LibShortText - 短文本分类

Chih-Jen Lin的新作。
http://www.csie.ntu.edu.tw/~cjlin/

• LIBSVM: a simple and easy-to-use support vector machines tool for classification (C-SVC, nu-SVC), regression (epsilon-SVR, nu-SVR), and distribution estimation.
  It includes a GUI for both classification and regression. Version 1.0 released in April 2000. Current Version: 3.17, April 2013.
  See software document published in ACM TIST: pdf, ps.gz, ACM digital lib.
• LIBLINEAR: a library for large linear classification. It is very suitable for document classification. Version 1.0 released in April 2007. Current Version: 1.93, January 2013.
• LibShortText: a library for short-text classification and analysis. Version 1.0 released in February 2013. Current Version: 1.0, February 2013.

Introduction

LibShortText is an open source tool for short-text classification and analysis. It can handle the classification of, for example, titles, questions, sentences, and short messages. Main features of LibShortText include

• It is more efficient than general text-mining packages. On a typical computer, processing and training 10 million short texts takes only around half an hour.
• The fast training and testing is built upon the linear classifier LIBLINEAR
• Default options often work well without tedious tuning.
• An interactive tool for error analysis is included. Based on the property that each short text contains few words, LibShortText provides details in predicting each text.