| Creator |
|
|
|
|
|
|
|
| Language |
|
| Publisher |
|
|
|
| Date |
|
| Source Title |
|
| Vol |
|
| Issue |
|
| First Page |
|
| Last Page |
|
| Publication Type |
|
| Access Rights |
|
| JaLC DOI |
|
| Related DOI |
|
|
|
|
|
| Related URI |
|
|
|
|
|
| Relation |
|
|
|
|
|
| Abstract |
There are a lot of information recommender systems on the Web. These systems aim to find and provide useful information for the users. For example, many online shopping sites recommend merchandise whi...ch the user is likely to purchase. However, practical uses do not always use these recommendations. In order to make an apt recommendation for the user, we need to give a plausible reason for it. However, the almost all existing systems give very simple or quantitative reasons. This paper aims to present clear and non-quantitative recommendation reasons which everybody is easy to understand. We make use of rules generated by Inductive Learning for the aim. We use Inductive Logic Programming (ILP) and Decision Tree. ILP generates first order predicate logic rules, and Decision Tree generates tree structures based on propositional logic. As an experience, we extract several rules from blogs with ILP and Decision Tree in order to recommend blogs or web pages which the bloggers are likely to be interested in. We succeeded in extracting several useful rules from blogs. We show that the rules could be used for not only recommendations but also giving their reason. The recommendation methods using in this paper are based on collaborative filtering and content-based filtering.show more
|
Export Mendeley
paper6