机构地区: 华南理工大学工商管理学院
出 处: 《图书情报工作》 2011年第2期103-106,共4页
摘 要: 面对海量、异构、动态的文本信息,对文本进行自动分类具有重要意义。文本分类的发展与模式识别的发展密切相关。文本分类具有的类目多、样本数目多、噪音多、各类别样本数目不均衡等特点,使各模式识别算法在应用于文本分类时存在许多缺点。近年来逐步发展起来的群集智能(Swarm Intelligence)理论和方法为文本分类提供一种新的智能化手段。将蚁群智能算法尝试性引入文本分类领域,构建基于蚁群智能的文本分类模型,并在文本数据集上进行测试和比较,结果表明该模型可以较好地应用于文本分类。 It's significance for us to study text auto classification, when we face so much dynamic information. The development of text classification has a close connection with pattern recognition. However, some peculiarity of text classification, such as it has many classes, much noise, and excessive samples, make pattern recognition difficult to classify texts, recently, swarm intelligence provides a new intellectualized method to text classification. This paper tentatively leads ant colony optimization, a ripe algorithm of swarm intel- ligence, into text classification. We construct a text classification model based on ant colony optimization, and test it. The result shows the model can accurately be used to classify texts.
领 域: [自动化与计算机技术] [自动化与计算机技术]