Narrow Results

This paper proposes document clustering using Kohonen neural network with pairwise constraints. This algorithm works directly on textual information without mapping document into some representation that has quantitative features. The input level of the proposed neural network can directly receive a qualitative value without mapping the qualitative value into the numerical value. The proposed neural network is based on the architecture of text processing Kohonen neural network, the concepts of dissimilarity measure of symbolic objects and pairwise constrained concepts. As a result, the model can successfully assign cluster label to the objects.

Document clustering is one of the most effective techniques to organize documents in an unsupervised manner. In this paper, an Incremental method for document Clustering based on Information Bottleneck theory (ICIB) is presented. The ICIB is designed to improve the accuracy and efficiency of document clustering, and resolve the issue that an arbitrary choice of document similarity measure could produce an inaccurate clustering result. In our approach, document similarity is calculated using information bottleneck theory and documents are grouped incrementally. A first document is selected randomly and classified as one cluster, then each remaining document is processed incrementally according to the mutual information loss introduced by the merger of the document and each existing cluster. If the minimum value of mutual information loss is below a certain threshold, the document will be added to its closest cluster; otherwise it will be classified as a new cluster. The incremental clustering process is low-precision and order-dependent, which cannot guarantee accurate clustering results. Therefore, an improved sequential clustering algorithm (SIB) is proposed to adjust the intermediate clustering results. In order to test the effectiveness of ICIB method, ten independent document subsets are constructed based on the 20NewsGroup and Reuters-21578 corpora. Experimental results show that our ICIB method achieves higher accuracy and time performance than K-Means, AIB and SIB algorithms.

This paper proposes a combination weighting algorithm using relative entropy for document clustering. Combination weighting is widely used in multiple attribute decision making (MADM) problem. However, there exist two difficulties to hinder the applications of combination weighting on document clustering. First, combination weighting is based on the integration of subjective weighting and objective weighting. However, there are so many attributes in documents that the subjective weights which rely on manual annotation by experts are impracticable. Secondly, a document data object might contain hundreds or even thousands of features. It is an extremely time-consuming task to calculate the combination weights. To address the issues, we suggest to simplify the combination weighting by not distinguishing subjective weight and objective weight. Meanwhile, we choose relative entropy method to reduce running time. In our algorithm, we obtain a combination weight set with 14 combination forms. The experiments on real document data show that both on the AC/PR/RE measures and the mutual information (MI) measure, the proposed CWRE-sIB algorithm is superior to the original sequential information bottleneck (sIB) algorithm and a series of weighting-sIB algorithms, which are built by applying a single weighting scheme to the original sIB algorithm.

Grouping documents based on their textual content is an important application of clustering referred to as text clustering. This paper deals with two issues in text clustering which are the detection of non-disjoint groups and the representation of textual data. In fact, a text document can discuss several topics and then, it must belong to several groups. The learning algorithm must be able to produce non-disjoint clusters and assigns documents to several clusters. Given that text documents are considered as unstructured data, the application of a learning algorithm requires to prepare a set of documents for numerical analysis by using the vector space model (VSM). This representation of text avoids correlation between terms and does not give importance to the order of words in the text. Therefore, we present in this paper an unsupervised learning method, based on the word sequence kernel, where the correlation between adjacent words in text and the possibility of document to belong to more than one cluster are not ignored. In addition, to facilitate the use of this method in text-analytic practice, we present the "DocCO" software which is publicly available. Experiments performed on several text collections show that the proposed method outperforms existing overlapping methods using VSM representation in terms of clustering accuracy.

XML is increasingly important in data exchange and information management. A great deal of efforts have been spent in developing efficient techniques for storing, querying, indexing and accessing XML documents. In this paper we propose a new approach to clustering XML data. In contrast to previous work, which focused on documents defined by different DTDs, the proposed method works for documents with the same DTD. Our approach is to extract features from documents, modeled by ordered labeled trees, and transform the documents to vectors in a high-dimensional Euclidean space based on the occurrences of the features in the documents. We then reduce the dimensionality of the vectors by principal component analysis (PCA) and cluster the vectors in the reduced dimensional space. The PCA enables one to identify vectors with co-occurrent features, thereby enhancing the accuracy of the clustering. We also discuss an extension of our techniques to XML retrieval. Experimental results based on documents obtained from Wisconsin's XML data bank show the effectiveness and good performance of the proposed techniques.

This paper describes a Web information search tool called WebYacht. The goal of WebYacht is to solve the problem of imprecise search results in current Web search engines. Due to incomplete information given by users and the diversified information published on the Web, conventional document ranking based on an automatic assessment of document relevance to the query may not be the best approach when little information is given as in most cases. In order to clarify the ambiguity of the short queries given by users, WebYacht adopts cluster-based browsing model as well as relevance feedback to facilitate Web information search. The idea is to have users give two to three times more feedback in the same amount of time that would be required to give feedback for conventional feedback mechanisms. With the assistance of cluster-based representation provided by WebYacht, a lot of browsing labor can be reduced. In this paper, we explain the techniques used in the design of WebYacht and compare the performances of feedback interface designs and to conventional similarity ranking search results.

