Abstract
Nowadays social semantic Web has become one of the most important sources of data. The quality of data is crucial for data integration and annotation in social Linked Open Data (LOD). The main goal of tweets classification and linking is to integrate all tweets that are related semantically in one place on the Web. The data inequality problems exist because authors post tweets and hashtag them based on their own preference. The authors can post tweets in different languages (English-Arabic). Therefore, the hashtags cannot be the only trackback function of the tweet to link it with the related tweets. Thus, the goal of this paper is to develop an automated tweet classifier which is not limited by the language or the category of the tweet. The approach can create new suggested multilingual (English, Arabic) hashtags from tweet's content and comments "autotagging" in social linked open data (LOD), matching the tweets in the same topic and provides high accuracy and performance using blocking and indexing techniques. The proposed framework has substantial improvements in tweets classification and linking compared to state of the art framework of classifying and linking techniques with better classifying rate which reach 95% in precision and 97% in recall.