{ "info": { "author": "Shlomi Hod", "author_email": "shlomi.hod@gmail.com", "bugtrack_url": null, "classifiers": [ "Development Status :: 4 - Beta", "Intended Audience :: Science/Research", "License :: OSI Approved :: MIT License", "Natural Language :: English", "Operating System :: OS Independent", "Programming Language :: Python", "Programming Language :: Python :: 3", "Programming Language :: Python :: 3.5", "Programming Language :: Python :: 3.6", "Programming Language :: Python :: 3.7", "Programming Language :: Python :: 3 :: Only", "Topic :: Scientific/Engineering :: Artificial Intelligence", "Topic :: Text Processing :: Linguistic" ], "description": "Responsibly\n===========\n\n.. image:: https://img.shields.io/badge/docs-passing-brightgreen.svg\n :target: https://docs.responsibly.ai\n\n.. image:: https://img.shields.io/gitter/room/nwjs/nw.js.svg\n :alt: Join the chat at https://gitter.im/ResponsiblyAI/responsibly\n :target: https://gitter.im/ResponsiblyAI/responsibly\n\n.. image:: https://img.shields.io/travis/ResponsiblyAI/responsibly/master.svg\n :target: https://travis-ci.org/ResponsiblyAI/responsibly\n\n.. image:: https://img.shields.io/coveralls/ResponsiblyAI/responsibly/master.svg\n :target: https://coveralls.io/r/ResponsiblyAI/responsibly\n\n.. image:: https://img.shields.io/scrutinizer/g/ResponsiblyAI/responsibly.svg\n :target: https://scrutinizer-ci.com/g/ResponsiblyAI/responsibly/?branch=master\n\n.. image:: https://img.shields.io/pypi/v/responsibly.svg\n :target: https://pypi.org/project/responsibly\n\n.. image:: https://img.shields.io/github/license/ResponsiblyAI/responsibly.svg\n :target: https://docs.responsibly.ai/about/license.html\n\n**Toolkit for Auditing and Mitigating Bias and Fairness**\n**of Machine Learning Systems \ud83d\udd0e\ud83e\udd16\ud83e\uddf0**\n\n*Responsibly* is developed for **practitioners** and **researchers** in mind,\nbut also for learners. Therefore, it is compatible with\ndata science and machine learning tools of trade in Python,\nsuch as Numpy, Pandas, and especially **scikit-learn**.\n\nThe primary goal is to be one-shop-stop for **auditing** bias\nand fairness of machine learning systems, and the secondary one\nis to mitigate bias and adjust fairness through\n**algorithmic interventions**.\nBesides, there is a particular focus on **NLP** models.\n\n*Responsibly* consists of three sub-packages:\n\n1. ``responsibly.dataset``\n Collection of common benchmark datasets from fairness research.\n\n2. ``responsibly.fairness``\n Demographic fairness in binary classification,\n including metrics and algorithmic interventions.\n\n3. ``responsibly.we``\n Metrics and debiasing methods for bias (such as gender and race)\n in word embedding.\n\nFor fairness, *Responsibly*'s functionality is aligned with the book\n`Fairness and Machine Learning\n- Limitations and Opportunities `_\nby Solon Barocas, Moritz Hardt and Arvind Narayanan.\n\nIf you would like to ask for a feature or report a bug,\nplease open a\n`new issue `_\nor write us in `Gitter `_.\n\nRequirements\n------------\n\n- Python 3.5+\n\nInstallation\n------------\n\nInstall responsibly with pip:\n\n.. code:: sh\n\n $ pip install responsibly\n\nor directly from the source code:\n\n.. code:: sh\n\n $ git clone https://github.com/ResponsiblyAI/responsibly.git\n $ cd responsibly\n $ python setup.py install\n\nCitation\n--------\n\nIf you have used *Responsibly* in a scientific publication,\nwe would appreciate citations to the following:\n\n::\n\n @Misc{,\n author = {Shlomi Hod},\n title = {{Responsibly}: Toolkit for Auditing and Mitigating Bias and Fairness of Machine Learning Systems},\n year = {2018--},\n url = \"http://docs.responsibly.ai/\",\n note = {[Online; accessed ]}\n }\n\nRevision History\n================\n\n0.1.1 (2019/08/04)\n------------------\n\n- Fix a dependencies issue with ``smart_open``\n\n- Change URLs to https\n\n0.1.0 (2019/07/31)\n------------------\n\n- Rename the project to ``responsibly`` from ``ethically``\n\n- Word embedding bias\n\n - Improve functionality of ``BiasWordEmbedding``\n\n- Threshold fairness interventions\n\n - Fix bugs with ROCs handling\n - Improve API and add functionality (``plot_thresholds``)\n\n0.0.5 (2019/06/14)\n------------------\n\n- Word embedding bias\n\n - Fix bug in computing WEAT\n\n - Computing and plotting factual property\n association to projections on a bias direction,\n similar to WEFAT\n\n\n0.0.4 (2019/06/03)\n------------------\n\n- Word embedding bias\n\n - Unrestricted ``most_similar``\n\n - Unrestricted ``generate_analogies``\n\n - Running specific experiments with ``calc_all_weat``\n\n - Plotting clustering by classification\n of biased neutral words\n\n\n0.0.3 (2019/04/10)\n------------------\n\n- Fairness in Classification\n\n - Three demographic fairness criteria\n\n - Independence\n - Separation\n - Sufficiency\n\n - Equalized odds post-processing algorithmic interventions\n - Complete two notebook demos (FICO and COMPAS)\n\n- Word embedding bias\n\n - Measuring bias with WEAT method\n\n- Documentation improvements\n\n- Fixing security issues with dependencies\n\n\n0.0.2 (2018/09/01)\n------------------\n\n- Word embedding bias\n\n - Generating analogies along the bias direction\n - Standard evaluations of word embedding (word pairs and analogies)\n - Plotting indirect bias\n - Scatter plot of bias direction projections between two word embedding\n - Improved verbose mode\n\n\n0.0.1 (2018/08/17)\n------------------\n\n- Gender debiasing for word embedding based on Bolukbasi et al.