DIAS, MARCELO TIBAU DE VASCONCELLOS. Understanding Web Search Patterns Through Exploratory Search as a Knowledge-Intensive Process. 18/01/2019. Instituição de Ensino: UNIVERSIDADE FEDERAL DO ESTADO DO RIO DE JANEIRO, Rio de Janeiro.
Dissertação Marcelo Tibau de Vasconcellos Dias.pdf
Dissertação de Mestrado
Understanding Web Search Patterns Through Exploratory Search as a Knowledge-Intensive Process
Author
Marcelo Tibau de Vasconcellos Dias (UNIRIO)
Abstract
As data grows exponentially and the Web encompasses most part of the knowledge Human Beings create the seeking for intelligent information systems capable of going beyond keyword searches increases. To better understands users’ intent, Web search engines need to transcend its information sorter utility and acquire a more relevant ability concerning semantics’ discernment. By aiding individuals to retrieve relevant information through formulation and modification of their queries, Web search engines gain importance as tools capable to associate informal and self-learning to formal learning. In a new context in which Web search engines can serve as a means to obtain content that supports learning, there is the demand that their performance involve a capacity for understanding and processing information that is more complex. Modeling them for searches with a more exploratory pattern, in which the available information needs to be explored in more detail, is a valid path to follow. Not being the scope of current Web search engines, exploratory search presents opportunities to clarify the best practices associated to users’ decision-making process regarding suitable and not suitable information and to use these behaviors to improve information system’s understanding about user’s search purpose and to enable the visualization of Web search pattern and learning process used in the search itself. This thesis presents the Exploratory Search KiP model, which helps clarify the reasons why a subject is searched and supports the visualization of how the information retrieved is used to define decision criteria about which data is worth extracting, to draw inferences, and to create a shortcut to understanding. It also introduces the ESKiP Taxonomy of Query States; a classification framework validated in an experiment in two different query logs datasets and that helps to represent the structure and behavior of query reformulation in search systems. A representation that can be used to Web search engines’ setups that organize and provide search results that are more accurate.
Keywords:
Information Systems, Exploratory Search, Knowledge-intensive Process, Query States, Query Reformulation.
Dissertação Marcelo Tibau de Vasconcellos Dias.pdf
