Abstract

Perception of social cues is a fundamental communicative skill that can be hampered by hearing and cognitive disorders. Understanding slang and sarcastic intent is often difficult in verbal communication, particularly for individuals who struggle with the perception of social cues. Misinterpretation of slang terms can cause discomfort or social isolation. Sarcasm is particularly difficult to recognize due to its inherently ambiguous and context-dependent nature. We have identified two problems of particular interest in social assistive technologies – slang word sentiment assessment and sarcasm detection. We propose combining a slang sentiment analysis model with a speech emotion analysis model to create an assistive tool, Chat-Box, which will detect social cues such as sarcasm, slang, and sentiment.

Original languageEnglish (US)
Title of host publicationHCI International 2018 – Posters’ Extended Abstracts - 20th International Conference, HCI International 2018, Proceedings
EditorsConstantine Stephanidis
PublisherSpringer Verlag
Pages394-401
Number of pages8
ISBN (Print)9783319922782
DOIs
StatePublished - 2018
Event20th International Conference on HCI, HCI International 2018 - Las Vegas, United States
Duration: Jul 15 2018Jul 20 2018

Publication series

NameCommunications in Computer and Information Science
Volume851
ISSN (Print)1865-0929

Conference

Conference20th International Conference on HCI, HCI International 2018
Country/TerritoryUnited States
CityLas Vegas
Period7/15/187/20/18

Keywords

  • Assistive technology
  • Sarcasm
  • Slang
  • Slang2Vector
  • Speech
  • Wire-frame
  • Word2Vec

ASJC Scopus subject areas

  • General Computer Science
  • General Mathematics

Fingerprint

Dive into the research topics of 'Chat-box: Proposing a mood analyzer for individuals with social interaction disabilities'. Together they form a unique fingerprint.

Cite this