Measuring the credibility of Arabic text content in Twitter

Recent research in measuring web content credibility automatically for text and multimedia have addressed many languages. Unfortunately, there is no research for credibility measurements of Arabic web content. Thus, this paper proposes a solution to automatically measure the credibility of Arabic web content. The automatic tool we present in this paper measures Arabic content credibility published in Twitter (a micro blogging service) targeting the news domain. Our tool is based on two approaches. The first approach uses the similarity between Twitter posts and authentic news sources, while the second approach is based on a set of proposed features, among them is the similarity with verified content. Preliminary evaluations of our tool have shown that the first approach for rating tweets credibility has higher precision and recall compared with the second approach.