With the rapid growth of the Internet, how to get information from this huge information space becomes an even more important problem. In this paper, An Intelligence Chinese Document Semantic Indexing System; ICDSIS, is proposed. Some new technologies are integrated in ICDSIS to obtain good performance. ICDSIS is composed of four key procedures. A parallel, distributed and configurable Spider is used for information gather; a multi-hierarchy document classification approach combining the information gain initially processes gathered web documents; a swarm intelligence based document clustering method is used for information organization; a concept-based retrieval interface is applied for user interactive retrieval. ICDSIS is an all-sided solution for information retrieval on the Internet.

Despite the rapid development, the field of data mining and knowledge discovery (DMKD) is still vaguely defined and lack of integrated descriptions. This situation causes difficulties in teaching, learning, research, and application. This paper surveys a large collection of DMKD literature to provide a comprehensive picture of current DMKD research and classify these research activities into high-level categories using grounded theory approach; it also evaluates the longitudinal changes of DMKD research activities during the last decade.

Information Retrieval is the most predominant topic in the field of Information Systems as the generation of data over various systems and channels is growing every day. The proposed system which works on the offline basis is designed for the purpose of organizing data in a defined manner and for the purpose of increasing relevancy in the retrieval of the required information from this largely generated data. In the proposed system two novel algorithms, Dynamic Path Selection Clustering (DPSC) algorithm for clustering and the Rearward Binary Window Match (RBWM) algorithm for search process are introduced to overcome the difficulty in data organization and search. The evaluation of the entire system is done and the results are compared along with the results of the existing techniques.

Document clustering is a significant research issue in information retrieval and text mining. Traditionally, most clustering methods were based on the vector space model which has a few limitations such as high dimensionality and weakness in handling synonymous and polysemous problems. Latent semantic indexing (LSI) is able to deal with such problems to some extent. Previous studies have shown that using LSI could reduce the time in clustering a large document set while having little effect on clustering accuracy. However, when conducting clustering upon a small document set, the accuracy is more concerned than efficiency. In this paper, we demonstrate that LSI can improve the clustering accuracy of a small document set and we also recommend the dimensions needed to achieve the best clustering performance.

Recently, MapReduce-based implementations of clustering algorithms have been developed to cope with the Big Data phenomenon, and they show promising results particularly for the document clustering problem. In this paper, we extend an efficient data partitioning method based on the relational analysis (RA) approach and applied to the document clustering problem, called PDC-Transitive. The introduced heuristic is parallelised using the MapReduce model iteratively and designed with a single reducer which represents a bottleneck when processing large data, we improved the design of the PDC-Transitive method to avoid the data dependencies and reduce the computation cost. Experiment results on benchmark datasets demonstrate that the enhanced heuristic yields better quality results and requires less computing time compared to the original method.

The clustering is an important data analysis technique. However, clustering high-dimensional data like documents needs more effort in order to extract the richness relevant information hidden in the multidimensionality space. Recently, document clustering algorithms based on metaheuristics have demonstrated their efficiency to explore the search area and to achieve the global best solution rather than the local one. However, most of these algorithms are not practical and suffer from some limitations, including the requirement of the knowledge of the number of clusters in advance, they are neither incremental nor extensible and the documents are indexed by high-dimensional and sparse matrix. In order to overcome these limitations, we propose in this paper, a new dynamic and incremental approach (CS_LSI) for document clustering based on the recent cuckoo search (CS) optimization and latent semantic indexing (LSI). Conducted Experiments on four well-known high-dimensional text datasets show the efficiency of LSI model to reduce the dimensionality space with more precision and less computational time. Also, the proposed CS_LSI determines the number of clusters automatically by employing a new proposed index, focused on significant distance measure. This later is also used in the incremental mode and to detect the outlier documents by maintaining a more coherent clusters. Furthermore, comparison with conventional document clustering algorithms shows the superiority of CS_LSI to achieve a high quality of clustering.

In this article, we examine an algorithm for document clustering using a similarity graph. The graph stores words and common phrases from the English language as nodes and it can be used to compute the degree of semantic similarity between any two phrases. One application of the similarity graph is semantic document clustering, that is, grouping documents based on the meaning of the words in them. Since our algorithm for semantic document clustering relies on multiple parameters, we examine how fine-tuning these values affects the quality of the result. Specifically, we use the Reuters-21578 benchmark, which contains $11,362$ newswire stories that are grouped in 82 categories using human judgment. We apply the k-means clustering algorithm to group the documents using a similarity metric that is based on keywords matching and one that uses the similarity graph. We evaluate the results of the clustering algorithms using multiple metrics, such as precision, recall, f-score, entropy, and purity.

Three linear and non-linear dimension reduction algorithms, LSI, Isomap and SIE for processing high dimensional data in natural language documents are presented in this paper. Document data processed by the three algorithms is then clustered by bisecting K-means clustering algorithm. Experimental results show that the performance of SIE algorithm and LSI algorithm in document clustering are basically equivalent and better than the benchmark, while the performance of Isomap algorithm is worse than the benchmark.