Track

Business Intelligence and Knowledge Management

Abstract

Consumer feedbacks have been widely used for product improvement. These consumer reviews revealcustomer sentiments (e.g., like/dislike, fulfilled/unfulfilled etc.) about products and the degree of sentiments aswell. These reviews are good sources to gauge customer feelings, which are important to make essentialbusiness decisions. In this research, we analyzed textual movie reviews semi-automatically using linguisticanalysis instead of using manual mechanisms. Generally, adjectives in text reviews express reviewers’ feelingsabout a product while adverbs (gradable) explain the degree of these feelings. Using a well-known moviereview database, we analyzed the pattern of adjectives and adverbs that appeared in reviewers’ comments. Wecompared the frequencies of these adjective and adverbial words with the symbolic ratings (A+ to F) of therespective reviews and found strong correlation between the positive/negative terms (adjectives and adverbs)embedded in the text and their corresponding symbolic ratings.

Share

COinS