Making social media analysis more efficient through taxonomy supported concept suggestion
| dc.contributor.author | Cardoso Coutinho, Fabio | |
| dc.contributor.author | Lang, Alexander | |
| dc.contributor.author | Mitschang, Bernhard | |
| dc.contributor.editor | Markl, Volker | |
| dc.contributor.editor | Saake, Gunter | |
| dc.contributor.editor | Sattler, Kai-Uwe | |
| dc.contributor.editor | Hackenbroich, Gregor | |
| dc.contributor.editor | Mitschang, Bernhard | |
| dc.contributor.editor | Härder, Theo | |
| dc.contributor.editor | Köppen, Veit | |
| dc.date.accessioned | 2018-10-24T09:56:24Z | |
| dc.date.available | 2018-10-24T09:56:24Z | |
| dc.date.issued | 2013 | |
| dc.description.abstract | Social Media sites provide consumers the ability to publicly create and shape the opinion about products, services and brands. Hence, timely understanding of content created in social media has become a priority for marketing departments, leading to the appearance of social media analysis applications. This article describes an approach to help users of IBM Cognos Consumer Insight, IBM's social media analysis offering, define and refine the analysis space more efficiently. It defines a Concept Suggestion Component (CSC) that suggests relevant as well as off-topic concepts within social media, and tying these concepts to taxonomies typically found in marketing around brands, products and campaigns. The CSC employs data mining techniques such as term extraction and clustering, and combines them with a sampling approach to ensure rapid and high-quality feedback. Initial evaluations presented in this article show that these goals can be accomplished for real-life data sets, simplifying the definition of the analysis space for a more comprehensive and focused analysis. | en |
| dc.identifier.isbn | 978-3-88579-608-4 | |
| dc.identifier.pissn | 1617-5468 | |
| dc.identifier.uri | https://dl.gi.de/handle/20.500.12116/17337 | |
| dc.language.iso | en | |
| dc.publisher | Gesellschaft für Informatik e.V. | |
| dc.relation.ispartof | Datenbanksysteme für Business, Technologie und Web (BTW) 2040 | |
| dc.relation.ispartofseries | Lecture Notes in Informatics (LNI) - Proceedings, Volume P-214 | |
| dc.title | Making social media analysis more efficient through taxonomy supported concept suggestion | en |
| dc.type | Text/Conference Paper | |
| gi.citation.endPage | 476 | |
| gi.citation.publisherPlace | Bonn | |
| gi.citation.startPage | 457 | |
| gi.conference.date | 13.-15. März 2013 | |
| gi.conference.location | Magdeburg | |
| gi.conference.sessiontitle | Regular Research Papers |
Dateien
Originalbündel
1 - 1 von 1