[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-2739":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":41,"readmeContent":42,"aiSummary":43,"trendingCount":16,"starSnapshotCount":16,"syncStatus":17,"lastSyncTime":44,"discoverSource":45},2739,"gensim","piskvorky\u002Fgensim","piskvorky","Topic Modelling for Humans","https:\u002F\u002Fradimrehurek.com\u002Fgensim",null,"Python",16438,4408,408,392,0,2,12,34,10,45,"GNU Lesser General Public License v2.1",false,"develop",true,[27,28,29,30,5,31,32,33,34,35,36,37,38,39,40],"data-mining","data-science","document-similarity","fasttext","information-retrieval","machine-learning","natural-language-processing","neural-network","nlp","python","topic-modeling","word-embeddings","word-similarity","word2vec","2026-06-12 02:00:43","gensim – Topic Modelling in Python\n==================================\n\n\u003C!--\nThe following image URLs are obfuscated = proxied and cached through\nGoogle because of Github's proxying issues. See:\nhttps:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Fissues\u002F2805\n-->\n\n[![Build Status](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Factions\u002Fworkflows\u002Ftests.yml\u002Fbadge.svg?branch=develop)](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Factions)\n[![GitHub release](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Frelease\u002Frare-technologies\u002Fgensim.svg?maxAge=3600)](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Freleases)\n[![Downloads](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fdm\u002Fgensim?color=blue)](https:\u002F\u002Fpepy.tech\u002Fproject\u002Fgensim\u002F)\n[![DOI](https:\u002F\u002Fzenodo.org\u002Fbadge\u002FDOI\u002F10.13140\u002F2.1.2393.1847.svg)](https:\u002F\u002Fdoi.org\u002F10.13140\u002F2.1.2393.1847)\n[![Mailing List](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F-Mailing%20List-blue.svg)](https:\u002F\u002Fgroups.google.com\u002Fg\u002Fgensim)\n[![Follow](https:\u002F\u002Fimg.shields.io\u002Ftwitter\u002Ffollow\u002Fgensim_py.svg?style=social&style=flat&logo=twitter&label=Follow&color=blue)](https:\u002F\u002Ftwitter.com\u002Fgensim_py)\n\nGensim is a Python library for *topic modelling*, *document indexing*\nand *similarity retrieval* with large corpora. Target audience is the\n*natural language processing* (NLP) and *information retrieval* (IR)\ncommunity.\n\n## ⚠️ Want to help out? [Sponsor Gensim](https:\u002F\u002Fgithub.com\u002Fsponsors\u002Fpiskvorky) ❤️\n\n## ⚠️ Gensim is in stable maintenance mode: we are not accepting new features, but bug and documentation fixes are still welcome! ⚠️\n\nFeatures\n--------\n\n-   All algorithms are **memory-independent** w.r.t. the corpus size\n    (can process input larger than RAM, streamed, out-of-core),\n-   **Intuitive interfaces**\n    -   easy to plug in your own input corpus\u002Fdatastream (trivial\n        streaming API)\n    -   easy to extend with other Vector Space algorithms (trivial\n        transformation API)\n-   Efficient multicore implementations of popular algorithms, such as\n    online **Latent Semantic Analysis (LSA\u002FLSI\u002FSVD)**, **Latent\n    Dirichlet Allocation (LDA)**, **Random Projections (RP)**,\n    **Hierarchical Dirichlet Process (HDP)** or **word2vec deep\n    learning**.\n-   **Distributed computing**: can run *Latent Semantic Analysis* and\n    *Latent Dirichlet Allocation* on a cluster of computers.\n-   Extensive [documentation and Jupyter Notebook tutorials].\n\nIf this feature list left you scratching your head, you can first read\nmore about the [Vector Space Model] and [unsupervised document analysis]\non Wikipedia.\n\nInstallation\n------------\n\nThis software depends on [NumPy], a Python package for\nscientific computing. Please bear in mind that building NumPy from source\n(e.g. by installing gensim on a platform which lacks NumPy .whl distribution)\nis a non-trivial task involving [linking NumPy to a BLAS library].  \nIt is recommended to provide a fast one (such as MKL, [ATLAS] or\n[OpenBLAS]) which can improve performance by as much as an order of\nmagnitude. On OSX, NumPy picks up its vecLib BLAS automatically,\nso you don’t need to do anything special.