QualiFood: An Intelligent Quality Food Evaluation Using Logical Satisfiability Reasoning On Spark

Abstract

There is an urgent need to address unhealthy dietary patterns for people. Therefore, having a high food quality is essential to health by adopting good eating habits. Furthermore, the evaluation of food quality is complex because it needs huge numbers of information related to food, for example its ingredients, the environmental conditions of food productions, the consumer state of health. Such information are scattered on multiple and non communicating systems including food producer systems, open data, medical data, etc. Applying the semantic to a very large collection of information from different data sources can highly contribute to enhance the food quality score. Hence, it will improve the individual diet/health by giving adequate nutritional recommendation and avoiding inconsistent food mixing by following set of rules.In this paper, we propose an intelligent and scalable approach to ensuring the food quality for meal healthiness query answering over big data related to food in a distributed way on a Spark ecosystem. For that, the cleaning inconsistent and contradictory big data approach is built by following the steps (1)modeling the consistency rules including inference and inconsistent rules (2) detecting inconsistency through rule evaluation on Apache Spark framework to discover the minimally subset of inconsistent data (3) cleaning the inconsistency through finding the cleaned meals for consistent query answering.