[3Rin4] Interactive 1

[3Rin4-91] Association Analysis of Pet Foods with Diseases in dogs Using E-Commerce Data

〇Naomi Sasaya¹, Shigefumi Kishida², Ryo Kikuchi², Shuji Yamaguchi¹, Akira Tajima¹ (1.Yahoo Japan Corporation, 2.Anicom Holdings, Inc.)

Keywords:machine learning, big data, pet food, healthcare, allergy

In this study, we investigated the effects of pet foods on pet’s health using five datasets: tens of thousands of questionnaire answers from pet insurance members, claim data from pet insurance members, the more than millions purchase histories of pet foods on the e-commerce site, ingredients of pet foods, and categories of ingredients. Since there is no information about the diseases on the e-commerce site, we used specific purchase activities as signs of diseases, and built a dataset with positive/negative labels based on purchase histories of e-commerce site. This time, as an example of the application of big data using alternative indicators, we investigated the ingredients of dog food that affect allergic diseases in dogs, and identified the ingredients that affected those disease.