Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering

Research output: Contribution in Book/Catalog/Report/Conference proceedingConference contribution

Abstract

Online reviews are an important source of information that customers use to make more informed purchase decisions. Attribute-centric reviews, in which the author supports her opinion with comments on the technical attributes of the product, are particularly insightful because they present deeper discussions about how technical specifications can meet the expectations of customers. However, as the number of available reviews grows, it becomes increasingly cumbersome to manually locate attribute-centric reviews as they get lost within a flood of less informative reviews. We propose a word clustering approach that uses the technical specifications of products to identify technical discussions in online reviews. Each output cluster represents a technical aspect of the products and can be used to extract its related attribute-centric reviews. We evaluate our approach by modeling technical aspects for 21,846 reviews for cameras and show that our approach can extract and rank relevant technical comments.

Original languageEnglish
Title of host publicationNatural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings
EditorsFlavius Frasincar, Ashwin Ittoo, Elisabeth Metais, Le Minh Nguyen
Pages97-109
Number of pages13
DOIs
Publication statusPublished - 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10260 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Fingerprint

Customers
Clustering
Attribute
Specification
Specifications
Review
Camera
Cameras
Evaluate
Output
Modeling

Cite this

Davril, J-M., Leclercq, T., Cordy, M., & Heymans, P. (2017). Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering. In F. Frasincar, A. Ittoo, E. Metais, & L. M. Nguyen (Eds.), Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings (pp. 97-109). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10260 LNCS). https://doi.org/10.1007/978-3-319-59569-6_10
Davril, Jean-Marc ; Leclercq, Tony ; Cordy, Maxime ; Heymans, Patrick. / Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering. Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings. editor / Flavius Frasincar ; Ashwin Ittoo ; Elisabeth Metais ; Le Minh Nguyen. 2017. pp. 97-109 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{feff9b5f66704414b2106bdbbe8bcccd,
title = "Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering",
abstract = "Online reviews are an important source of information that customers use to make more informed purchase decisions. Attribute-centric reviews, in which the author supports her opinion with comments on the technical attributes of the product, are particularly insightful because they present deeper discussions about how technical specifications can meet the expectations of customers. However, as the number of available reviews grows, it becomes increasingly cumbersome to manually locate attribute-centric reviews as they get lost within a flood of less informative reviews. We propose a word clustering approach that uses the technical specifications of products to identify technical discussions in online reviews. Each output cluster represents a technical aspect of the products and can be used to extract its related attribute-centric reviews. We evaluate our approach by modeling technical aspects for 21,846 reviews for cameras and show that our approach can extract and rank relevant technical comments.",
author = "Jean-Marc Davril and Tony Leclercq and Maxime Cordy and Patrick Heymans",
year = "2017",
doi = "10.1007/978-3-319-59569-6_10",
language = "English",
isbn = "9783319595689",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "97--109",
editor = "Flavius Frasincar and Ashwin Ittoo and Elisabeth Metais and Nguyen, {Le Minh}",
booktitle = "Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings",

}

Davril, J-M, Leclercq, T, Cordy, M & Heymans, P 2017, Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering. in F Frasincar, A Ittoo, E Metais & LM Nguyen (eds), Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10260 LNCS, pp. 97-109. https://doi.org/10.1007/978-3-319-59569-6_10

Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering. / Davril, Jean-Marc; Leclercq, Tony; Cordy, Maxime; Heymans, Patrick.

Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings. ed. / Flavius Frasincar; Ashwin Ittoo; Elisabeth Metais; Le Minh Nguyen. 2017. p. 97-109 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10260 LNCS).

Research output: Contribution in Book/Catalog/Report/Conference proceedingConference contribution

TY - GEN

T1 - Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering

AU - Davril, Jean-Marc

AU - Leclercq, Tony

AU - Cordy, Maxime

AU - Heymans, Patrick

PY - 2017

Y1 - 2017

N2 - Online reviews are an important source of information that customers use to make more informed purchase decisions. Attribute-centric reviews, in which the author supports her opinion with comments on the technical attributes of the product, are particularly insightful because they present deeper discussions about how technical specifications can meet the expectations of customers. However, as the number of available reviews grows, it becomes increasingly cumbersome to manually locate attribute-centric reviews as they get lost within a flood of less informative reviews. We propose a word clustering approach that uses the technical specifications of products to identify technical discussions in online reviews. Each output cluster represents a technical aspect of the products and can be used to extract its related attribute-centric reviews. We evaluate our approach by modeling technical aspects for 21,846 reviews for cameras and show that our approach can extract and rank relevant technical comments.

AB - Online reviews are an important source of information that customers use to make more informed purchase decisions. Attribute-centric reviews, in which the author supports her opinion with comments on the technical attributes of the product, are particularly insightful because they present deeper discussions about how technical specifications can meet the expectations of customers. However, as the number of available reviews grows, it becomes increasingly cumbersome to manually locate attribute-centric reviews as they get lost within a flood of less informative reviews. We propose a word clustering approach that uses the technical specifications of products to identify technical discussions in online reviews. Each output cluster represents a technical aspect of the products and can be used to extract its related attribute-centric reviews. We evaluate our approach by modeling technical aspects for 21,846 reviews for cameras and show that our approach can extract and rank relevant technical comments.

UR - http://www.scopus.com/inward/record.url?scp=85021776028&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-59569-6_10

DO - 10.1007/978-3-319-59569-6_10

M3 - Conference contribution

SN - 9783319595689

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 97

EP - 109

BT - Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings

A2 - Frasincar, Flavius

A2 - Ittoo, Ashwin

A2 - Metais, Elisabeth

A2 - Nguyen, Le Minh

ER -

Davril J-M, Leclercq T, Cordy M, Heymans P. Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering. In Frasincar F, Ittoo A, Metais E, Nguyen LM, editors, Natural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings. 2017. p. 97-109. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-59569-6_10