Soon Myoung Chung

Text document clustering based on frequent word meaning sequences

Most of existing text clustering algorithms use the vector space model, which treats documents as bags of words. Thus, word sequences in the documents are ignored, while the meaning of natural languages strongly depends on them. In this paper, we propose two new text clustering algorithms, named Clustering based on Frequent Word Sequences (CFWS) and Clustering based on Frequent Word Meaning Sequences (CFWMS). A word is the word form showing in the document, and a word meaning is the concept expressed by synonymous word forms. A word (meaning) sequence is frequent if it occurs in more than certain percentage of the documents in the text database. The frequent word (meaning) sequences can provide compact and valuable information about those text documents. For experiments, we used the Reuters-21578 text collection, CISI documents of the Classic data set [Classic data set, ftp://ftp.cs.cornell.edu/pub/smart/], and a corpus of the Text Retrieval Conference (TREC) [High Accuracy Retrieval from Documents (HARD) Track of Text Retrieval Conference, 2004]. Our experimental results show that CFWS and CFWMS have much better clustering accuracy than Bisecting k-means (BKM) [M. Steinbach, G. Karypis, V. Kumar, A Comparison of Document Clustering Techniques, KDD-2000 Workshop on Text Mining, 2000], a modified bisecting k-means using background knowledge (BBK) [A. Hotho, S. Staab, G. Stumme, Ontologies improve text document clustering, in: Proceedings of the 3rd IEEE International Conference on Data Mining, 2003, pp. 541-544] and Frequent Itemset-based Hierarchical Clustering (FIHC) [B.C.M. Fung, K. Wang, M. Ester, Hierarchical document clustering using frequent itemsets, in: Proceedings of SIAM International Conference on Data Mining, 2003] algorithms.

Text Clustering with Feature Selection by Using Statistical Data

Feature selection is an important method for improving the efficiency and accuracy of text categorization algorithms by removing redundant and irrelevant terms from the corpus. In this paper, we propose a new supervised feature selection method, named CHIR, which is based on the chi2 statistic and new statistical data that can measure the positive term-category dependency. We also propose a new text clustering algorithm, named text clustering with feature selection (TCFS). TCFS can incorporate CHIR to identify relevant features (i.e., terms) iteratively, and the clustering becomes a learning process. We compared TCFS and the K-means clustering algorithm in combination with different feature selection methods for various real data sets. Our experimental results show that TCFS with CHIR has better clustering accuracy in terms of the F-measure and the purity.

Text document clustering based on neighbors

Clustering is a very powerful data mining technique for topic discovery from text documents. The partitional clustering algorithms, such as the family of k-means, are reported performing well on document clustering. They treat the clustering problem as an optimization process of grouping documents into k clusters so that a particular criterion function is minimized or maximized. Usually, the cosine function is used to measure the similarity between two documents in the criterion function, but it may not work well when the clusters are not well separated. To solve this problem, we applied the concepts of neighbors and link, introduced in [S. Guha, R. Rastogi, K. Shim, ROCK: a robust clustering algorithm for categorical attributes, Information Systems 25 (5) (2000) 345-366], to document clustering. If two documents are similar enough, they are considered as neighbors of each other. And the link between two documents represents the number of their common neighbors. Instead of just considering the pairwise similarity, the neighbors and link involve the global information into the measurement of the closeness of two documents. In this paper, we propose to use the neighbors and link for the family of k-means algorithms in three aspects: a new method to select initial cluster centroids based on the ranks of candidate documents; a new similarity measure which uses a combination of the cosine and link functions; and a new heuristic function for selecting a cluster to split based on the neighbors of the cluster centroids. Our experimental results on real-life data sets demonstrated that our proposed methods can significantly improve the performance of document clustering in terms of accuracy without increasing the execution time much.

A new image segmentation technique based on partition mode test

Abstract Image segmentation is one of the most important steps in a modern computerized machine vision system. This paper describes a simple, systematic one-pass image segmentation algorithm which is based on the partition mode test of pixels within a (2 × 2) window and assigning and updating label fields to the pixels of this window. A number of well chosen examples are shown to demonstrate the capabilities of the new algorithm.

