Clustering text data streams is an important issue in data mining community and has a number of applications such as news group filtering, text crawling, document organization and topic detection and tracing etc. However, most methods are similarity-based approaches and use the TF*IDF scheme to represent the semantics of text data and often lead to poor clustering quality. In this paper, we fir...

Text clustering is of great importance in data mining, information fusion, artificial intelligence and some other fields. There are many methods in literatures that can be used to classify text. Most of them require some parameters, such as the number of categories, which should be assigned in advance or estimated in classifying process. However, it is difficult to determine these quantities in...

In order to build human cognition features into the procedure of clustering, this paper introduces a novel text clustering system, CogTCA (Cognitive Text Clustering with Ants), which (1) represents texts according to four cognitive situation dimensions in form of cognitive situation matrices and vectors rather than canonical sparse matrices of high dimensions, (2) proposes several new similarit...

In this project report, I will evaluate the several text clustering approaches and how they can be used for the purpose of text classification. The particular task is topic classification of 20 Newsgroup dataset and sentiment classification restaurant reviews dataset. Future direction for improving the results will also be discussed.

In this paper, a novel text clustering technique is proposed to summarize text documents. The clustering method, so called ‘Ensemble Clustering Method’, combines both genetic algorithms (GA) and particle swarm optimization (PSO) efficiently and automatically to get the best clustering results. The summarization with this clustering method is to effectively avoid the redundancy in the summarized...

This work presents document clustering experiments performed over noisy texts (i.e. text that have been extracted through an automatic process like speech or character recognition). The effect of recognition errors on different clustering techniques is measured through the comparison of the results obtained with clean (manually typed texts) and noisy (automatic speech transcripts affected by 30...

Text clustering typically involves clustering in a high dimensional space, which appears difficult with regard to virtually all practical settings. In addition, given a particular clustering result it is typically very hard to come up with a good explanation of why the text clusters have been constructed the way they are. In this paper, we propose a new approach for applying background knowledg...

This paper addresses the problem of learning to classify texts by exploiting information derived from both training and testing sets. To accomplish this, clustering is used as a complementary step to text classification, and is applied not only to the training set but also to the testing set. This approach allows us to estimate the location of the testing examples and the structure of the whole...

The vast amount of textual information available in electronic form is growing at a staggering rate in recent times. The task of mining useful or interesting frequent itemsets (words/terms) from very large text databases that are formed as a result of the increasing number of textual data still seems to be a quite challenging task. A great deal of attention in research community has been receiv...