{"id":49186887,"url":"https://github.com/zjunlp/scinet","last_synced_at":"2026-04-27T08:01:01.439Z","repository":{"id":353058811,"uuid":"1215961599","full_name":"zjunlp/SciNet","owner":"zjunlp","description":"A Large-Scale Knowledge Graph for Automated Scientific Research","archived":false,"fork":false,"pushed_at":"2026-04-25T16:31:46.000Z","size":3356,"stargazers_count":4,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-26T07:02:40.563Z","etag":null,"topics":["agent","ai-agent","ai-for-science","ai4research","automated-scientific-research","autonomous-agent","idea-generation","knowledge-graph","large-language-models","llm","research-automation","resource","retrieval-augmented-generation","science-discovery","scientific-knowledge-graph","scinet"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/zjunlp.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","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,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-20T12:31:02.000Z","updated_at":"2026-04-25T16:31:49.000Z","dependencies_parsed_at":"2026-04-27T08:00:42.470Z","dependency_job_id":null,"html_url":"https://github.com/zjunlp/SciNet","commit_stats":null,"previous_names":["zjunlp/scinet"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/zjunlp/SciNet","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zjunlp%2FSciNet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zjunlp%2FSciNet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zjunlp%2FSciNet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zjunlp%2FSciNet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/zjunlp","download_url":"https://codeload.github.com/zjunlp/SciNet/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/zjunlp%2FSciNet/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32327701,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-26T23:26:28.701Z","status":"online","status_checked_at":"2026-04-27T02:00:06.769Z","response_time":128,"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":["agent","ai-agent","ai-for-science","ai4research","automated-scientific-research","autonomous-agent","idea-generation","knowledge-graph","large-language-models","llm","research-automation","resource","retrieval-augmented-generation","science-discovery","scientific-knowledge-graph","scinet"],"created_at":"2026-04-23T04:30:32.414Z","updated_at":"2026-04-27T08:01:01.384Z","avatar_url":"https://github.com/zjunlp.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cdiv align=\"center\"\u003e\n  \u003ch1\u003eSciNet: A Large-Scale Knowledge Graph for Automated Scientific Research\u003c/h1\u003e\n\u003c/div\u003e\n\n\u003cp align=\"center\"\u003e\n  Open-source client for running literature-grounded scientific research tasks on top of SciNet API.\n\u003c/p\u003e\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://arxiv.org/abs/2602.14367\"\u003e📄arXiv\u003c/a\u003e\n\u003c/p\u003e\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://github.com/zjunlp/SciNet\"\u003e\n  \t\u003cimg src=\"https://awesome.re/badge.svg\" alt=\"Awesome\"\u003e\n  \u003c/a\u003e\n  \u003ca href=\"https://github.com/zjunlp/SciNet/blob/main/LICENSE\"\u003e\n    \u003cimg src=\"https://img.shields.io/badge/License-MIT-green.svg\" alt=\"License: MIT\"\u003e\n  \u003c/a\u003e\n  \u003cimg src=\"https://img.shields.io/github/last-commit/zjunlp/SciNet?color=blue\" alt=\"Last Commit\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/PRs-Welcome-red\" alt=\"PRs Welcome\"\u003e\n\u003c/p\u003e\n\n------\n\n## 📑 Table of Contents\n\n- [✨ Overview](#-overview)\n- [🧩 Supported Tasks](#-supported-tasks)\n- [🛠️ GROBID](#-grobid)\n- [📂 Layout](#-repository-layout)\n- [🚀 Quick Start](#-quick-start)\n- [🧪 Run Tasks](#-run-tasks)\n  - [`Idea Grounding and Evaluation`](#idea-grounding-and-evaluation)\n  - [`Idea Generation`](#idea-generation)\n  - [`Research Trend Predicting`](#research-trend-predicting)\n  - [`Related Author Retrieval`](#related-author-retrieval)\n  - [`Researcher Background Review`](#researcher-background-review)\n- [📝 TODO](#-todo)\n- [✍️ Citation](#-citation)\n\n## ✨ Overview\n\nSciNet is a large-scale, multi-disciplinary, heterogeneous academic resource knowledge graph designed as a panoramic scientific evolution network. By integrating over 43M papers from 26 disciplines, and a total of 157M entites and 3B triplets, SciNet provides a structured topological cognitive substrate that dismantles disciplinary barriers and furnishes AI agents with a global perspective.