Herramientas/Tool Index
Textar
Textar consists of Python algorithms that aim to consolidate the natural language processing tools used by Argentina’s National Directorate of Public Data and Information.
This tool was developed by the Argentine Ministry of Modernization’s National Directorate of Public Data and Information. Its development arose from a need to assist the Chief of the Cabinet of Minister’s Secretariat of Parliamentary Relations and Administration in creating reports. These algorithms assist in the search and categorization of the questions that make up the secretariat’s reports upon integration into an interface.
FUNCTIONS:
Text search by semantic similarity.
Text tagging based on trainable categories.
Governments in other countries can repurpose or adapt this tool to improve the flow of the management of questions and answers received by various offices, as well as to improve communication with civil society.
You can see the source code here: http://github.com/datosgobar/textar