An Improved Adaptive model for Information Recommending and Spreading

People in the Internet era have to cope with information overload and expend great effort on finding what they need. Recent experiments indicate that recommendations based on users' past activities are usually less favored than those based on social relationships, and thus many researchers have proposed adaptive algorithms on social recommendation. However, in those methods, quite a number of users have little chance to recommend information, which might prevent valuable information from spreading. We present an improved algorithm that allows more users to have enough followers to spread information. Experimental results demonstrate that both recommendation precision and spreading effectiveness of our method can be improved significantly.