\n\n\u003cimg src=\"imgs/field_distribution_pie.png\" alt=\"field_distribution_pie\" style=\"zoom:7%;\" /\u003e\n\n\u003cdiv align=\"center\"\u003e\n  Discipline Distribution in SciNet\n\u003c/div\u003e\n\n\u003cimg src=\"imgs/schema.png\" alt=\"schema\" style=\"zoom:10%;\" /\u003e\n\n\u003cdiv align=\"center\"\u003e\n  Schema of SciNet\n\u003c/div\u003e\n\nThis repository provides a runnable client for several scientific research workflows, including idea evaluation, topic review, author discovery, author profiling, and idea generation.\n\nEach run is driven by CLI inputs plus optional runtime parameter overrides. The client also writes a `request.json` file into the run directory so every execution remains easy to inspect and reproduce later.\n\nThe local client is responsible for:\n\n- building a structured request\n- calling a hosted `SciNet API`\n- running client-side post-processing such as reranking, PDF parsing, grounding, and Markdown report generation\n\nUsers do **not** need to connect to Neo4j or other database components directly.\n\n## 🧩 Supported Tasks\n\n| Task Type | Required Input | Main Output |\n| --- | --- | --- |\n| `Idea Grounding and Evaluation` | `--idea-text` or `--pdf-path` | grounded evidence, paragraph matches, and idea-level analysis |\n| `Idea Generation` | `--topic-text` | generated ideas grounded in retrieved literature |\n| `Research Trend Predicting` | `--topic-text` | topic evolution summary and representative papers |\n| `Related Author Retrieval` | `--idea-text` or `--pdf-path` | related authors and supporting papers |\n| `Researcher Background Review` | `--author-name` | research trajectory and representative works |\n\n## 🛠️ GROBID\n\nGROBID extracts structured metadata from scientific PDFs, including titles, authors, abstracts, and references.\n\nGROBID is needed for:\n\n- `Idea Grounding and Evaluation`\n- `Related Author Retrieval` when using `--pdf-path`\n\nExample startup with Docker:\n\n```bash\ndocker pull lfoppiano/grobid:latest\ndocker run -d --rm --name grobid -p 8070:8070 lfoppiano/grobid:latest\ncurl http://127.0.0.1:8070/api/isalive\n```\n\n## 📂 Layout\n\nThis repository is a lightweight client for SciNet workflows.\n\n- `run_scinet.py`: main entrypoint for runs\n- `scinet/`: main runtime package, including CLI handling, task dispatch, retrieval, grounding, and Markdown rendering\n- `references/search/`: reference and demonstration code for the search logic; it is kept for inspection and illustration, and is not part of the default runtime path\n\n## 🚀 Quick Start\n\nUse the following steps to get a working run from a clean checkout.\n\n### 1. Create an environment and install dependencies\n\n```bash\npython3 -m venv .venv\nsource .venv/bin/activate\npip install -U pip\npip install -r requirements.txt\n```\n\n### 2. Create the environment file\n\n```bash\ncp .env.example .env\n```\n\nWe currently provide a public `SciNet API` endpoint for community use. Fill in the required variables:\n\n```env\nSCINET_API_BASE_URL=http://scinet.openkg.cn\nSCINET_API_KEY=scinet-public-key\n\nLLM_PROVIDER=openai_compatible\nLLM_API_KEY=replace-me\nLLM_BASE_URL=https://your-openai-compatible-endpoint/v1\nLLM_MODEL=your-model-name\n\n# Legacy compatibility keys. New setups should prefer LLM_*.\nOPENAI_API_KEY=replace-me\nOPENAI_BASE_URL=https://your-openai-compatible-endpoint/v1\nOPENAI_MODEL=your-model-name\n\nGROBID_BASE_URL=http://127.0.0.1:8070\nOA_API_KEY=\nOPENALEX_MAILTO=\n```\n\nYou can get an OpenAlex API key from [here](https://openalex.org/settings/api-key).