New TOIS paper: "From Users' Intentions to IF-THEN Rules in the Internet of Things"
Our paper "From Users' Intentions to IF-THEN Rules in the Internet of Things" has been accepted to ACM Transactions on Information Systems (TOIS) as part of the special issue on conversational search and recommendation!
The paper presents HeyTAP2, a semantic Conversational Search and Recommendation (CSR) system that combines a semantic recommendation algorithm with a navigation-by-preference approach. By leveraging a conversational agent, users can communicate their current personalization intentions by specifying functionality at a high level, such as decreasing the temperature of a room when they leave it. Based on this input, HeyTAP2 implements a semantic recommendation process that considers the current user's intention, the connected entities owned by the user, and the user's long-term preferences revealed through their profile. If not satisfied with the suggestions, users can converse with the system to provide further feedback (i.e., short-term preferences), allowing HeyTAP2 to provide refined recommendations that better align with their original intention. Evaluation results show that the system outperforms similar baseline recommender systems, with recommendation accuracy and similarity to target items increasing as the interaction between HeyTAP2 and the user progresses.
More information: