Is it possible to develop a reliable QA-Corpus using social media data? What are the challenges faced when attempting such a task? In this paper, we discuss these questions and present our findings when developing a QA-Corpus on the topic of Brazilian finance. In order to populate our corpus, we relied on opinions from experts on Brazilian finance that are active on the Twitter application. From these experts, we extracted information from news websites that are used as answers in the corpus. Moreover, to effectively provide rankings of answers to questions, we employ novel word vector based similarity measures between short sentences (that accounts for both questions and Tweets). We validated our methods on a recently released dataset of similarity between short Portuguese sentences. Finally, we also discuss the effectiveness of our approach when used to rank answers to questions from real users.
%0 Book Section
%1 Cavalin2016
%A Cavalin, Paulo
%A Figueiredo, Flavio
%A de Bayser, Maíra
%A Moyano, Luis
%A Candello, Heloisa
%A Appel, Ana
%A Souza, Renan
%B Computational Processing of the Portuguese Language: 12th International Conference, PROPOR 2016, Tomar, Portugal, July 13-15, 2016, Proceedings
%C Cham
%D 2016
%E Silva, João
%E Ribeiro, Ricardo
%E Quaresma, Paulo
%E Adami, André
%E Branco, António
%I Springer International Publishing
%K qa
%P 353--358
%R 10.1007/978-3-319-41552-9_36
%T Building a Question-Answering Corpus Using Social Media and News Articles
%U https://doi.org/10.1007/978-3-319-41552-9_36
%X Is it possible to develop a reliable QA-Corpus using social media data? What are the challenges faced when attempting such a task? In this paper, we discuss these questions and present our findings when developing a QA-Corpus on the topic of Brazilian finance. In order to populate our corpus, we relied on opinions from experts on Brazilian finance that are active on the Twitter application. From these experts, we extracted information from news websites that are used as answers in the corpus. Moreover, to effectively provide rankings of answers to questions, we employ novel word vector based similarity measures between short sentences (that accounts for both questions and Tweets). We validated our methods on a recently released dataset of similarity between short Portuguese sentences. Finally, we also discuss the effectiveness of our approach when used to rank answers to questions from real users.
%@ 978-3-319-41552-9
@inbook{Cavalin2016,
abstract = {Is it possible to develop a reliable QA-Corpus using social media data? What are the challenges faced when attempting such a task? In this paper, we discuss these questions and present our findings when developing a QA-Corpus on the topic of Brazilian finance. In order to populate our corpus, we relied on opinions from experts on Brazilian finance that are active on the Twitter application. From these experts, we extracted information from news websites that are used as answers in the corpus. Moreover, to effectively provide rankings of answers to questions, we employ novel word vector based similarity measures between short sentences (that accounts for both questions and Tweets). We validated our methods on a recently released dataset of similarity between short Portuguese sentences. Finally, we also discuss the effectiveness of our approach when used to rank answers to questions from real users.},
added-at = {2018-01-08T20:56:10.000+0100},
address = {Cham},
author = {Cavalin, Paulo and Figueiredo, Flavio and de Bayser, Ma{\'i}ra and Moyano, Luis and Candello, Heloisa and Appel, Ana and Souza, Renan},
biburl = {https://www.bibsonomy.org/bibtex/2c1e5903b08690853a802cdadad0471d5/defeatnelly},
booktitle = {Computational Processing of the Portuguese Language: 12th International Conference, PROPOR 2016, Tomar, Portugal, July 13-15, 2016, Proceedings},
doi = {10.1007/978-3-319-41552-9_36},
editor = {Silva, Jo{\~a}o and Ribeiro, Ricardo and Quaresma, Paulo and Adami, Andr{\'e} and Branco, Ant{\'o}nio},
interhash = {da5b5b4994b73273b99241b2d68b6c5e},
intrahash = {c1e5903b08690853a802cdadad0471d5},
isbn = {978-3-319-41552-9},
keywords = {qa},
pages = {353--358},
publisher = {Springer International Publishing},
timestamp = {2018-01-08T20:56:10.000+0100},
title = {Building a Question-Answering Corpus Using Social Media and News Articles},
url = {https://doi.org/10.1007/978-3-319-41552-9_36},
year = 2016
}