A Method for Eliciting Customer Requirements from Question Sentences

The modern development of the internet has led to product information being provided in online markets. User reviews provide various types of information in addition to product information; they comprise a channel for communicating product information and affect consumer purchasing decisions. As such, user reviews can have a significant influence on sales. However, they are also noted as a place for identifying user needs because a large number of user reviews can be obtained at low cost. User review analysis generally uses opinion mining methods, which largely divide reviews into being positive, negative, or neutral; analysis is performed only on positive and negative user reviews after neutral reviews are discarded. User reviews containing question sentences are judged to be neutral, yet they can contain content important for identifying user needs. Therefore, this paper proposes a method to derive customer requirements from the user reviews containing question sentences. It is shown that user needs are contained within these user reviews through analysis of the results. New user needs can be found through this method and a large number of user needs can be obtained at low cost.