\n\n\n", "description_content_type": "text/x-rst", "docs_url": null, "download_url": "", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "https://docs.responsibly.ai", "keywords": "", "license": "MIT", "maintainer": "", "maintainer_email": "", "name": "responsibly", "package_url": "https://pypi.org/project/responsibly/", "platform": "", "project_url": "https://pypi.org/project/responsibly/", "project_urls": { "Homepage": "https://docs.responsibly.ai" }, "release_url": "https://pypi.org/project/responsibly/0.1.1/", "requires_dist": [ "numpy (>=1.15)", "scipy (>=1.1)", "pandas (>=0.23)", "matplotlib (<3,>=2.2)", "seaborn (>=0.9)", "scikit-learn (>=0.19)", "gensim (>=3.7)", "tabulate (>=0.8)", "six (>=1.10)", "click (>=6.0)", "tqdm (>=4.24)", "mlxtend (<0.17,>=0.13)" ], "requires_python": "", "summary": "Toolkit for Auditing and Mitigating Bias and Fairness of Machine Learning Systems \ud83d\udd0e\ud83e\udd16\ud83e\uddf0", "version": "0.1.1" }, "last_serial": 5631536, "releases": { "0.1.0": [ { "comment_text": "", "digests": { "md5": "1cfc2f13654cc465021939d124bbb044", "sha256": "3ae8dc2855fbc9afe3605f08fcd4ed67fb82af1a32bbcb790348542cd5664c63" }, "downloads": -1, "filename": "responsibly-0.1.0-py3-none-any.whl", "has_sig": false, "md5_digest": "1cfc2f13654cc465021939d124bbb044", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": null, "size": 28201168, "upload_time": "2019-08-04T07:55:53", "url": "https://files.pythonhosted.org/packages/5f/11/be4eddcb29418446f067a414c74982f711dc9079a7abed5f1a944ea224ad/responsibly-0.1.0-py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "e457e63d2ae385e4bf99fa16303c16b8", "sha256": "7f55fce3db8944596e0e9a88763d29c8690c593e5efee676fdddfb7c833f229c" }, "downloads": -1, "filename": "responsibly-0.1.0.tar.gz", "has_sig": false, "md5_digest": "e457e63d2ae385e4bf99fa16303c16b8", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 28139232, "upload_time": "2019-08-04T07:56:22", "url": "https://files.pythonhosted.org/packages/0f/9a/3edcb1336e6bb65784465e8ad129502d497515d6beb44f90b322e31fa4f1/responsibly-0.1.0.tar.gz" } ], "0.1.1": [ { "comment_text": "", "digests": { "md5": "1a87e9fa58f3c914f89d5cddf956ab2a", "sha256": "67b378855640585cbec7c88ba41d98dfb5ae0ee0dd0c5d653703037239d8d8f2" }, "downloads": -1, "filename": "responsibly-0.1.1-py3-none-any.whl", "has_sig": false, "md5_digest": "1a87e9fa58f3c914f89d5cddf956ab2a", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": null, "size": 28201190, "upload_time": "2019-08-04T20:40:47", "url": "https://files.pythonhosted.org/packages/10/f6/dc0af2236f1b5095d619065ed15d16784e15230151b6750fa7777645a972/responsibly-0.1.1-py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "2576c89faad15d9d03e0d18829c62148", "sha256": "64e6701389a671543edc46b911203fffeddaab12180e246ddfc2d2bb8b98404c" }, "downloads": -1, "filename": "responsibly-0.1.1.tar.gz", "has_sig": false, "md5_digest": "2576c89faad15d9d03e0d18829c62148", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 28139249, "upload_time": "2019-08-04T20:41:39", "url": "https://files.pythonhosted.org/packages/34/16/e87dc9b02659174646ecc023265c4de84fe326011d3466ada54313db1004/responsibly-0.1.1.tar.gz" } ] }, "urls": [ { "comment_text": "", "digests": { "md5": "1a87e9fa58f3c914f89d5cddf956ab2a", "sha256": "67b378855640585cbec7c88ba41d98dfb5ae0ee0dd0c5d653703037239d8d8f2" }, "downloads": -1, "filename": "responsibly-0.1.1-py3-none-any.whl", "has_sig": false, "md5_digest": "1a87e9fa58f3c914f89d5cddf956ab2a", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": null, "size": 28201190, "upload_time": "2019-08-04T20:40:47", "url": "https://files.pythonhosted.org/packages/10/f6/dc0af2236f1b5095d619065ed15d16784e15230151b6750fa7777645a972/responsibly-0.1.1-py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "2576c89faad15d9d03e0d18829c62148", "sha256": "64e6701389a671543edc46b911203fffeddaab12180e246ddfc2d2bb8b98404c" }, "downloads": -1, "filename": "responsibly-0.1.1.tar.gz", "has_sig": false, "md5_digest": "2576c89faad15d9d03e0d18829c62148", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 28139249, "upload_time": "2019-08-04T20:41:39", "url": "https://files.pythonhosted.org/packages/34/16/e87dc9b02659174646ecc023265c4de84fe326011d3466ada54313db1004/responsibly-0.1.1.tar.gz" } ] }