Mixture of Topic-Based Distributional Semantic and Affective Models

Fenia Christopoulou, Eleftheria Briakou, Elias Iosif, Alexandros Potamianos

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Typically, Distributional Semantic Models (DSMs) estimate semantic similarity between words using a single-model, where the multiple senses of polysemous words are conflated in a single representation. Similarly, in textual affective analysis tasks, ambiguous words are usually not treated differently when estimating word affective scores. In this work, a semantic mixture model is proposed enabling the combination of word similarity scores estimated across multiple topic-specific DSMs (TDSMs). Based on the assumption that semantic similarity implies affective similarity, we extend this model to perform sentence-level affect estimation. The proposed model outperforms the baseline approach achieving state-of-the-art results for semantic similarity estimation and sentence-level polarity detection.

Original languageEnglish
Title of host publicationProceedings - 12th IEEE International Conference on Semantic Computing, ICSC 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages203-210
Number of pages8
ISBN (Electronic)9781538644072
DOIs
Publication statusPublished - 9 Apr 2018
Externally publishedYes
Event12th IEEE International Conference on Semantic Computing, ICSC 2018 - Laguna Hills, United States
Duration: 31 Jan 20182 Feb 2018

Publication series

NameProceedings - 12th IEEE International Conference on Semantic Computing, ICSC 2018
Volume2018-January

Conference

Conference12th IEEE International Conference on Semantic Computing, ICSC 2018
Country/TerritoryUnited States
CityLaguna Hills
Period31/01/182/02/18

Keywords

  • Affective Models
  • Distributional Semantic Models
  • Semantic Mixture Models
  • Semantic Similarity
  • Topic Modeling

Fingerprint

Dive into the research topics of 'Mixture of Topic-Based Distributional Semantic and Affective Models'. Together they form a unique fingerprint.

Cite this