THUCTC: 一个高效的中文文本分类工具包

THUCTC(THU Chinese Text Classification)是由清华大学自然语言处理实验室推出的中文文本分类工具包，能够自动高效地实现用户自定义的文本分类语料的训练、评测、分类功能。文本分类通常包括特征选取、特征降维、分类模型学习三个步骤。如何选取合适的文本特征并进行降维，是中文文本分类的挑战性问题。

THUCTC (THU Chinese Text Classification) is a Chinese text classification toolkit launched by the Natural Language Processing Laboratory of Tsinghua University, which can automatically and efficiently implement the training, evaluation and classification functions of user-customized text classification corpora. Text classification generally consists of three steps: feature selection, feature dimensionality reduction and classification model learning. How to select appropriate text features and perform dimensionality reduction is a challenging issue in Chinese text classification.

搜集汇总

背景与挑战

背景概述

THUCTC是清华大学开发的中文文本分类工具包，支持自定义语料训练与评测，采用bigram特征和高效分类算法，适用于开放领域长文本分类。