Dissertation > Excellent graduate degree dissertation topics show

Research and Application of K-Means Algorithm Based on Concept Lattice

Author: LiYanXia
Tutor: ShiYiMin
School: Dalian Maritime University
Course: Computer Science and Technology
Keywords: Concept lattice Text Clustering Text representation K-Means algorithm Concept Similarity
CLC: TP18
Type: Master's thesis
Year: 2010
Downloads: 108
Quote: 0
Read: Download Dissertation

Abstract


With the rapid development of the Internet, search engines become the main channel for people to obtain information. However, a search engine search results move thousands of all categories of information mixed with the user to find the information they want, such as finding a needle in a haystack. An effective way to improve the quality of the search engines is a web application text clustering technology similar to web search results text gathered into a class. Web search results clustering can provide users with easy to navigate information navigation to help users quickly navigate to comply with its own query subject categories, thereby improving the efficiency of the search engine search. Clustering is not know in advance in class case, the collection of objects at the relevant degree of similarity grouping process. Clustering before the majority of the representation of the text is used in the vector space model, and on this basis to calculate the similarity. TF X IDF (Term Frequency X Inverse Document Frequency) vector space model to calculate the weight. Its advantage is to reflect the importance of the keywords for the text, but this indicates that the model brings two problems: (1) a dimension of the feature vector of the text is too high; (2) text is seen by a group of n pay entry vector vector space, the assumption is no semantic relationship between words, but the reality of the text, the wording is often semantic association, and therefore the reliability of the results have been affected. The concept lattice is an ordered set of a set of concepts, the process of building concept lattice is the concept of clustering process. Concept lattice, the extension of the concept for the collection of all objects belonging to the concept, and the connotation is a set of attributes shared by all of these objects. Given a formal context can be constructed based on concept lattice, and construct the concept lattice is unique. K-Means algorithm is currently the most widely used Partition-based clustering algorithm. In this paper, the concept of grid combined with the K-Means algorithm, proposed a new clustering method-of K-MeansBCC (K-Means Algorithm Based on Concept, Lattice). The text as an object, generate the feature words in the text as a property concept lattice; extraction of concept lattice concept, using the concept of the text, and the definition of the concept of similarity function; Finally, the K-Means algorithm for clustering. Concepts used to represent text, reducing the number of dimensions of the characteristic words, to improve the performance of the clustering. In addition, K-Means algorithm artificially determine the value of K, randomly selected from the center point of the shortcomings of a density-based solution. The the K-MeansBCC algorithm used in the sea search clustering module, to do with the K-Means algorithm comparing experimental results show that the the K-MeansBCC algorithm has reasonable and effectiveness.

Related Dissertations

  1. Research and Improvement on K-Means Clustering Algorithm,TP311.13
  2. Research on K-means Optimization Clustering Algorithm,TP311.13
  3. Evolutionary Clustering Algorithm and Its Application,TP311.13
  4. Web Usage Mining and the Research of Personalized Recommendation,TP311.13
  5. Web news hot discovery system design and implementation,TP393.09
  6. Library management system of personalized service Design and Implementation,TP311.52
  7. Subway construction project risk evaluation methods and criteria for research,U231.3
  8. Rough concept lattice based multi-attribute decision analysis,O159
  9. The Algorithms of Generating Concept Lattice,O153.1
  10. Criterion of Isomorphism for Trees and the Application of Tree in Concept Lattice and Inverse Matrix,O157.5
  11. Reseraching of Transfer Learning Methods Based on Optimized Ontology,TP391.1
  12. Concept Lattice Construction Algorithm and Application Based on the Ontology,TP391.1
  13. Research on Cluster Analysis for Spatial Data Mining,TP311.13
  14. Research on Attribute Reduction and Constructing Algorithm of Concept Lattice,O159
  15. Application of Concept Lattices in Distributed Power Networks Fault Diagnosis,TM711
  16. Attribute reduction of concept lattices,TP18
  17. Based on rough set concept lattice reduction and construction,TP18
  18. Study on Mining Maximal Frequent Itemset Based on Iceberg Concept Lattice,TP311.13
  19. Moving grid security policy storage mechanism study,TN929.5
  20. The Research of Concept Lattice Pruning Method and Its Application in Web Mining,TP311.13
  21. The Design and Implementation of Semantic Retrieval Prototype System Based-on Ontology,TP391.3

CLC: > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory
© 2012 www.DissertationTopic.Net  Mobile