Title :
Product review sentiment classification using parts of speech
Author :
Tanawongsuwan, Patrawadee
Author_Institution :
Comput. Sci. Dept., Nat. Inst. of Dev. Adm. (NIDA), Bangkok, Thailand
Abstract :
A prospective buyer interested in a particular item may find out information about the item from various sources, including product reviews. With interactive information sharing facilitated by Web 2.0, a lot of product reviews are available on the web. For a popular item with a large number of reviews, a prospective buyer could use some help in selecting only reviews of interest, such as, only positive or negative reviews, when only particular kind of information is being sought for. This research work implemented a system that classified a product review as having either positive or negative tone, through the analysis of parts of speech of the review´s textual content. The system used machine learning algorithms for training positive-negative classification models. Experiments were performed particularly on textbook reviews.
Keywords :
Internet; learning (artificial intelligence); marketing data processing; pattern classification; text analysis; Web 2.0; interactive information sharing; machine learning algorithm; positive-negative classification model training; product review sentiment classification; textbook review; textual content; Accuracy; Book reviews; Observers; USA Councils; book; classification; machine learning; mining; parts of speech; review; sentiment; textbook;
Conference_Titel :
Computer Science and Information Technology (ICCSIT), 2010 3rd IEEE International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-5537-9
DOI :
10.1109/ICCSIT.2010.5563883
