Logo des Repositoriums
 
Konferenzbeitrag

Improvement of automated social media sentiment analysis methods - a context-based approach

Lade...
Vorschaubild

Volltext URI

Dokumententyp

Text/Conference Paper

Zusatzinformation

Datum

2019

Zeitschriftentitel

ISSN der Zeitschrift

Bandtitel

Verlag

Gesellschaft für Informatik e.V.

Zusammenfassung

The sentiment analysis of social media data increasingly gains importance in business and research. But still, topical algorithms cope with problems, since it is reasonably manageable to extract the tonality of a social media post, but not the authors attitude towards a given topic. However, in most cases, this is the relevant information users of social media analysis tools are looking for. To tackle this problem, we propose a context-based algorithm that not only focuses on isolated postings, but also takes the authorsŠ earlier postings and their interactions with other usersŠ posts into account to derive their actual opinion on a subject. To evaluate this approach, we implemented a test system and compared the algorithmŠs results to manually assessed sentiments.

Beschreibung

Debeyem Dennis; Eder, Tim; Guigas, Paul Vincent; Schuberth, Viktoria (2019): Improvement of automated social media sentiment analysis methods - a context-based approach. SKILL 2019 - Studierendenkonferenz Informatik. Bonn: Gesellschaft für Informatik e.V.. PISSN: 1614-3213. ISBN: 978-3-88579-449-3. pp. 33-44. Natural Language Processing. Kassel. 25.-26. September 2019

Zitierform

DOI

Tags