More and more devices featuring internet connectivity are being created every day. The value of s... more More and more devices featuring internet connectivity are being created every day. The value of such devices is in their ability to interact with one another, as part of automation workflows created by users to improve their home experience. Various platforms offer systems that allow for the creation of automation rules between different connected devices, but formulating and defining such automations can be quite complex. This work explores the automated, personalized recommendation of automation rules for users of connected devices. The application of recommender systems’ techniques to this domain is novel, so that exploring the applicability of the recommendation approaches discussed in the literature is one of the main contributions of this work. The hypothesis that certain groups of automations provide synergies was explored, and a system that exploits this by providing recommendations based on learning association rules was developed. Various strategies to make association rule mining feasible on an automation dataset were devised, such as generalizing automation rules into recommendable items, and applying a similarity operator to all items, making automations identifiable across users. The developed system showed very promising results, under an evaluation methodology consisting of gathering precision, recall and coverage metrics for the most important hyperparameters, as well as when comparing it against a naive recommender developed as a baseline. Finally, it was discussed how the techniques developed in this work are generalizable to domains outside of automation rule territory.
More and more devices featuring internet connectivity are being created every day. The value of s... more More and more devices featuring internet connectivity are being created every day. The value of such devices is in their ability to interact with one another, as part of automation workflows created by users to improve their home experience. Various platforms offer systems that allow for the creation of automation rules between different connected devices, but formulating and defining such automations can be quite complex. This work explores the automated, personalized recommendation of automation rules for users of connected devices. The application of recommender systems’ techniques to this domain is novel, so that exploring the applicability of the recommendation approaches discussed in the literature is one of the main contributions of this work. The hypothesis that certain groups of automations provide synergies was explored, and a system that exploits this by providing recommendations based on learning association rules was developed. Various strategies to make association rule mining feasible on an automation dataset were devised, such as generalizing automation rules into recommendable items, and applying a similarity operator to all items, making automations identifiable across users. The developed system showed very promising results, under an evaluation methodology consisting of gathering precision, recall and coverage metrics for the most important hyperparameters, as well as when comparing it against a naive recommender developed as a baseline. Finally, it was discussed how the techniques developed in this work are generalizable to domains outside of automation rule territory.
Uploads
Papers by Diogo Franco
This work explores the automated, personalized recommendation of automation rules for users of connected devices. The application of recommender systems’ techniques to this domain is novel,
so that exploring the applicability of the recommendation approaches discussed in the literature is one of the main contributions of this work. The hypothesis that certain groups of automations provide synergies was explored, and a system that exploits this by providing recommendations based on learning association rules was developed. Various strategies to make association rule mining feasible
on an automation dataset were devised, such as generalizing automation rules into recommendable items, and applying a similarity operator to all items, making automations identifiable across users.
The developed system showed very promising results, under an evaluation methodology consisting of gathering precision, recall and coverage metrics for the most important hyperparameters, as well as when comparing it against a naive recommender developed as a baseline. Finally, it was discussed how the techniques developed in this work are generalizable to domains outside of automation rule territory.
This work explores the automated, personalized recommendation of automation rules for users of connected devices. The application of recommender systems’ techniques to this domain is novel,
so that exploring the applicability of the recommendation approaches discussed in the literature is one of the main contributions of this work. The hypothesis that certain groups of automations provide synergies was explored, and a system that exploits this by providing recommendations based on learning association rules was developed. Various strategies to make association rule mining feasible
on an automation dataset were devised, such as generalizing automation rules into recommendable items, and applying a similarity operator to all items, making automations identifiable across users.
The developed system showed very promising results, under an evaluation methodology consisting of gathering precision, recall and coverage metrics for the most important hyperparameters, as well as when comparing it against a naive recommender developed as a baseline. Finally, it was discussed how the techniques developed in this work are generalizable to domains outside of automation rule territory.