M. Ogawa, F. Harada, H. Shimakawa, and Hirokazu Fukuda (Japan)
Individual advertisement, Characteristics of the product,Correction of the interest indicators, Recommendation
The advertisement expenditure including Internet advertisement in Japan grows larger every year. Companies emphasize advertisements when they sell new products and services. Advertisements are also essential for consumers because they refer advertisements to obtain information and features of the products and services. Especially, the internet advertisement enables a customer to inspect, search for, and buy products and services. In internet advertisement, information on the product should be provided to each customer individually according to his/her interest. In this paper, we propose a method to create an individual advertisement about a product based on each customer’s interest and information on the product. The interest and information are numerically represented. In this method, information on the product is divided with various granularities and classified with a tree structure. An advertisement is created by combining the classified information based on the user interest. The user interest numerically expressed is gradually corrected based on the browsing history of the created advertisement. The proposed method can provide an internet advertisement where the included contents agree with the features which are interesting to each customer. Using this method, customers may consider buying the product. The experimental result shows that 80% of examinees received appropriate advertisements, where more than half of received advertisements interest each examinee.
Important Links:
Go Back
