{ "info": { "author": "Moustafa Badawy", "author_email": "moustafa.badawym@gmail.com", "bugtrack_url": null, "classifiers": [], "description": "Blazor\n======\nA [Voci V-Blaze][1] Natural Language Processor\n\nPrerequisites\n-------------\n- **Python 2.7**\n- **NLTK 2.0** (with data)\n\nInstallation\n-------------\n > pip install blazor\n\nNLTK data packages are required as well. For installation instructions see http://www.nltk.org/data.html.\n\n----------\nThe transcript output of the V-Blaze web service is parsed and processed as follows:\n\nPreprocessing\n-------------\nThis is where the transcript generated from the Voci V-Blaze is loaded, tokenized into full sentences and cleaned from vocal tags (``, `++BRTH++`, `++AH++`, etc.). Call is divided into `Utterance`s, each utterance contains its respective properties (`id`, `confidence`, `start`, `end`, `channel`, `sentiment`, `gender`, `breath_count`, `silence_count`, and `sentences`).\n\nSentiment Analysis\n------------------\nA Sentiment Analyzer is trained on movie reviews and then utilized to classify (using NaiveBayesClassifier) each utterance sentence to one of `pos` or `neg` classes. Positive and Negative fractions are then stored for later indexing.\n\n\n [1]: http://www.vocitec.com/solutions/compare.php#blaze", "description_content_type": null, "docs_url": null, "download_url": "UNKNOWN", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "http://pypi.python.org/pypi/Blazor/", "keywords": null, "license": "LICENSE.txt", "maintainer": null, "maintainer_email": null, "name": "Blazor", "package_url": "https://pypi.org/project/Blazor/", "platform": "UNKNOWN", "project_url": "https://pypi.org/project/Blazor/", "project_urls": { "Download": "UNKNOWN", "Homepage": "http://pypi.python.org/pypi/Blazor/" }, "release_url": "https://pypi.org/project/Blazor/2.0.1/", "requires_dist": null, "requires_python": null, "summary": "Voci Natural Language Processor.", "version": "2.0.1" }, "last_serial": 1208006, "releases": { "2.0.1": [] }, "urls": [] }