A 'Catchy' Copy and Concept Evaluation System Using a Natural Language Processing Approach

摘要

In advertising, marketing research, CF (Commercial Film) creation, and drama creation, it is very important to be able to predict the words or concepts that will come into vogue. Initially, a creator comes up with many candidates for future vogue words or concepts. Next, the creator selects the best word or concept for the most "catchy" copy. This selection process is very hard because it is essentially a sensory and intuitive job. Thus, the creator often cannot explain why he/she thinks a particular candidate word is the best one. To resolve this problem, this paper demonstrates a new approach using natural language processing techniques and the "Contemporary Word Dictionary" (Jiyukokumin-sha, Tokyo) to support creators in selecting the best word or concept from several candidates. This system is based on the assumption that a vogue word or concept is a reflection of the social, economic, or political situation. Using this system, the "Most Popular Vogue Words" selected in the year 2000 by the publishing company Jiyukokumin-sha were evaluated by comparison with the former "Most Popular Vogue Words" selected in the years 1997-1999. The experimental results obtained were also tested statistically. The statistical test showed that the result was expressive. This means that the proposed approach is both practical and effective.