Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering

Résultats de recherche: Contribution dans un livre/un catalogue/un rapport/dans les actes d'une conférenceArticle dans les actes d'une conférence/un colloque

Résumé

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.

langue originaleAnglais
titreNatural Language Processing and Information Systems - 22nd International Conference on Applications of Natural Language to Information Systems, NLDB 2017, Proceedings
rédacteurs en chefFlavius Frasincar, Ashwin Ittoo, Elisabeth Metais, Le Minh Nguyen
Pages97-109
Nombre de pages13
Les DOIs
Etat de la publicationPublié - 2017

Série de publications

NomLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10260 LNCS
ISSN (imprimé)0302-9743
ISSN (Electronique)1611-3349

Empreinte digitale Examiner les sujets de recherche de « Technical Aspect Extraction from Customer Reviews based on Seeded Word Clustering ». Ensemble, ils forment une empreinte digitale unique.

Contient cette citation