В рамках данной работы было проведено исследование существующих типов вопросно-ответных систем и методов анализа текста. Был проведен анализ существующих вопросно-ответных систем. Приведена обобщённая схема работы вопросно-ответных систем и для каждого типа систем приведена детальная схема работы. Описаны и исследованы методы выбора кандидатов ответа и методы их оценки. Также в работе описаны возможные критерии оценки работы таких систем. В ходе исследования был разработан рабочий прототип вопросно-ответной системы, основанный на системе BERT для русского языка. Используемая модель RuBERT была предобучена и протестирована на стандартных задачах SQuAD. В ходе работы модель была протестирована и оценена в разработанном рабочем прототипе и показала высокие результаты по предложенным критериям оценки. / Within the framework of this work, a study was made of the existing types of question-answer systems and text analysis methods. An analysis of the existing question-answer systems was carried out. A generalized scheme of operation of question-answer systems is given, and a detailed scheme of operation is given for each type of system. Methods for selecting response candidates and methods for their evaluation are described and investigated. The paper also describes possible criteria for evaluating the operation of such systems. In the course of the study, a working prototype of a question-answer system based on the BERT system for the Russian language was developed. The RuBERT model used was pre-trained and tested on standard SQuAD problems. During the work, the model was tested and evaluated in the developed working prototype and showed high results according to the proposed evaluation criteria.

Recommendation systems are information filtering systems that help users effectively and efficiently explore large amount of information and identify items of interest. Accurate predictions of users' interests improve user satisfaction and are beneficial to business or service providers. Researchers have been making tremendous efforts to improve the accuracy of recommendations. Emerging trends of technologies and application scenarios, however, lead to challenges other than accuracy for recommendation systems. Three new challenges include: (1) opinion spam results in untrustworthy content and makes recommendations deceptive; (2) users prefer diversified content; (3) in some applications user behavior data may not be available to infer users' preference. This thesis tackles the above challenges. We identify features of untrustworthy commercial campaigns on a question and answer website, and adopt machine learning-based techniques to implement an adaptive detection system which automatically detects commercial campaigns. We incorporate diversity requirements into a classic theoretical model and develop efficient algorithms with performance guarantees. We propose a novel and robust approach to infer user preference profile from recommendations using copula models. The proposed approach can offer in-depth business intelligence for physical stores that depend on Wi-Fi hotspots for mobile advertisement. / Graduate / 0984 / cchenv@uvic.ca

Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-05-07T14:58:00Z No. of bitstreams: 1 MILENA SALES ARAÚJO - DISSERTAÇÃO PPGCC 2015..pdf: 2915607 bytes, checksum: ae3e37b001051a5fd440be9134f6e832 (MD5) / Made available in DSpace on 2018-05-07T14:58:00Z (GMT). No. of bitstreams: 1 MILENA SALES ARAÚJO - DISSERTAÇÃO PPGCC 2015..pdf: 2915607 bytes, checksum: ae3e37b001051a5fd440be9134f6e832 (MD5) Previous issue date: 2015-07-15 / Sites de perguntas e respostas são conhecidos por serem uma fonte de conhecimento simples e confiável. Os sites do StackExchange se destacam por estarem dentre as maiores comunidades de perguntas e respostas online, principalmente se considerarmos aquelas relacionadas à STEM. Assim como em outras comunidades online, mulheres são uma minoria nos sites do Stackexchange, fazendo com que estes sites não se beneficiem de pontos de vista diversificados na criação de seu conteúdo. Esta pesquisa investiga se mulheres que se dedicam aos sites do StackExchange contribuem menos e/ou deixam a comunidade precocemente. Para isso, examinamos diferenças entre os gêneros quanto ao número de contribuições feitas aos sites, o tempo que passaram contribuindo e quanto a diferença de qualidade. das suas contribuições segundo a comunidade. Nossos resultados mostram que, na maioria dos sites, mulheres contribuem tanto quanto homens, sendo suas contribuições com qualidade e por intervalos de tempo similares aos homens. Quando percebe-se diferença no número de contribuições, são as mulheres que tendem a contribuir mais. Além disso, encontramos que a proporção de contribuições provenientes de mulheres nos sites relacionados a STEM do StackExchange está aumentando na maioria das comunidades onde encontramos diferença estatisticamente significativa. Estes resultados ajudam a melhorar a visão que temos das contribuições feitas por mulheres neste tipo de site, gerando repercussão para pesquisas futuras, assim como discussão de design para os administradores. / Social Question and answer (Q&A) sites are known as simple and reliable knowledge databases. SitesfromtheStackExchangeplatformpresentlystandoutassomeofthelargest and most social Q&A spaces, particularly the sites related to STEM. As in some other collaborative online communities, women are underrepresented in StackExchange platform, hindering its sites from contemplating more diverse views in their content creation process. This work investigates whether women who are engaged with sites from StackExchange are influenced to contribute less and to leave the community. For that, we examined gender differencces in the number of contributions made to the sites, the time spent contributing, and on the evaluation of quality that the community provides for posted content. Our results point that, in the greater part of sites, women contributors contribute as much as men, with similar quality and for a similar period. When difference do occur, it happens most often that women tend to contribute more than men. Nevertheless, we found that the proportion of contributions coming from women in the STEM sites of StackExchange is increasing in several sites. These results point to aricher picture of female contribution in those sites, and raise several implications for future research, as well as to site administrators.