Feature network-driven quadrant mapping for summarizing customer reviews

Su Gon Cho, Seoung Bum Kim

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

With the rapid growth of e-commerce, customers increasingly write online reviews of the product they purchase. These customer reviews are one of the most valuable sources of information affecting selection of products or services. Summarizing these customer reviews is becoming an interesting area of research, inspiring researchers to develop a more condensed, concise summarization for users. However, most of the current efforts at summarization are based on general product features without feature’s relationship. As a result, these summaries either ignore feedback from customers or do a poor job of reflecting the opinions expressed in customer reviews. To remedy this summarization shortcoming, we propose a feature network-driven quadrant mapping that captures and incorporates opinions from customer reviews. Our focus is on construction of a feature network, which is based on co-occurrence and sematic similarities, and a quadrant display showing the opinions polarity of feature groups. Moreover, the proposed approach involves clustering similar product features, and thus, it is different from standard text summarization based on abstraction and extraction. The summarized results can help customers better understand the overall opinions about a product.

Original languageEnglish
Pages (from-to)1-19
Number of pages19
JournalJournal of Systems Science and Systems Engineering
DOIs
Publication statusAccepted/In press - 2017 Jan 13

Fingerprint

Display devices
Feedback

Keywords

  • Customer review
  • feature network
  • text mining
  • text summarization
  • visualization

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Information Systems

Cite this

Feature network-driven quadrant mapping for summarizing customer reviews. / Cho, Su Gon; Kim, Seoung Bum.

In: Journal of Systems Science and Systems Engineering, 13.01.2017, p. 1-19.

Research output: Contribution to journalArticle

@article{d664fbf62b064c83960bcf50b0a9163c,
title = "Feature network-driven quadrant mapping for summarizing customer reviews",
abstract = "With the rapid growth of e-commerce, customers increasingly write online reviews of the product they purchase. These customer reviews are one of the most valuable sources of information affecting selection of products or services. Summarizing these customer reviews is becoming an interesting area of research, inspiring researchers to develop a more condensed, concise summarization for users. However, most of the current efforts at summarization are based on general product features without feature’s relationship. As a result, these summaries either ignore feedback from customers or do a poor job of reflecting the opinions expressed in customer reviews. To remedy this summarization shortcoming, we propose a feature network-driven quadrant mapping that captures and incorporates opinions from customer reviews. Our focus is on construction of a feature network, which is based on co-occurrence and sematic similarities, and a quadrant display showing the opinions polarity of feature groups. Moreover, the proposed approach involves clustering similar product features, and thus, it is different from standard text summarization based on abstraction and extraction. The summarized results can help customers better understand the overall opinions about a product.",
keywords = "Customer review, feature network, text mining, text summarization, visualization",
author = "Cho, {Su Gon} and Kim, {Seoung Bum}",
year = "2017",
month = "1",
day = "13",
doi = "10.1007/s11518-017-5329-5",
language = "English",
pages = "1--19",
journal = "Journal of Systems Science and Systems Engineering",
issn = "1004-3756",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Feature network-driven quadrant mapping for summarizing customer reviews

AU - Cho, Su Gon

AU - Kim, Seoung Bum

PY - 2017/1/13

Y1 - 2017/1/13

N2 - With the rapid growth of e-commerce, customers increasingly write online reviews of the product they purchase. These customer reviews are one of the most valuable sources of information affecting selection of products or services. Summarizing these customer reviews is becoming an interesting area of research, inspiring researchers to develop a more condensed, concise summarization for users. However, most of the current efforts at summarization are based on general product features without feature’s relationship. As a result, these summaries either ignore feedback from customers or do a poor job of reflecting the opinions expressed in customer reviews. To remedy this summarization shortcoming, we propose a feature network-driven quadrant mapping that captures and incorporates opinions from customer reviews. Our focus is on construction of a feature network, which is based on co-occurrence and sematic similarities, and a quadrant display showing the opinions polarity of feature groups. Moreover, the proposed approach involves clustering similar product features, and thus, it is different from standard text summarization based on abstraction and extraction. The summarized results can help customers better understand the overall opinions about a product.

AB - With the rapid growth of e-commerce, customers increasingly write online reviews of the product they purchase. These customer reviews are one of the most valuable sources of information affecting selection of products or services. Summarizing these customer reviews is becoming an interesting area of research, inspiring researchers to develop a more condensed, concise summarization for users. However, most of the current efforts at summarization are based on general product features without feature’s relationship. As a result, these summaries either ignore feedback from customers or do a poor job of reflecting the opinions expressed in customer reviews. To remedy this summarization shortcoming, we propose a feature network-driven quadrant mapping that captures and incorporates opinions from customer reviews. Our focus is on construction of a feature network, which is based on co-occurrence and sematic similarities, and a quadrant display showing the opinions polarity of feature groups. Moreover, the proposed approach involves clustering similar product features, and thus, it is different from standard text summarization based on abstraction and extraction. The summarized results can help customers better understand the overall opinions about a product.

KW - Customer review

KW - feature network

KW - text mining

KW - text summarization

KW - visualization

UR - http://www.scopus.com/inward/record.url?scp=85009253080&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85009253080&partnerID=8YFLogxK

U2 - 10.1007/s11518-017-5329-5

DO - 10.1007/s11518-017-5329-5

M3 - Article

AN - SCOPUS:85009253080

SP - 1

EP - 19

JO - Journal of Systems Science and Systems Engineering

JF - Journal of Systems Science and Systems Engineering

SN - 1004-3756

ER -