Articolul precedent |
Articolul urmator |
184 5 |
Ultima descărcare din IBN: 2023-09-26 03:59 |
SM ISO690:2012 SOKOLOVA, Marina, BOBICEV, Victoria. Learning Relationship between Authors' Activity and Sentiments: A case study of online medical forums. In: Recent Advances in Natural Language Processing: RANLP, Ed. 10, 7-9 septembrie 2015, Hissar. Stroudsburg PA: Association for Computational Linguistics (ACL), 2015, Ediția 10, pp. 604-610. ISSN 13138502. |
EXPORT metadate: Google Scholar Crossref CERIF DataCite Dublin Core |
Recent Advances in Natural Language Processing Ediția 10, 2015 |
||||||
Conferința "10th International Conference on Recent Advances in Natural Language Processing" 10, Hissar, Bulgaria, 7-9 septembrie 2015 | ||||||
|
||||||
Pag. 604-610 | ||||||
|
||||||
Descarcă PDF | ||||||
Rezumat | ||||||
Our current work analyses relations betweensentiments and activity of authors of online In- Vitro Fertilization forums. We focus on twotypes of active authors: those who start new discussions and those who post significantlymore messages than other authors. By incorporating authors' activity information into adomain-specific lexical representation of messages, we were able to improve multi-classclassification of sentiments by 9% for Support Vector Machines and by 15.3 % for ConditionalRandom Fields. |
||||||
Cuvinte-cheie e-learning, Support vector machines |
||||||
|
Dublin Core Export
<?xml version='1.0' encoding='utf-8'?> <oai_dc:dc xmlns:dc='http://purl.org/dc/elements/1.1/' xmlns:oai_dc='http://www.openarchives.org/OAI/2.0/oai_dc/' xmlns:xsi='http://www.w3.org/2001/XMLSchema-instance' xsi:schemaLocation='http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd'> <dc:creator>Sokolova, M.</dc:creator> <dc:creator>Bobicev, V.</dc:creator> <dc:date>2015</dc:date> <dc:description xml:lang='en'><p>Our current work analyses relations betweensentiments and activity of authors of online In- Vitro Fertilization forums. We focus on twotypes of active authors: those who start new discussions and those who post significantlymore messages than other authors. By incorporating authors' activity information into adomain-specific lexical representation of messages, we were able to improve multi-classclassification of sentiments by 9% for Support Vector Machines and by 15.3 % for ConditionalRandom Fields.</p></dc:description> <dc:source>Recent Advances in Natural Language Processing (Ediția 10) 604-610</dc:source> <dc:subject>e-learning</dc:subject> <dc:subject>Support vector machines</dc:subject> <dc:title>Learning Relationship between Authors' Activity and Sentiments: A case study of online medical forums</dc:title> <dc:type>info:eu-repo/semantics/article</dc:type> </oai_dc:dc>