Dissertation > Excellent graduate degree dissertation topics show

Research on Web Based Expert Search

Author: DuanHuiZhong
Tutor: YuYong
School: Shanghai Jiaotong University
Course: Applied Computer Technology
Keywords: Expert search Experts identify Search model Experts rank Experts names disambiguation
CLC: TP391.3
Type: Master's thesis
Year: 2009
Downloads: 142
Quote: 1
Read: Download Dissertation

Abstract


The so-called experts, is a certain expertise personnel. In today's era surge in the amount of information within the enterprise network and the World Wide Web search expert, is an increasingly important requirement. However, the existing expert search techniques and tools can not meet this demand. For intranet search experts have been some studies, but there is still not a unified theoretical framework, and retrieve the effect is still unable to meet the actual needs of users, and for experts search the World Wide Web research is currently very few. To compensate for these research gaps, this paper presents a Web-based expert search research. Specifically, this paper mainly studied in depth the following topics. Expert discovery and identification is the automatic discovery documents set all the experts, and the experts appear to identify and mark the location. It is to establish an expert search engine the primary problem. This paper presents an analysis of the E-mail through a method to automatically obtain a list of experts in the documentation set, and propose an efficient recognition expert position in the document focus on the emergence of the algorithm. Expert search model is a list of experts in a given case, for a given query in the document focused and effective for expert search. This paper presents an evidence-based expert search model. In this model, an expert with the query relevance of the evidence by the document focused decision. Experimental results show that the proposed model can effectively search the Web document set retrieval experts. To distinguish between experts and ordinary users, this paper proposes a static ranking (Static Ranking) method to discover the importance of the expert. This paper presents the ranking algorithm based on link analysis expert (ExpertRank), and further develop the theme sensitive expert ranking algorithm (Topic Sensitive ExpertRank). The World Wide Web, names Ambiguity very serious. The search result is often a name corresponding to the number of different individuals. This became an expert search problem must be solved. This paper presents a problem for the use of additional feedback page classification framework and in the framework of the proposed two features for classification: \Experimental results show that these two features to improve the disambiguation effect plays an important role. In addition, the paper also discussed in depth the effect of different classification methods, and the use of different sampling methods handle data imbalance brought about stability issues.

Related Dissertations

  1. Chinese expert retrieval key technology research,TP18
  2. Expert search systems based on enterprise-class corpus,TP391.3
  3. Construction and Visualization of social network - based technology consulting expert database,G312
  4. Research on Enterprise Information Retrieval,TP391.3
  5. Feature search task switching does not affect the efficiency of,B842.1
  6. Inversion Effect for Non-Face Objects,B842.1
  7. Enterprise Information Retrieval object retrieval method,TP391.3
  8. Mining Web Social Networks,TP393.09
  9. Research on Entity Retrieval and Mining with the Web,TP391.3
  10. Research on the Search of Target Firm in Merger and Acquisition of China,F271
  11. A Study of Targets Independence and Modeling of Multiple-Target Random Visual Search,TP391.3
  12. Research on Selecting M&A’s Target Firms Based on Synergistic Effects,F271
  13. Study on Scheduling Flexible Manufacturing Systems Using Petri Nets,TH166
  14. Research on Analyzing Model and Prevention & Emergency System for Power System Blackouts,TM711
  15. The Research of Domain-Oriented Search Engine Model Based on Meta-Search Scheduling,TP391.3
  16. Social Capital on the students' employment impact analysis,F249.2;F224
  17. Research and Implementation of Intelligent Information Retrieval Technology in American Health Care System,TP391.3
  18. Research on Intelligent Search Model Based Semantic Web,TP391.3
  19. The Construction of the Amino Acid Patent Database,O6-39
  20. Fast active tabu search and its application in image classification,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Retrieval machine
© 2012 www.DissertationTopic.Net  Mobile