An artificial intelligence framework for compensating transgressions and its application to diet management

Abstract

Today, there is considerable interest in personal healthcare. The pervasiveness of technology allows to precisely track human behavior; however, when dealing with the development of an intelligent assistant exploiting data acquired through such technologies, a critical issue has to be taken into account; namely, that of supporting the user in the event of any transgression with respect to the optimal behavior. In this paper we present a reasoning framework based on Simple Temporal Problems that can be applied to a general class of problems, which we called cake&carrot problems, to support reasoning in presence of human transgression. The reasoning framework offers a number of facilities to ensure a smart management of possible "wrong behaviors" by a user to reach the goals defined by the problem. This paper describes the framework by means of the prototypical use case of diet domain. Indeed, following a healthy diet can be a difficult task for both practical and psychological reasons and dietary transgressions are hard to avoid. Therefore, the framework is tolerant to dietary transgressions and adapts the following meals to facilitate users in recovering from such transgressions. Finally, through a simulation involving a real hospital menu, we show that the framework can effectively achieve good results in a realistic scenario.

A Survey on Food Computing

“…[Min et al 2018b] explored an audio-kinetic model of well-formed multi-sensory earable devices for dietary monitoring. In addition, there are also some works, which applied constraint reasoning to dynamically adapt dietary recommendations for compensating diet transgressions [Anselma and Mazzei 2015;Anselma et al 2017Anselma et al , 2018.…”

Building a Persuasive Virtual Dietitian

“…In MADiMan, we represent the DRVs as an STP [12]. An STP consists of a conjunction of bounds on differences.…”

“…It consists of three nested loops that compute the all-pairs shortest paths of the STP and it corresponds to the Floyd-Warshall algorithm. In [12], we described how such a basic algorithm can be exploited to determine the consistency of some meals with respect to a diet and to perform a what-if analysis to foresee the consequences of a diet violation. For the sake of simplicity, in the figure, we do not represent the macronutrients, nor the single meals, but only the total energy in a day.…”

“…In previous work, we discussed the general architecture of the MADiMan system [10,11] and the capacity of the reasoning module with a number of simulation experiments with virtual agents based on hospital menus [12] and on Mediterranean menus [13,14]. The research questions that we explore in this paper concern the NLG module of MADiMan, which has been partially previously described in some preliminary work [15,16].…”

Designing and testing the messages produced by a virtual dietitian

“…Moreover, he is recommended to assume (LARN, 2014), e.g., 260 kcal/day of proteins, 735 kcal/day of lipids and 1455 kcal/day of carbohydrates. In MADiMan we represent the DRVs as STP constraints (Anselma et al, 2017). STP models a set of constraints as a conjunction of bounds on differences.…”