z","repository":{"id":306416315,"uuid":"1025681946","full_name":"QuantumCoderrr/Adobe-India-Hackathon25","owner":"QuantumCoderrr","description":null,"archived":false,"fork":false,"pushed_at":"2025-07-25T14:43:42.000Z","size":25134,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-07-25T16:40:34.524Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/QuantumCoderrr.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-07-24T16:20:24.000Z","updated_at":"2025-07-25T14:43:46.000Z","dependencies_parsed_at":"2025-07-25T16:50:40.782Z","dependency_job_id":null,"html_url":"https://github.com/QuantumCoderrr/Adobe-India-Hackathon25","commit_stats":null,"previous_names":["quantumcoderrr/adobe-india-hackathon25"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/QuantumCoderrr/Adobe-India-Hackathon25","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QuantumCoderrr%2FAdobe-India-Hackathon25","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QuantumCoderrr%2FAdobe-India-Hackathon25/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QuantumCoderrr%2FAdobe-India-Hackathon25/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QuantumCoderrr%2FAdobe-India-Hackathon25/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/QuantumCoderrr","download_url":"https://codeload.github.com/QuantumCoderrr/Adobe-India-Hackathon25/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QuantumCoderrr%2FAdobe-India-Hackathon25/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":268326739,"owners_count":24232496,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-08-02T02:00:12.353Z","response_time":74,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2025-08-02T02:12:58.174Z","updated_at":"2025-08-02T02:13:01.847Z","avatar_url":"https://github.com/QuantumCoderrr.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🚀 Adobe India Hackathon 2025 – \"Connecting the Dots\"\n\n## 🔍 Rethink Reading. Rediscover Knowledge.\n\nImagine a world where PDFs aren’t just passive documents — but intelligent, interactive companions that **understand structure**, **connect ideas**, and **respond meaningfully**. That’s the mission of Adobe’s *Connecting the Dots* challenge — and this repository is our response to it.\n\n---\n\n## 📌 Problem Statement\n\nIn an era where we’re flooded with digital documents, the real power lies not in reading more — but in reading smarter. Adobe’s challenge asked us to:\n- ✅ Extract intelligent outlines from PDFs (**Challenge 1A**)\n- ✅ Identify section-specific content based on user personas (**Challenge 1B**)\n- 🧠 Do it all with lightweight models, on-device, and with high accuracy\n- 📦 Wrap everything in reproducible, portable Docker containers\n\n---\n\n## 🧠 Solutions Overview\n\n### 🔹 Challenge 1A – Structured PDF Outline Extraction\n\n**Objective**: Build a Python script that processes a directory of PDFs and returns JSON-formatted outlines — capturing headings, structure, and page numbers.\n\n- 🛠 Built with `PyMuPDF` (fitz) for PDF parsing\n- 📂 Input/output via CLI arguments\n- 🐳 Packaged in a Docker container for seamless execution\n- 📄 Outputs: Individual `.json` files per PDF with structural metadata\n\n📁 Folder: [`Challenge_1a`](./Challenge_1a)\n\n➡️ Includes:\n- `process_pdfs.py`\n- `Dockerfile`\n- `sample_dataset/` (PDFs)\n- `output/` (Generated JSON files)\n\n---\n\n### 🔹 Challenge 1B – Persona-Driven Section Extraction\n\n**Objective**: For a given set of PDFs and a `challenge1b_input.json`, extract and rank the top 5 most relevant sections based on a specified user persona.\n\n- 🤖 Used `sentence-transformers (MiniLM)` for semantic embeddings\n- 📊 Applied `cosine similarity` (via scikit-learn) for ranking sections\n- 🧾 Output format aligned with provided sample files\n- 🐳 Docker-ready, CPU-efficient, \u003c1GB\n\n📁 Folder: [`Challenge_1b`](./Challenge_1b)\n\n➡️ Includes:\n- `process_documents.py`\n- `Dockerfile`\n- Collections 1–3 with:\n  - PDFs\n  - Input prompts\n  - Output JSONs (predicted sections)\n\n---\n\n## 🐳 Docker Instructions (For Judges)\n\nEach challenge can be run independently in Docker.\n\n### 🏗 Build Image\n\n```bash\ndocker build --platform linux/amd64 -t adobe_round1a ./Challenge_1a\ndocker build --platform linux/amd64 -t adobe_round1b ./Challenge_1b\n```\n\n### ▶️ Run Container\n```bash\ndocker run --rm \\\n  -v $(pwd)/Challenge_1a/input:/app/input \\\n  -v $(pwd)/Challenge_1a/output:/app/output \\\n  --network none \\\n  adobe_round1a\n```\n```bash\ndocker run --rm \\\n  -v $(pwd)/Challenge_1b/input:/app/input \\\n  -v $(pwd)/Challenge_1b/output:/app/output \\\n  --network none \\\n  adobe_round1b\n```\n\n## 🧑‍💻 Team\n\n- **Sandip Ghosh** — [GitHub: @QuantumCoderrr](https://github.com/QuantumCoderrr)  \n- **Sandhita Poddar** — [GitHub: @CelestialCoderrr](https://github.com/CelestialCoderrr)  \n\nTogether, we built something that doesn't just read PDFs — it *understands* them.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fquantumcoderrr%2Fadobe-india-hackathon25","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fquantumcoderrr%2Fadobe-india-hackathon25","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fquantumcoderrr%2Fadobe-india-hackathon25/lists"}