Temporal reasoning for web services composition for personal assistants

Abstract

Virtual Personal Assistants (VPAs) are digital assistants like Amazon Alexa, Apple Siri, and Google Home that handle the queries requested by the users. To answer the complex queries, VPAs may need to do some reasoning to process those queries right. In this thesis, we focus on the temporal aspects of the input queries for the VPA, which is currently not catered completely to having an area of improvement. Temporal reasoning is a type of reasoning for the values based on time, so the focus is on the queries having time-based values. Some examples are experimented personally with selected VPAs to know the problem and the dealing of VPAs with temporal values. During these experiments, some anomalies are observed in the response of the VPAs. For the use case, initially, the selection of a domain is done, and then, a Smart Home (SH) environment is selected as a domain to consider the use cases and do the possible temporal reasoning of the queries a VPA can get as input. For the solution, we propose a system with a temporal reasoner component, and a simulation of the system is implemented as a proof of concept. The proposed system has the integration of different technologies with a component of the temporal reasoner. The concept of a temporal reasoner is implemented and some pre-defined scenarios in an SH are simulated. The simulation executing input queries with temporal reasoning concludes the bright side of the application of temporal reasoning in an SH environment.