A Query Construction Method Based on Data Dependency for Gesture Interaction

It is becoming more popular to explore data by multi-touch gesture at present. In order to improve the fluency of interaction, it is necessary to construct queries and provide real-time feedback in the process of a gesture. In this paper, a query construction method based on data dependency for gesture interaction, called DDQC(Data Dependency based Query Construction), is proposed. The movement characteristics of multi-touch gesture and the data dependency on table are taken into account to provide query results many times during a continuous multi-touch gesture. In DDQC, the division point of touching trajectory is presented to dynamically determine the time to query, and the query context of current gesture and the functional dependency on table are used to construct the query event. By collecting the real behavior data of gestural query, the experiments compare the query performance of DDQC, SnapToQuery, TouchViz and Binned Aggregation, and the results show that DDQC has better performance in Query Execution Time, I/O Times and Discovery Time.