RT Dissertation/Thesis
T1 Exploiting multiple sources of evidence for opinion search in the web
A1 González Chenlo, José Manuel
A2 Universidade de Santiago de Compostela. E.T.S. de Enxeñaría. Facultade de Física. Departamento de Electrónica e Computación, 
K1 sistemas Miner  a de Opiniones y An alisis de Sentimientos
K1 web
AB In this thesis we study Opinion Mining and Sentiment Analysis and proposea  ne-grained analysis of the opinions conveyed in texts. Concretely, the aim ofthis research is to gain an understanding on how to combine di erent types ofevidence to e ectively determine on-topic opinions in texts. To meet this aim,we consider content-match evidence, obtained at document and passage level,as well as di erent structural aspects of the text.Current Opinion Mining technology is not mature yet. As a matter of fact,people often use regular search engines, which lack evolved opinion search ca-pabilities, to  nd opinions about their interests. This means that the e ort ofdetecting what are the key relevant opinions relies on the user. The lack ofwidely accepted Opinion Mining technology is due to the limitations of cur-rent models, which are simplistic and perform poorly. In this thesis we studya speci c set of factors that are indicative of subjectivity and relevance and wetry to understand how to e ectively combine them to detect opinionated docu-ments, to extract relevant opinions and to estimate their polarity. We proposeinnovative methods and models able to incorporate di erent types of evidenceand it is our intention to contribute in di erent areas, including those relatedto i) search for opinionated documents, ii) detection of subjectivity at docu-ment and passage level, and iii) estimation of polarity. An important concernthat guides this research is e ciency. Some types of evidence, such as discoursestructure, have only been tested with small collections from narrow domains(e.g., movie reviews). We demonstrate here that evolved linguistic features {based on discourse analysis{ can potentially lead to a better understanding ofhow subjectivity ows in texts. And we show that this type of features can bee ciently injected into general-purpose opinion retrieval solutions that operateat large scale.
YR 2014
FD 2014-10-07
LK http://hdl.handle.net/10347/11523
UL http://hdl.handle.net/10347/11523
LA eng
DS Minerva
RD 24 abr 2026