Smart speakers, such as Google Home or Amazon Echo, are entering our homes and enriching the ecosystem of the Internet of Things (IoT) already present in them. The Intelligent Personal Assistants (IPAs) they include allow users to ask for different information (e.g. the weather or a recipe), set reminders and lists, and directly control other IoT devices (e.g. lamps). These assistants, through a companion app installed on the owner’s smartphone, provide advanced features like the possibility to set up some personalization rules in the form of trigger-action: if something happens, then do something else. In alternative, there are visual programming tools such as IFFTT or Node-RED that allow for more complex management of IoT by creating also complicated rules, but these tools lack simple management and require some effort by less technical users. The voice processing capabilities of the IPAs and their knowledge of the IoT ecosystem in which they are inserted could be exploited for the creation of the rules through a purely vocal interaction.

This thesis aims to explore novel approaches for creating personalization rules through conversation between the user and an IPA. To this end, interviews were held focused on identifying a set of strategies used to create rules in trigger-action format by voice. Later the study illustrates the design and the implementation of two prototypes with different rules composition approaches: one completely based on voice interaction and the other requiring physical action on home devices to show the IPA what to do. The work continues with the evaluation conducted through an in-the-lab experiment with 10 users; the evaluation focused on the usability and the efficiency of the conversational agents, then we have carried out a comparison between the interfaces based on the considerations and the results achieved by testing.


Carlo Borsarelli

Thesis Details

Luigi De Russis, Alberto Monge Roffarello
Master Degree in Computer Engineering