\n\nRequired variables:\n\n| Variable | Required For | Notes |\n| --- | --- | --- |\n| `SCINET_API_BASE_URL` | all tasks | hosted `SciNet API` base URL |\n| `SCINET_API_KEY` | all tasks | sent as `X-API-Key` |\n| `LLM_PROVIDER` | all tasks | provider selector, currently `openai_compatible` |\n| `LLM_API_KEY` | all tasks | used for planning, reranking, and summarization |\n| `LLM_BASE_URL` | all tasks | OpenAI-compatible base URL |\n| `LLM_MODEL` | all tasks | chat model name |\n| `GROBID_BASE_URL` | PDF tasks | needed for `--pdf-path` flows |\n| `OA_API_KEY` | optional | OpenAlex fallback support |\n| `OPENALEX_MAILTO` | optional | OpenAlex contact email |\n\n### 3. Make sure the required services are ready\n\n- a hosted `SciNet API`\n- an LLM endpoint exposed in OpenAI-compatible format\n- GROBID if you want to use `--pdf-path`\n\n### 4. Run a task\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Research Trend Predicting\" \\\n  --topic-text \"research idea evaluation with large language models\" \\\n  --pretty\n```\n\n### 5. Check the output\n\nEach run creates a directory under `runs/` containing:\n\n- `request.json`\n- `result.json`\n- `result.md`\n\n## 🧪 Run Tasks\n\n### `Idea Grounding and Evaluation`\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Idea Grounding and Evaluation\" \\\n  --idea-text \"Use literature-grounded evidence to evaluate research ideas.\" \\\n  --pretty\n```\n\nWith PDF input:\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Idea Grounding and Evaluation\" \\\n  --pdf-path /absolute/path/to/paper.pdf \\\n  --params-json '{\"search_final_top_k\": 15, \"manifest_top_k\": 10}' \\\n  --pretty\n```\n\n### `Idea Generation`\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Idea Generation\" \\\n  --topic-text \"scientific idea generation with retrieval-augmented large language models\" \\\n  --pretty\n```\n\n### `Research Trend Predicting`\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Research Trend Predicting\" \\\n  --topic-text \"research idea evaluation with large language models\" \\\n  --pretty\n```\n\n### `Related Author Retrieval`\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Related Author Retrieval\" \\\n  --idea-text \"knowledge-grounded evaluation of scientific research ideas\" \\\n  --pretty\n```\n\n### `Researcher Background Review`\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Researcher Background Review\" \\\n  --author-name \"Geoffrey Hinton\" \\\n  --pretty\n```\n\nYou can also override task parameters from your own JSON file:\n\n```bash\npython3 run_scinet.py \\\n  --task-type \"Idea Grounding and Evaluation\" \\\n  --idea-text \"Use literature-grounded evidence to evaluate research ideas.\" \\\n  --params-file /absolute/path/to/params.json \\\n  --pretty\n```\n\nFor `Idea Grounding and Evaluation`, model-related overrides can be supplied through `--params-file` or `--params-json`.\nYou can also set local model paths once in `.env`:\n\n```bash\nSCINET_EMBEDDING_MODEL_PATH=/absolute/path/to/BAAI--bge-large-en-v1.5\nSCINET_RERANKER_MODEL_PATH=/absolute/path/to/BAAI--bge-reranker-large\n```\n\n`grounded_review` also accepts `query_provider`, `query_model`, and `query_api_url` overrides in `params`.\nIf omitted, it resolves them from `LLM_PROVIDER`, `LLM_MODEL`, and `LLM_BASE_URL`.\n\nBy default, `Idea Grounding and Evaluation` uses:\n\n- embedding model: `BAAI/bge-large-en-v1.5` [huggingface_url](https://huggingface.co/BAAI/bge-large-en-v1.5)\n- reranker model: `BAAI/bge-reranker-large` [huggingface_url](https://huggingface.co/BAAI/bge-reranker-large)\n\nThe first run may download these models into the local Hugging Face cache.\n\n## 📝 TODO\n\n- [ ] **CLI Tools.** Add more user-facing CLI capabilities so downstream users and AI agents can invoke retrieval workflows without touching database internals.\n- [ ] **Skills.** Package reusable agent skills for common scientific discovery workflows and expose best practices as easier-to-load components.\n- [ ] **More Knowledge.** Integrate more knowledge forms beyond paper-centric entities, such as datasets, code, standards, theorems, and experimental experience.\n- [ ] **Benchmark and Evaluation.** Build dedicated benchmarks and evaluation protocols for downstream scientific research tasks supported by SciNet.\n- [ ] **Dynamic Update**Improve dynamic knowledge updates toward a more systematic and frequent refresh mechanism.\n\n## ✍️ Citation\n\nIf you find our work helpful, please use the following citations.\n\n```\n\n```\n\n### License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzjunlp%2Fscinet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fzjunlp%2Fscinet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fzjunlp%2Fscinet/lists"}