Role-based access control for grid database services using the community authorization service

In this paper, we propose a role-based access control (RBAC) method for grid database services in open grid services architecture-data access and integration (OGSA-DAI). OGSA-DAI is an efficient grid-enabled middleware implementation of interfaces and services to access and control data sources and sinks. However, in OGSA-DAI, access control causes substantial administration overhead for resource providers in virtual organizations (VOs) because each of them has to manage a role-map file containing authorization information for individual grid users. To solve this problem, we used the community authorization service (CAS) provided by the globus toolkit to support the RBAC within the OGSA-DAI framework. The CAS grants the membership on VO roles to users. The resource providers then need to maintain only the mapping information from VO roles to local database roles in the role-map files, so that the number of entries in the role-map file is reduced dramatically. Furthermore, the resource providers control the granting of access privileges to the local roles. Thus, our access control method provides increased manageability for a large number of users and reduces day-to-day administration tasks of the resource providers, while they maintain the ultimate authority over their resources. Performance analysis shows that our method adds very little overhead to the existing security infrastructure of OGSA-DAI

Mining association rules using inverted hashing and pruning

In this paper, we propose a new algorithm named Inverted Hashing and Pruning (IHP) for mining association rules between items in transaction databases. The performance of the IHP algorithm was evaluated for various cases and compared with those of two well-known mining algorithms, Apriori algorithm [Proc. 20th VLDB Conf., 1994, pp. 487-499] and Direct Hashing and Pruning algorithm [IEEE Trans. on Knowledge Data Engrg. 9 (5) (1997) 813-825]. It has been shown that the IHP algorithm has better performance for databases with long transactions.

Multipass algorithms for mining association rules in text databases

Abstract. In this paper, we propose two new algorithms for mining association rules between words in text databases. The characteristics of text databases are quite different from those of retail transaction databases, and existing mining algorithms cannot handle text databases efficiently because of the large number of itemsets (i.e., words) that need to be counted. Two well-known mining algorithms, Apriori algorithm and Direct Hashing and Pruning (DHP) algorithm, are evaluated in the context of mining text databases, and are compared with the new proposed algorithms named Multipass-Apriori (M-Apriori) and Multipass-DHP (M-DHP). It has been shown that the proposed algorithms have better performance for large text databases.

Computer Architecture for a Surrogate File to a Very Large Data/Knowledge Base

This article presents techniques for managing a very large data/knowledge base to support multiple inference-mechanisms for logic programming. Because evaluation of goals can require accessing data from the extensional database, or EDB, in very general ways, one must often resort to indexing on all fields of the extensional database facts. This presents a formidable management problem in that the index data may be larger than the EDB itself. This problem becomes even more serious in this case of very large data/knowledge bases (hundreds of gigabytes), since considerably more hardware will be required to process and store the index data. In order to reduce the amount of index data considerably without losing generality, the authors form a surrogate file, which is a hashing transformation of the facts. Superimposed code words (SCW), concatenated code words (CCW), and transformed inverted lists (TIL) are possible structures for the surrogate file. since these transformations are quite regular and compact, the authors consider possible computer architecture for the processing of the surrogate file.

Parallel bisecting k-means with prediction clustering algorithm

In this paper, we propose a new parallel clustering algorithm, named Parallel Bisecting k-means with Prediction (PBKP), for message-passing multiprocessor systems. Bisecting k-means tends to produce clusters of similar sizes, and according to our experiments, it produces clusters with smaller entropy (i.e., purer clusters) than k-means does. Our PBKP algorithm fully exploits the data-parallelism of the bisecting k-means algorithm, and adopts a prediction step to balance the workloads of multiple processors to achieve a high speedup. We implemented PBKP on a cluster of Linux workstations and analyzed its performance. Our experimental results show that the speedup of PBKP is linear with the number of processors and the number of data points. Moreover, PBKP scales up better than the parallel k-means with respect to the dimension and the desired number of clusters.

A scalable algorithm for mining maximal frequent sequences using a sample

In this paper, we propose an efficient scalable algorithm for mining Maximal Sequential Patterns using Sampling (MSPS). The MSPS algorithm reduces much more search space than other algorithms because both the subsequence infrequency-based pruning and the supersequence frequency-based pruning are applied. In MSPS, a sampling technique is used to identify long frequent sequences earlier, instead of enumerating all their subsequences. We propose how to adjust the user-specified minimum support level for mining a sample of the database to achieve better overall performance. This method makes sampling more efficient when the minimum support is small. A signature-based method and a hash-based method are developed for the subsequence infrequency-based pruning when the seed set of frequent sequences for the candidate generation is too big to be loaded into memory. A prefix tree structure is developed to count the candidate sequences of different sizes during the database scanning, and it also facilitates the customer sequence trimming. Our experiments showed MSPS has very good performance and better scalability than other algorithms.