Trustworthy services selection mechanism based on machine learning techniques

With the rapid growth of web services, users select the available services according to not only functional requirements, but also non-functional QoS characteristics. Therefore, research on how to find suitable and trustworthy service becomes increasingly important and challenging. In this paper, we propose an efficient Trustworthy Services Selection model named TSSelector by employing state-of-the-art machine learning techniques to improve services´ semantic representation and predict services´ QoS characteristics. Compared with the existing approaches, the TSSelector has at least two advantages. First, it imports WordNet and Latent Semantic Index to extend web services´ semantic and represent it as the low-dimensional compact feature vectors. Second, it employs matrix completion technique to predict and correct the missing and corrupted QoS values. Preliminary experimental results performed on the real-world data set demonstrate the feasibility of the proposed approaches.