Supervised Identification of the User Intent of Web Search Queries
As the Web continues to increase both in size and complexity, Web search is a ubiquitous service that allows users to find all kind of information, resources, and activities. However, as the Web evolves so do the needs of the users. Nowadays, users have more complex interests that go beyond of the t...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Estado: | Versión publicada |
| Fecha de publicación: | 2011 |
| País: | España |
| Institución: | CBUC, CESCA |
| Repositorio: | TDR. Tesis Doctorales en Red |
| OAI Identifier: | oai:www.tdx.cat:10803/51300 |
| Acceso en línea: | http://hdl.handle.net/10803/51300 |
| Access Level: | acceso abierto |
| Palabra clave: | User intent Web queries Web searching Search engines Intención del usuario consultas Web búsqueda Web motores de búsqueda 62 |
| Sumario: | As the Web continues to increase both in size and complexity, Web search is a ubiquitous service that allows users to find all kind of information, resources, and activities. However, as the Web evolves so do the needs of the users. Nowadays, users have more complex interests that go beyond of the traditional informational queries. Thus, it is important for Web-search engines, not only to continue answering effectively informational and navigational queries, but also to be able to identify and provide accurate results for new types of queries. This Ph.D. thesis aims to analyze the impact of the query intent in the search behavior of the users. In order to achieve this, we first study the behavior of users with different types of query intent on search engine result pages (SERP), using eye tracking techniques. Our study shows that the query intent of the user affects all the decision process in the SERP. Users with different query intent prefer different type of search results (organic, sponsored), they attend to different main areas of interest (title, snippet, URL, image) and focus on search results with different ranking position. To be able to accurately identify the intent of the user query is an important issue for search engines, as this will provide useful elements that allow them adapting their results to changing user behaviors and needs. Therefore, in this thesis we propose a method to identify automatically the intent behind user queries. Our hypothesis is that the performance of single-faceted classification of queries can be improved by introducing information of multi-faceted training samples into the learning process. Hence, we study a wide set of facets that can be considered for the characterization of the query intent of the user and we investigate whether combining multiple facets can improve the predictability of these facets. Our experimental results show that this idea can significantly improve the quality of the classification. Since most of previous works in query intent classification are oriented to the study of single facets, these results are a first step to an integrated query intent classification model. |
|---|