Meal plan search engine inspired by ant colony’s foraging

Abstract

mobile phones, consumer electrical appliances, and other information devices, we observe a continuous improvement in the quality of consumer services. This environment, known as "Ubiquitous computing" or "Smart home appliance," has the potential for further development [1].Smart home appliances such as the refrigerator not only keep food fresh, but also manage information about food ingredients, act as a search engine for recipes, and communicate with the "digital kitchen". Such digital collaboration among appliances and other information devices is a leading example of this environment.However, a conventional, keyword-based search engine is not effective for such a system, making the system inefficient to search for recipes that can be cooked using only the ingredients in the refrigerator [2,3].In this paper, we propose a system that recommends meal plans and recipes that can be cooked using only the available food items in the refrigerator. To realize this new search engine, we focus on the mechanism of an ant colony's foraging [4][5][6]. It is reported that ants repeat simple actions such as "lifting up", "carrying", and "lifting down," and the ant colonies make food clusters. The food clusters are built randomly, and therefore, the size and place of each cluster is different.In the proposed search engine, we consider the food items in the refrigerator to be the ants' food and the cookable dishes to be the ants' food clusters. Hence, the search engine (for meal plans) represents an ant colony's foraging in the virtual world. We also validate the effectiveness of the proposed search engine by performing simulations. Goals of the proposed systemThe proposed system recommends meal plans and recipes that can be cooked only using the available food Abstract Currently, the communication link between consumer electrical appliances and information devices is developing; for example, a refrigerator, which keeps food fresh, now also manages information about the food ingredients it contains. This environment, known as "Ubiquitous computing" or "Smart home appliance," has the potential for further research and development. In this environment, we consider a system that recommends food recipes or meal plans that can be cooked with the available ingredients in the refrigerator. However, a conventional keyword search is not effective for such a system, and therefore, in this paper, we propose an algorithm for the search engine that is inspired by an ant colony's foraging. In addition, we validate the effectiveness of the proposed search engine through simulations.

Managing Knowledge in Computational Models for Global Food, Nutrition and Health Technologies

“…Mathematical models are based on some mathematical formulation such as Linear Programming (LP) which has been used to model the MPP [40,41]. Computational models use relevant computing paradigms and formalisms such as: GA [22], Quantum Particle Swarm Optimisation [42], Ant Colony Optimisation [42,43] and Knapsack Problem [44]. Hybrid models integrate two or more models [29,37,45].…”

“…This finding confirms what is in found in literature since there are some works [46], [22], and [48] which incorporated domain knowledge in GAs for the MPP even though the knowledge was hard to manage. However, there are very few works in literature which modelled the MPP using GA but without incorporating much domain knowledge such as [43]. As can be seen these works are now aged and the current trend is to incorporate domain knowledge in GA in order to solve the MPP.…”