• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 2
  • Tagged with
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Avalia??o qualitativa da Consulta Espa?o- Textual Top-k

Naziazeno Neto, Luiz Felipe 04 August 2017 (has links)
Submitted by Jadson Francisco de Jesus SILVA (jadson@uefs.br) on 2018-04-11T23:46:51Z No. of bitstreams: 1 Dissertacao_LuizFelipeNaziazeno.pdf: 1544725 bytes, checksum: 715a0d41004634943067cd177b927c35 (MD5) / Made available in DSpace on 2018-04-11T23:46:51Z (GMT). No. of bitstreams: 1 Dissertacao_LuizFelipeNaziazeno.pdf: 1544725 bytes, checksum: 715a0d41004634943067cd177b927c35 (MD5) Previous issue date: 2017-08-04 / It is increasingly common to find large databases with textual and spatial (latitude and longitude) information. Therefore, there is great interest in new strategies to efficiently retrieve relevant information from these datasets. One solution, which has attracted the attention of the research community, is the use of the Top-k Spatial- Textual Query. This query returns the best k spatio-textual documents considering the spatial distance between the query location and the documents and the textual relevance between the query keywords and the text associated with the documents. This work, however, studies the Top-k Spatial-Textual Query from a qualitative perspective, proposing a methodology for creating test collections using metrics to evaluate the quality of the results of the two existing ranking functions, from a textual and spatial perspective. The generated collections have a great diversity of spatial-textual data, which is vital for a qualitative study. The gained knowledge in this research provides an important source for those interested in evaluating information retrievals systems with spatio-textual characteristics. / ? cada vez mais comum encontrar grandes bancos de dados com informa??es textual e espacial (latitude e longitude). Portanto, existe um grande interesse por novas estrat?gias para recuperar eficientemente informa??es relevantes a partir destes conjuntos de dados. Uma solu??o, que tem atra?do a aten??o da comunidade cient?fica, ? a utiliza??o da Consulta Espa?o Textual Top-k. Esta consulta retorna os k melhores documentos espa?o textuais considerando a dist?ncia espacial entre o local da consulta e os documentos e a relev?ncia textual entre as palavras-chave da consulta e o texto associado aos documentos. As pesquisas anteriores concentram-se nos aspectos de performance dessa consulta, como a redu??o do tempo de resposta e I/O. Este trabalho, no entanto, estuda a Consulta Espa?o Textual Top-k a partir de uma perspectiva qualitativa, propondo uma metodologia para cria??o de cole??es de teste espa?o textuais, utilizando m?tricas para avaliar a qualidade dos resultados das duas fun??es de ranqueamento existente, sob a perspectiva textual e espacial. As cole??es geradas proporcionaram uma grande diversidade de dados espa?o textuais, o que ? vital para uma an?lise qualitativa. O conhecimento adquirido, nesta pesquisa, proporciona uma fonte importante aos interessados em avaliar sistemas de recupera??o com caracter?sticas espa?o textuais.
2

Processamento distribu?do da consulta espa?o textual top-k

Novaes, Tiago Fernandes de Athayde 17 July 2017 (has links)
Submitted by Ricardo Cedraz Duque Moliterno (ricardo.moliterno@uefs.br) on 2017-11-28T21:38:06Z No. of bitstreams: 1 dissertacao-versao-final.pdf: 2717503 bytes, checksum: a1476bba65482b40daa1a139191ea912 (MD5) / Made available in DSpace on 2017-11-28T21:38:06Z (GMT). No. of bitstreams: 1 dissertacao-versao-final.pdf: 2717503 bytes, checksum: a1476bba65482b40daa1a139191ea912 (MD5) Previous issue date: 2017-07-17 / With the popularization of databases containing objects with spatial and textual information (spatio-textual object), the interest in new queries and techniques for retrieving these objects have increased. In this scenario, the main query is the the top-k spatio-textual query. This query retrieves the k best spatio-textual objects considering the distance of the object to the query location and the textual similarity between the query keywords and the textual information of the objects. However, most the studies related to top-k spatio-textual query are performed in centralized environments, not addressing real world problems such as scalability. In this paper, we study different strategies for partitioning the data and processing the top-k spatio-textual query in a distributed environment. We evaluate each strategy in a real distributed environment, employing real datasets. / Com a populariza??o de bases de dados contendo objetos que possuem informa??o espacial e textual (objeto espa?o-textual), aumentou o interesse por novas consultas e t?cnicas capazes de recuperar esses objetos de forma eficiente. Uma das principais consultas para objetos espa?o-textuais ? a consulta espa?o-textual top-k. Essa consulta visa recuperar os k melhores objetos considerando a dist?ncia do objeto at? um local informado na consulta e a similaridade textual entre palavras-chave de busca e a informa??o textual dos objetos. No entanto, a maioria dos estudos para consultas espa?o-textual top-k assumem ambientes centralizados, n?o abordando problemas frequentes em aplica??es do mundo real como escalabilidade. Nesta disserta??o s?o estudadas diferentes formas de particionar os dados e o impacto destes particionamentos no processamento da consulta espa?o-textual top-k em um ambiente distribu?do. Todas as estrat?gias propostas s?o avaliadas em um ambiente distribu?do real, utilizando dados reais.

Page generated in 0.1039 seconds