Publication Date: 2020/02/08
Abstract: Product review is one of the criteria that is useful for prospective buyers to make decisions in purchasing a product. The large number of product reviews makes it difficult to make conclusions on the contents of product reviews so that consumers have difficulty in deciding to buy a product. To overcome this problem, we need a system that can automatically identify product features in product reviews. There are two steps before entering the summary generation: the first step is the extraction of product features which is carried out using the association mining method to get frequent itemsets with two word selection schemes, namely noun filtering and noun phrase filtering. The second step is the classification of extracted product features using a supervised learning approach with the Random Forest algorithm. Summarization of product reviews on each feature is carried out extractively by displaying product features with an orientation to separate positive and negative reviews.
Keywords: Association Mining, Classification, Opinion Summarization, Product Feature Extraction, Product Review, Random Forest.
DOI: No DOI Available
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT20JAN631.pdf
REFERENCES