Extracting social semantics from multimodal meeting content

Zhiwen Yu, Xingshe Zhou, Yuichi Nakamura

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

The article presents a framework for extracting both low-level (individual) and high-level (group) semantics. Meetings are a human collaboration process encapsulating a large amount of social and communicative information. The key challenges in creating effective collaborations might be social rather than technical, owing to the dynamics of information exchange. Social factors are thus a fundamental aspect of meeting efficacy. Social semantics are particularly important for understanding the social dynamics occurring during a meeting. Once the basic features have been detected, they are used to infer the interaction occasion. There are two types of interaction occasions: spontaneous or reactive. The interaction occasion can be automatically recognized. An interaction flow is a list of all interactions in a discussion session, with triggering relationships between them.

Original languageEnglish
Article number6253196
Pages (from-to)68-75
Number of pages8
JournalIEEE Pervasive Computing
Volume12
Issue number2
DOIs
StatePublished - Apr 2013

Fingerprint

Dive into the research topics of 'Extracting social semantics from multimodal meeting content'. Together they form a unique fingerprint.

Cite this