Comparing N-gram Models for Tag Suggestion in Tagging System

On the web, a tag is a significant keyword of content, including a photo, video, and blog article. A tagging is an action of adding a tag to content. Many web sites, such as del.icio.us 1 and CiteULike 2 , are providing tagging system. These tags can be used for web search. Users have already recognized the value and the importance of tags, but some users do not use tags. These users might feel annoyed to be forced to add tags, or they might simply not know what to add in order to obtain a good search result. These problems are the reasons why tag suggestion system would be beneficial. In this paper, we use n-gram models for tag suggestion in tagging system. We gathered tag data from various web sites. Based on crawled tag data, we will employ various n-gram models and compare obtained results in the next paper. This is a progress paper. KeywordsWeb 2.0; Folksonomy; Tag Suggestion; Tagging; Ngram; Natural Language Processing