Open Directory Project based universal taxonomy for Personalization of Online (Re)sources

作者：

Highlights：

• We examine the overall quality of Open Directory Project as a classifier.

• We compare several data grouping schemes and evaluate each of them.

• We show that ODP can be used both as a horizontal as well as vertical classifier.

• GENERAL grouping scheme yields best results for horizontal classification.

• Number of documents (limit models) yield best results for vertical classification.

摘要

•We examine the overall quality of Open Directory Project as a classifier.•We compare several data grouping schemes and evaluate each of them.•We show that ODP can be used both as a horizontal as well as vertical classifier.•GENERAL grouping scheme yields best results for horizontal classification.•Number of documents (limit models) yield best results for vertical classification.

论文关键词：Recommendation systems,Content personalization,Automatic content classification,Automatic content labeling,Information extraction,Information retrieval,Open Directory Project,Vector Space Modeling,TF-IDF

论文评审过程：Available online 25 April 2015, Version of Record 15 May 2015.

论文官网地址：https://doi.org/10.1016/j.eswa.2015.04.033