Document Clustering using Improved K- meansAlgorithm

2016

In today's era of World Wide Web, there is a<br> tremendous proliferation in the amount of digitized text<br> documents. As there is huge collection of documents on the web,<br> there is a need of grouping the set of documents into clusters.<br> Document clustering plays an important role in effectively<br> navigating and organizing the documents. K-Means clustering<br> algorithm is the most commonly document clustering algorithm<br> because it can be easily implemented and is the most efficient one<br> in terms of execution times. The major problem with this<br> algorithm is that it is quite sensitive to selection of initial cluster<br> centroids. The algorithm takes the initial cluster center arbitrarily<br> so it does not always promise good clustering results. If the initial<br> centroids are incorrectly determined, the remaining data points<br> with the same similarity scores may fall into the dif...

2014

Clustering is an automatic learning technique aimed at grouping a set of objects into subsets or clusters. Objects in the same cluster should be as similar as possible, whereas objects in one cluster should be as dissimilar as possible from objects in the other clusters. Document clustering has become an increasingly important task in analysing huge documents. The challenging aspect to analyse the enormous documents is to organise them in such a way that facilitates better search and knowledge extraction without introducing extra cost and complexity. Document clustering has played an important role in many fields like information retrieval and data mining. In this paper, first Document Clustering has been proposed using Hierarchical Agglomerative Clustering and K-Means Clustering Algorithm.Here, the approach is purely based on the frequency count of the terms present in the documents where context of the documents are totally ignored. Therefore, the method is modified by incorporati...

International Journal of Applied Information Systems, 2012

Document clustering is automatic organization of documents into clusters so that documents within a cluster have high similarity in comparison to documents in other clusters. It has been studied intensively because of its wide applicability in various areas such as web mining, search engines, and information retrieval. It is measuring similarity between documents and grouping similar documents together. It provides efficient representation and visualization of the documents; thus helps in easy navigation also. In this paper, we have given overview of various document clustering methods studied and researched since last few years, starting from basic traditional methods to fuzzy based, genetic, coclustering, heuristic oriented etc. Also, the document clustering procedure with feature selection process, applications, challenges in document clustering, similarity measures and evaluation of document clustering algorithm is explained.

Document Clustering is the collection of similar documents into classes and the similarity is some function on the document. Document Clustering need not require any separate training process and manual tagging group in advance. The documents used in the same clusters are more similar, while the documents used in different clusters are more dissimilar. It is one of the familiar technique used in data analysis and is used in many areas including data mining, statistics and image analysis. The traditional clustering approaches lose its algorithmic approach when handling high dimensional data. For this, a new K-Means Clustering technique is proposed in this work. Here Cosine Similarity of Vector Space Model is used as the centroid for clustering. Using this approach, the documents can be clustered efficiently even when the dimension is high because it uses vector space representation for documents which is suitable for high dimensions.

2018

In document clustering system, some documents with the same similarity scores may fall into different clusters instead of same cluster due to calculate similarity distance between pairs of documents based on geometric measurements. To tackle this point, probability distribution of K-Means (PD K-Means) algorithm is proposed. In this system, documents are clustered based on proposed probability distribution equation instead of similarity measure between objects. It can also solve initial centroids problems of K-Means by using Systematic Selection of Initial Centroid (SSIC) approach. So, it not only can generate compact and stable results but also eliminates initial cluster problem of K-Means. According to the experiment, F-measure values increase about 0.28 in 20 NewsGroup dataset, 0.26 in R8 and 0.14 in R52 from Reuter21578 datasets. The evaluations demonstrate that the proposed solution outperforms than original method and can be applied for various standard and unsupervised

Text clustering is a text mining technique used to group text documents into groups (or clusters) based on similarity of content. This organization (i.e. clustering) is so as to make documents more understandable and easier to search the relevant information, easier to process, and even more efficient in utilizing communication bandwidth and storage space. An example is clustering results of a web search engine operation into groups of similar documents. Many text clustering algorithms have been developed using different approaches, but none can be said to be the best. The choice of a particular algorithm is a big issue to text clustering system developers. K Means is arguably the most popular text clustering algorithm. However, just like the others, it must be having its own weaknesses. In this paper, we explore the K Means algorithm as well as its variants and discuss their appropriateness in text clustering. We describe the characteristics of the algorithms accompanied by some examples and illustrations in an attempt to discover the strengths and weaknesses. The paper thus gives an in depth view of the K Means algorithms, discusses the appropriateness of the algorithms, and also gives guidance to researchers of text mining concerning the choice of K Means for text clustering.

