6533b86efe1ef96bd12ccaf5
RESEARCH PRODUCT
Detecting Emotions in Comments on Forums
Marius CiocaDaniela Gîfusubject
Computer Networks and Communicationsbusiness.industryRomanianmedia_common.quotation_subjectPerspective (graphical)Sentiment analysisEmotion identificationPassionPublic relationslanguage.human_languageField (computer science)Computer Science ApplicationsComputational Theory and MathematicslanguageNatural (music)SociologyArtificial intelligencebusinessConstruct (philosophy)media_commondescription
The paper presents one of the most important issues in Natural Language Processing (NLP), emotion identification and classification to implement a computational technology based on existing resources, open-source or freely available for research purposes. Furthermore, we are interested to use it for establishing Gold standards in sentiment analysis area, such as SentiWordNet. In this sense, we propose to recognize and classify the emotions (sentiments) of the public consumer from the written texts which appeared on the various Forums. We analyse the writing style which refers to how consumers construct sentences together when they write comments to indicate their passion about an entity (persons, brand, location, etc.). We present in this paper a method for integrating Romanian lexical resources from motional perspective, in developing, which can be used in sentiment analysis. This study is intend to help direct beneficiaries (public consumer, marketing managers, PR firms, politicians, investors), but, also, specialists and researchers in the field of natural language processing, linguists, psychologists, sociologists, economists, etc.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2014-12-01 | International Journal of Computers Communications & Control |