상세정보

Search engines have become the necessities of life, today. Especially, a great part of searches are informational types. The purpose of informational search is getting useful information what user wants to know about topics such as movie, restaurant or even bio-medical. Due to the traditional structure of search engine, it has a limitation to retrieve rich information as results. Typical search system returns a list of documents. It’s very efficient for navigational or transnational search, but not for informational. In
case of an informational search, users have to go through the contents of documents to get information.
In order to address this problem, we introduce a new entity search framework named Consento designed to answer informational queries. This framework performs subdocument level indexing, as opposed to document indexing, to return semantic enhanced results. In particular, we define a new indexing unit, Maximal Coherent Semantic Unit (MCSU). An MCSU represents a segment of a document, which captures a single coherent
semantic. We also introduce a new ranking method, called ConsensusRank that counts captured semantics referring to an entity as a weighted vote. In order to validate the
efficacy of the proposed framework, we compare with standard retrieval models and their recent extensions for entity ranking. Experiments using bio medical, movie, hotel and restaurant data show the effectiveness of our framework.

Abstract
Contents i
List of Figures iii
List of Tables iv
1 Introduction 1
2 Related Work 7
3 Consento architecture 10
3.1 Indexing Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.1 Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.2 Content Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.3 Indexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 Search Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.1 Query Parsing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.2 Retrieval and Ranking . . . . . . . . . . . . . . . . . . . . . . . . . 18
4 Validation 22
4.1 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.2 Comparison with Baselines . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5 Smith Search: Consento based Restaurant Search System 31
5.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.2 Smith Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3 Demonstration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.4 System Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6 Conclusion 38
Bibliography 40
Acknowledgement
ii