Incorporating product description to sentiment topic models for improved aspect-based sentiment analysis

Reinald Kim Amplayo, Seanie Lee, Min Song

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Sentiment topic models are used as unsupervised methods to solve the specific problems of the general aspect-based sentiment analysis (ABSA) problem. One of the main problems of the technique is its substandard aspect term extraction, which leads to difficulties in aspect label determination. This paper is focused on improving the aspect term extraction of topic models by incorporating product descriptions to the current state-of-the-art sentiment topic model, Aspect Sentiment Unification Model (ASUM). We present two models that extend from ASUM differently to leverage on the information found in the product description: Seller-aided Aspect-based Sentiment Model (SA-ASM) and Seller-aided Product-based Sentiment Model (SA-PSM). SA-ASM has its topic distribution inside the review while SA-PSM has its topic distribution inside the product description. Based on experiments conducted to reviews of laptops and mobile phones, results show that SA-ASM performs better in micro-level problems such as sentiment classification and aspect assignment and SA-PSM performs better in macro-level problems like aspect category detection. Both models achieve better performances compared to current topic modeling methods for the ABSA problem.

Original languageEnglish
Pages (from-to)200-215
Number of pages16
JournalInformation Sciences
Volume454-455
DOIs
Publication statusPublished - 2018 Jul 1

Fingerprint

Sentiment Analysis
Model
Unification
Sentiment
Topic model
Sentiment analysis
Seller
Term
Mobile Phone
Modeling Method
Mobile phones
Leverage
Byproducts
Macros
Labels

All Science Journal Classification (ASJC) codes

  • Software
  • Control and Systems Engineering
  • Theoretical Computer Science
  • Computer Science Applications
  • Information Systems and Management
  • Artificial Intelligence

Cite this

@article{88e1bb7868ff40e1818d70a11e801029,
title = "Incorporating product description to sentiment topic models for improved aspect-based sentiment analysis",
abstract = "Sentiment topic models are used as unsupervised methods to solve the specific problems of the general aspect-based sentiment analysis (ABSA) problem. One of the main problems of the technique is its substandard aspect term extraction, which leads to difficulties in aspect label determination. This paper is focused on improving the aspect term extraction of topic models by incorporating product descriptions to the current state-of-the-art sentiment topic model, Aspect Sentiment Unification Model (ASUM). We present two models that extend from ASUM differently to leverage on the information found in the product description: Seller-aided Aspect-based Sentiment Model (SA-ASM) and Seller-aided Product-based Sentiment Model (SA-PSM). SA-ASM has its topic distribution inside the review while SA-PSM has its topic distribution inside the product description. Based on experiments conducted to reviews of laptops and mobile phones, results show that SA-ASM performs better in micro-level problems such as sentiment classification and aspect assignment and SA-PSM performs better in macro-level problems like aspect category detection. Both models achieve better performances compared to current topic modeling methods for the ABSA problem.",
author = "Amplayo, {Reinald Kim} and Seanie Lee and Min Song",
year = "2018",
month = "7",
day = "1",
doi = "10.1016/j.ins.2018.04.079",
language = "English",
volume = "454-455",
pages = "200--215",
journal = "Information Sciences",
issn = "0020-0255",
publisher = "Elsevier Inc.",

}

Incorporating product description to sentiment topic models for improved aspect-based sentiment analysis. / Amplayo, Reinald Kim; Lee, Seanie; Song, Min.

In: Information Sciences, Vol. 454-455, 01.07.2018, p. 200-215.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Incorporating product description to sentiment topic models for improved aspect-based sentiment analysis

AU - Amplayo, Reinald Kim

AU - Lee, Seanie

AU - Song, Min

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Sentiment topic models are used as unsupervised methods to solve the specific problems of the general aspect-based sentiment analysis (ABSA) problem. One of the main problems of the technique is its substandard aspect term extraction, which leads to difficulties in aspect label determination. This paper is focused on improving the aspect term extraction of topic models by incorporating product descriptions to the current state-of-the-art sentiment topic model, Aspect Sentiment Unification Model (ASUM). We present two models that extend from ASUM differently to leverage on the information found in the product description: Seller-aided Aspect-based Sentiment Model (SA-ASM) and Seller-aided Product-based Sentiment Model (SA-PSM). SA-ASM has its topic distribution inside the review while SA-PSM has its topic distribution inside the product description. Based on experiments conducted to reviews of laptops and mobile phones, results show that SA-ASM performs better in micro-level problems such as sentiment classification and aspect assignment and SA-PSM performs better in macro-level problems like aspect category detection. Both models achieve better performances compared to current topic modeling methods for the ABSA problem.

AB - Sentiment topic models are used as unsupervised methods to solve the specific problems of the general aspect-based sentiment analysis (ABSA) problem. One of the main problems of the technique is its substandard aspect term extraction, which leads to difficulties in aspect label determination. This paper is focused on improving the aspect term extraction of topic models by incorporating product descriptions to the current state-of-the-art sentiment topic model, Aspect Sentiment Unification Model (ASUM). We present two models that extend from ASUM differently to leverage on the information found in the product description: Seller-aided Aspect-based Sentiment Model (SA-ASM) and Seller-aided Product-based Sentiment Model (SA-PSM). SA-ASM has its topic distribution inside the review while SA-PSM has its topic distribution inside the product description. Based on experiments conducted to reviews of laptops and mobile phones, results show that SA-ASM performs better in micro-level problems such as sentiment classification and aspect assignment and SA-PSM performs better in macro-level problems like aspect category detection. Both models achieve better performances compared to current topic modeling methods for the ABSA problem.

UR - http://www.scopus.com/inward/record.url?scp=85046696205&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046696205&partnerID=8YFLogxK

U2 - 10.1016/j.ins.2018.04.079

DO - 10.1016/j.ins.2018.04.079

M3 - Article

VL - 454-455

SP - 200

EP - 215

JO - Information Sciences

JF - Information Sciences

SN - 0020-0255

ER -