An Empirical Study on Learning to Rank of Tweets

摘要: Twitter, as one of the most popular micro-blogging services, provides large quantities fresh information including real-time news, comments, conversation, pointless babble and advertisements. Twitter presents tweets in chronological order. Recently, introduced a new ranking strategy that considers popularity terms number retweets. This method, however, has not taken into account content relevance or twitter account. Therefore amount inevitably flood relevant tweets. paper proposes which uses only tweet, but also authority tweet-specific features such whether URL link is included tweet. We employ learning to rank algorithms determine best set with series experiments. It demonstrated tweet contains not, length are conjunction.