Lexicon-based sentiment analysis approach for ranking event entities

Chamass, Sajida (Faculty of Sciences, Lebanese University, Beirut, Lebanon) ; Hazimeh, Hussein (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland) ; Makki, Jawad (Faculty of Sciences Lebanese University, Beirut, Lebanon) ; Mugellini, Elena (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland) ; Abou Khaled, Omar (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland)

Social media platforms (SMP) are new resource for data analytics. Multiple aspects can be studied by using its variety of features. Sentiment analysis (SA) is a rising research topic in SMPs. SA approaches on studying and analysing events are still missing several shortcomings. In this paper, we address the problem of ranking event entities and propose a novel approach for this goal. An entity is a person who presents some task in such event, for e.g., a researcher in a conference. To achieve our target, we employ the lexical approach, in addition to associating features from both Facebook and Twitter platforms. We used Facebook reactions also, that not been used in the state-of-the-art approaches. Our results have shown that by associating both features from Facebook and Twitter and by using reactions, we can successfully rank entities participating in a specific event having high precision.


Keywords:
Article Type:
scientifique
Faculty:
Ingénierie et Architecture
School:
HEIA-FR
Institute:
HumanTech - Technology for Human Wellbeing Institute
Subject(s):
Arts visuels
Ingénierie
Date:
2018-05
Pagination:
14 p.
Published in:
International Journal of Services and Standards
Numeration (vol. no.):
2018, vol. 12, no. 2, pp. 126-139
DOI:
ISSN:
1740-8849
Appears in Collection:

Note: The status of this file is: public


 Record created 2019-01-22, last modified 2019-03-26

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)