The growth of inbound travel is fully coordinate with the successful urban development. Increasing the number of inbound travelers not only creates more jobs and economic opportunities but also drives the country toward prosperity. Thus, inbound traveler analysis through trajectory pattern mining, a subfield of urban computing, is regarded as a promising solution. This paper introduces large-scale mobile ad requests as an alternative data source of trajectory pattern mining in order to eliminate the limitations of conventional data sources, such as GPS data, cellular data, and IP address data. In addition, to expedite a comprehensive inbound traveler analysis, we build TraVis, a real-world system for efficiently exploring the inbound travelers' activities through the interactive visualization interface. By incorporating various modules, such as mobile users' home country and travel intent prediction, frequent trajectory pattern mining, and interactive visualization, TraVis proves the capability of profiling the travelers' behavior pattern. We use Japan inbound travelers in the case study to present the mining insights, and we also demonstrate the extensive system functionalities. Our system has been assisting Japan government agencies to formulate travel marketing strategies, including tourist experience enhancement and attractions marketing.

There is great interest in exploiting the opportunity provided by cloud computing platforms for large-scale analytics. Among these platforms, Apache Spark is growing in popularity for machine learning and graph analytics. Developing efficient complex analytics in Spark requires deep understanding of both the algorithm at hand and the Spark API or subsystem APIs (e.g., Spark SQL, GraphX). Our BigDatalog system addresses the problem by providing concise declarative specification of complex queries amenable to efficient evaluation. Towards this goal, we propose compilation and optimization techniques that tackle the important problem of efficiently supporting recursion in Spark. We perform an experimental comparison with other state-of-the-art large-scale Datalog systems and verify the efficacy of our techniques and effectiveness of Spark in supporting Datalog-based analytics.