Author(s): Suryakumar B, Dr. E.Ramadevi

Nowadays there have been many technologies which provide positioning services eg., smart phone sensors, location estimation of 802.11, ultrasonic systems and so on. As a consequence, it is becoming easier to generate a large scale trajectory data of tracking traces of moving objects. The explosion of location-based social networks such as facebook, watsapp and etc provides number of ways for tracing human mobility including user generated geo-tagged contents, checkin-services and mobile applications. However, the issues in many applications are analyzing and mining trajectory data due to complex characteristics reflected in human mobility which is affected by multiple contextual information. Hence, this paper is focused on the challenges obtained in mining human trajectory data. In this paper, multi-context trajectory embedding model (MC-TEM) is proposed which is developed in advanced deep neural network called as Convolutional Neural Network (CNN). It characterizes the several types of contexts for different applications. The CNN is used in the contextual feature learning process. This CNN based MC-TEM model reduce the computation time of feature learning and also this method needs fewer parameters to tune. This method is applied for location recommendation and social link prediction. The proposed method is tested in three real time dataset to prove the effectiveness of the proposed method in terms of precision, recall and F-measure metrics.