\n\nInstall the latest version of gensim:\n\n```bash\n    pip install --upgrade gensim\n```\n\nOr, if you have instead downloaded and unzipped the [source tar.gz]\npackage:\n\n```bash\n    tar -xvzf gensim-X.X.X.tar.gz\n    cd gensim-X.X.X\u002F\n    pip install .\n```\n\nFor alternative modes of installation, see the [documentation].\n\nGensim is being [continuously tested](https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002F#testing) under all\n[supported Python versions](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Fwiki\u002FGensim-And-Compatibility).\nSupport for Python 2.7 was dropped in gensim 4.0.0 – install gensim 3.8.3 if you must use Python 2.7.\n\nHow come gensim is so fast and memory efficient? Isn’t it pure Python, and isn’t Python slow and greedy?\n--------------------------------------------------------------------------------------------------------\n\nMany scientific algorithms can be expressed in terms of large matrix\noperations (see the BLAS note above). Gensim taps into these low-level\nBLAS libraries, by means of its dependency on NumPy. So while\ngensim-the-top-level-code is pure Python, it actually executes highly\noptimized Fortran\u002FC under the hood, including multithreading (if your\nBLAS is so configured).\n\nMemory-wise, gensim makes heavy use of Python’s built-in generators and\niterators for streamed data processing. Memory efficiency was one of\ngensim’s [design goals], and is a central feature of gensim, rather than\nsomething bolted on as an afterthought.\n\nDocumentation\n-------------\n\n-   [QuickStart]\n-   [Tutorials]\n-   [Official API Documentation]\n\n  [QuickStart]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002Fauto_examples\u002Fcore\u002Frun_core_concepts.html\n  [Tutorials]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002Fauto_examples\u002F\n  [Official Documentation and Walkthrough]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002F\n  [Official API Documentation]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002Fauto_examples\u002Findex.html#documentation\n\nSupport\n-------\n\nFor commercial support, please see [Gensim sponsorship](https:\u002F\u002Fgithub.com\u002Fsponsors\u002Fpiskvorky).\n\nAsk open-ended questions on the public [Gensim Mailing List](https:\u002F\u002Fgroups.google.com\u002Fg\u002Fgensim).\n\nRaise bugs on [Github](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Fblob\u002Fdevelop\u002FCONTRIBUTING.md) but please **make sure you follow the [issue template](https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002Fblob\u002Fdevelop\u002FISSUE_TEMPLATE.md)**. Issues that are not bugs or fail to provide the requested details will be closed without inspection.\n\n\n---------\n\nAdopters\n--------\n\n| Company | Logo | Industry | Use of Gensim |\n|---------|------|----------|---------------|\n| [RARE Technologies](https:\u002F\u002Frare-technologies.com\u002F) | ![rare](docs\u002Fsrc\u002Freadme_images\u002Frare.png) | ML & NLP consulting | Creators of Gensim – this is us! |\n| [Amazon](http:\u002F\u002Fwww.amazon.com\u002F) |  ![amazon](docs\u002Fsrc\u002Freadme_images\u002Famazon.png) | Retail |  Document similarity. |\n| [National Institutes of Health](https:\u002F\u002Fgithub.com\u002FNIHOPA\u002Fpipeline_word2vec) | ![nih](docs\u002Fsrc\u002Freadme_images\u002Fnih.png) | Health | Processing grants and publications with word2vec. |\n| [Cisco Security](http:\u002F\u002Fwww.cisco.com\u002Fc\u002Fen\u002Fus\u002Fproducts\u002Fsecurity\u002Findex.html) | ![cisco](docs\u002Fsrc\u002Freadme_images\u002Fcisco.png) | Security |  Large-scale fraud detection. |\n| [Mindseye](http:\u002F\u002Fwww.mindseyesolutions.com\u002F) | ![mindseye](docs\u002Fsrc\u002Freadme_images\u002Fmindseye.png) | Legal | Similarities in legal documents. |\n| [Channel 4](http:\u002F\u002Fwww.channel4.com\u002F) | ![channel4](docs\u002Fsrc\u002Freadme_images\u002Fchannel4.png) | Media | Recommendation engine. |\n| [Talentpair](http:\u002F\u002Ftalentpair.com) | ![talent-pair](docs\u002Fsrc\u002Freadme_images\u002Ftalent-pair.png) | HR | Candidate matching in high-touch recruiting. |\n| [Juju](http:\u002F\u002Fwww.juju.com\u002F)  | ![juju](docs\u002Fsrc\u002Freadme_images\u002Fjuju.png) | HR | Provide non-obvious related job suggestions. |\n| [Tailwind](https:\u002F\u002Fwww.tailwindapp.com\u002F) | ![tailwind](docs\u002Fsrc\u002Freadme_images\u002Ftailwind.png) | Media | Post interesting and relevant content to Pinterest. |\n| [Issuu](https:\u002F\u002Fissuu.com\u002F) | ![issuu](docs\u002Fsrc\u002Freadme_images\u002Fissuu.png) | Media | Gensim's LDA module lies at the very core of the analysis we perform on each uploaded publication to figure out what it's all about. |\n| [Search Metrics](http:\u002F\u002Fwww.searchmetrics.com\u002F) | ![search-metrics](docs\u002Fsrc\u002Freadme_images\u002Fsearch-metrics.png) | Content Marketing | Gensim word2vec used for entity disambiguation in Search Engine Optimisation. |\n| [12K Research](https:\u002F\u002F12k.com\u002F) | ![12k](docs\u002Fsrc\u002Freadme_images\u002F12k.png)| Media |   Document similarity analysis on media articles. |\n| [Stillwater Supercomputing](http:\u002F\u002Fwww.stillwater-sc.com\u002F) | ![stillwater](docs\u002Fsrc\u002Freadme_images\u002Fstillwater.png) | Hardware | Document comprehension and association with word2vec. |\n| [SiteGround](https:\u002F\u002Fwww.siteground.com\u002F) |  ![siteground](docs\u002Fsrc\u002Freadme_images\u002Fsiteground.png) | Web hosting | An ensemble search engine which uses different embeddings models and similarities, including word2vec, WMD, and LDA. |\n| [Capital One](https:\u002F\u002Fwww.capitalone.com\u002F) | ![capitalone](docs\u002Fsrc\u002Freadme_images\u002Fcapitalone.png) | Finance | Topic modeling for customer complaints exploration. |\n\n-------\n\nCiting gensim\n------------\n\nWhen [citing gensim in academic papers and theses], please use this\nBibTeX entry:\n\n    @inproceedings{rehurek_lrec,\n          title = {{Software Framework for Topic Modelling with Large Corpora}},\n          author = {Radim {\\v R}eh{\\r u}{\\v r}ek and Petr Sojka},\n          booktitle = {{Proceedings of the LREC 2010 Workshop on New\n               Challenges for NLP Frameworks}},\n          pages = {45--50},\n          year = 2010,\n          month = May,\n          day = 22,\n          publisher = {ELRA},\n          address = {Valletta, Malta},\n          note={\\url{http:\u002F\u002Fis.muni.cz\u002Fpublication\u002F884893\u002Fen}},\n          language={English}\n    }\n\n  [citing gensim in academic papers and theses]: https:\u002F\u002Fscholar.google.com\u002Fcitations?view_op=view_citation&hl=en&user=9vG_kV0AAAAJ&citation_for_view=9vG_kV0AAAAJ:NaGl4SEjCO4C\n\n  [design goals]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002Fintro.html#design-principles\n  [RaRe Technologies]: https:\u002F\u002Frare-technologies.com\u002Fwp-content\u002Fuploads\u002F2016\u002F02\u002Frare_image_only.png%20=10x20\n  [rare\\_tech]: \u002F\u002Frare-technologies.com\n  [Talentpair]: https:\u002F\u002Favatars3.githubusercontent.com\u002Fu\u002F8418395?v=3&s=100\n  [citing gensim in academic papers and theses]: https:\u002F\u002Fscholar.google.cz\u002Fcitations?view_op=view_citation&hl=en&user=9vG_kV0AAAAJ&citation_for_view=9vG_kV0AAAAJ:u-x6o8ySG0sC\n\n  [documentation and Jupyter Notebook tutorials]: https:\u002F\u002Fgithub.com\u002FRaRe-Technologies\u002Fgensim\u002F#documentation\n  [Vector Space Model]: https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVector_space_model\n  [unsupervised document analysis]: https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FLatent_semantic_indexing\n  [NumPy]: https:\u002F\u002Fnumpy.org\u002Finstall\u002F\n  [linking NumPy to a BLAS library]: https:\u002F\u002Fnumpy.org\u002Fdevdocs\u002Fbuilding\u002Fblas_lapack.html\n  [ATLAS]: https:\u002F\u002Fmath-atlas.sourceforge.net\u002F\n  [OpenBLAS]: https:\u002F\u002Fxianyi.github.io\u002FOpenBLAS\u002F\n  [source tar.gz]: https:\u002F\u002Fpypi.org\u002Fproject\u002Fgensim\u002F\n  [documentation]: https:\u002F\u002Fradimrehurek.com\u002Fgensim\u002F#install\n\n","Gensim 是一个用于主题建模、文档索引和相似度检索的 Python 库，特别适用于处理大规模文本数据。其核心功能包括多种高效的主题建模算法（如 LSA\u002FLSI\u002FSVD、LDA、HDP）及词向量模型（如 word2vec），支持内存独立处理超大语料库，并具备多核与分布式计算能力以加速处理过程。Gensim 提供了直观易用的接口设计，方便用户自定义输入流或扩展其他向量空间算法。该库非常适合自然语言处理与信息检索领域的研究人员和开发者使用，尤其是在需要对海量文本数据进行主题分析、文档相似度计算等任务时。","2026-06-11 02:51:03","top_language"]