Poet

Abstract

Interaction-based systems have been widely used in many enterprises like Grab to enable quick and easy analysis of large-scale spatial data. Unlike traditional instruction-based query processing systems, modern interaction-based systems allow users to issue complex queries through simple interactions with a Graphical User Interface (GUI). While such systems have significantly transformed the process of spatial query processing, they still rely on a process-after-query approach for executing the queries. Even though the user is continuously interacting with the GUI, the actual processing is only initiated after the user completes their interactions, thus wasting the opportunities to reduce the response time of query processing. Inside Grab, we develop Poet, a progressive execution framework to continuously analyze user interactions and to perform progressive execution as soon as the system gains reasonable confidence regarding the user intentions. By integrating Poet, the interaction-based system can begin processing before the query is expressed in its whole by the user. The user interactions are captured and modelled in Markov chains, which guide the probability of progressive execution. For handling large-scale trajectory data in Grab, the progressive execution engine of Poet has been designed on top of Apache Flink. Our experiments show that Poet is able to reduce the latency in generating the output, providing a more interactive experience. Our experiments find that Poet helps reduce the query execution latency by up to 25x.