ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-09212016-155533


Tipo di tesi
Tesi di laurea magistrale
Autore
TRANI, ROBERTO
URN
etd-09212016-155533
Titolo
Come combinare Learning To Rank e Query Expansion per migliorare le performance dei motori di ricerca web
Dipartimento
INFORMATICA
Corso di studi
INFORMATICA
Relatori
relatore Prof. Venturini, Rossano
relatore Ing. Nardini, Franco Maria
Parole chiave
  • web search engine
  • query expansion
  • learning to rank
Data inizio appello
07/10/2016
Consultabilità
Completa
Riassunto
Hundreds of millions of users each day look for the information they need on the Web using search engines trying to find out the relevant results from billions of indexed documents.
Queries submitted by web search engines users are short and intrinsically ambiguous since the terms that compose the queries can be inaccurate and general. As a consequence of that the relevant documents for the query could not be returned to users. The trend of the last years is to deal with this problem adding terms (synonyms, disambiguation, etc.) to the original query
trying to cover the user’s true intent.
This technique of adding terms to the original query isknown in literature as Query Expansion (QE) which is an effective method to improve the retrieval performance of the search engine.
The aim of this thesis is to study the theoretical and practical impact of QE in a query processor of a real-world Web search engine. We study how the expanded query can be handled within the search process and we propose an experimental framework enabling QE to test the effectiveness and the efficiency of a generic given expansion strategy. We perform this study in collaboration with the well-known Italian Search Engine (istella). The framework above has been successfully employed within the istella query processor. The results of the activity show important improvements of the effectiveness of the istella search engine thus confirming the validity of the approach taken.
File