A content-based recommendation system for animal adoption using the weighted cosine technique

Abstract

An adopted animal is considered a grand gesture with significant social responsibility since it needs daily care and affection from the adopter. Far beyond appearance and aesthetic issues, the adopter must consider their natural conditions, respecting the specific characteristics of the animal. The lack of expertise in choosing the animal can lead to an irresponsible and fruitless adoption, compromising the animal’s quality of life. Animal adoption portals are prone to this problem if they do not provide controls that try to approximate the adopter’s real motivations and conditions with the animal’s characteristics. Faced with this problem, this article proposes a content-based recommendation system to suggest animals for users to consume drugs, applying the weighted cosine technique. The weighting is fundamental in calculating the similarity between the users’ emotions and the animal’s characteristics. The experiments increased the accuracy of the recommendations when weighting is applied over the traditional method without considering the prescribed method of users by specific characteristics of the animals to be adopted.

Keywords:
Recommender system; Content-based recommendation; Adoption of animals; Dog adoption; Weighted cosine technique