—Rapid advancements of smart technologies, permits the individuals and organizations to store large number of documents in repositories. But it is quite difficult to retrieve the relevant documents from these massive collections. Document clustering is the process of organizing such massive document collections into meaningful clusters. It is simple and less tedious to find relevant documents, if documents are clustered on the basis of topic or category. There are various document clustering algorithms available for effectively organizing the documents such that a document is close to its related documents. This paper presents various clustering techniques that are being used in text mining.

Document clustering is becoming more and more important with the abundance of text documents available through World Wide Web and corporate document management systems. Document clustering is the process of categorizing text document into a systematic cluster or group, such that the documents in the same cluster are similar whereas the documents in the other clusters are dissimilar. This survey includes the information about data mining clustering technique for unstructured data.

International Journal of Computer Applications, 2014

With the growth of Internet, large amount of text data is increasing, which are created by different media like social networking sites, web, and other informatics sources, etc. This data is in unstructured format which makes it tedious to analyze it, so we need methods and algorithms which can be used with various types of text formats. Clustering is an important part of the data mining. Clustering is the process of dividing the large &similar type of text into the same class. Clustering is widely used in many applications like medical, biology, signal processing, etc. This paper briefly covers the various kinds of text clustering algorithm, present scenario of the text clustering algorithm, analysis and comparison of various aspects which contain sensitivity, stability. Algorithm contains traditional clustering like hierarchal clustering, density based clustering and self-organized map clustering.

2012

There are two important problems worth conducting research in the fields of personalized information services based on user model. One is how to get and describe user personal information, i.e. building user model, the other is how to organize the information resources, i.e. document clustering. It is difficult to find out the desired information without a proper clustering algorithm. Several new ideas have been proposed in recent years. But most of them only took into account the text information, but some other useful information may have more contributions for documents clustering, such as the text size, font and other appearance characteristics, so called visual features.In this paper we introduce a new technique called Closed Document Clustering Method (CDCM) by using advanced clustering metrics. This method enhances the previous method of cluster the scientific documents based on visual features, so called VF-Clustering algorithm. Five kinds of visual features of documents are...

This paper presents a new document clustering method based on frequent co-occurring words. We first employ the Singular Value Decomposition, and then group the words into clusters called word representatives as substitution of the corresponding words in the original documents. Next, we extract the frequent word representative sets by Apriori. Subsequently, each document is designated to a basic unit described by the frequent word representative set, from which we can get the ultimate clusters by hierarchical clustering. The major advantage of our method is that it can produce the cluster description by the frequent word representatives and then by the corresponding words in the clustering process without any extra works. Compared with the state-of-the-art UPGMA method on benchmark datasets, our method has better performance in terms of the entropy and cluster purity.

This paper deals with e-government documents multilayered clustering based on hybrid approach that combines Fuzzy-C-mean algorithm, cosine similarity and semantic similarity measures. The system described here is intended to reduce response time between citizen's questions and government answers, either to eliminate or to minimize the role of subject matter experts. Layers of documents are defined by key terms that are discovered by a clustering engine that we named ADVANSE. After short overview of clustering algorithms the paper concentrates step by step on the functionality of ADVANSE. Finally, concluding remarks emphasize some important features of this approach and gave future research directions.

International Journal of Knowledge Based Computer Systems , 2016

Cluster analysis is an unsupervised learning approach that aims to group the objects into different groups or clusters. So that each cluster can contain similar objects with respect to any predefined condition. Text document clustering is the important technique of text mining in efficiently organizing the large volume of documents into a small number of significant clusters. The main objective of this research work is to cluster the collection of documents into related groups based on the contents of the particular documents. In order to perform this clustering task, this research work makes use of two existing algorithms, namely K-means and Bisecting K-means algorithm, and also this research work proposes a new clustering algorithm namely Enhanced-Bisecting K-means algorithm. From the experimental results it is observed that the proposed algorithm gives the better clustering accuracy than other algorithms.

2015

Data mining is the process of non-trivial discovery from implied, previously unknown, and potentially useful information from data in large databases. Hence it is a core element in knowledge discovery, often used synonymously. Clustering, one of technique for data mining used for grouping similar terms together. Earlier statistical analysis used in text mining depends on term frequency. Then, new concept based text mining model was introduced which analyses terms. Clustering of document is useful for the purpose of document organization, summarization, and information retrieval in an efficient way. Initially, clustering is applied for enhancing the information retrieval techniques. Of late, clustering techniques have been applied in the areas which involve browsing the gathered data or in categorizing the outcome provided by the search engines for the reply to the query raised by the users. In this paper, we are providing a comprehensive survey over the document clustering.