JP7559010B2 - Food recommendation system and food recommendation method - Google Patents

Description

The present invention relates to a food recommendation system and a food recommendation method for pets, and more particularly to a food recommendation system and a food recommendation method that recommend food suitable for a pet based on the insurance claim rate for each breed to which the pet belongs.

Pets (hereafter sometimes abbreviated to "animals") are irreplaceable to humans. In recent years, while the average lifespan of pet dogs and other animals has increased significantly, the number of animals suffering from illnesses during their lives has increased, and the increasing medical expenses borne by owners has become a problem.

As a result, interest in the health of pet animals is growing. Diet in particular is considered to be important for the health of pet animals, and a variety of pet foods (hereafter sometimes abbreviated to "food") have been proposed and are on sale.

Foods currently on the market are advertised as having unique nutritional compositions and blends, but are not tailored to the constitution or health condition of individual pet animals. Some foods are sold specifically for animals in each category, categorizing animals into categories such as for pregnant animals, for elderly animals, and for obese animals, but even when referring to obese animals, there are many different causes and ways to solve obesity, and it may not be the optimal solution to feed all pets the same food for obesity.

Therefore, there is a demand for food that is tailored to the constitution of each pet animal.

Patent Document 1 discloses a food suggestion system that suggests foods for preventing hereditary diseases that are effective in preventing or inhibiting the progression of hereditary diseases in pet animals and that can be safely given to the animals over the long term.

However, the food recommendation system described in Patent Document 1 requires genetic analysis of pet animals as a prerequisite for making suggestions, and is therefore not easy to use. In addition, research into hereditary diseases is less advanced than in humans, making it difficult to identify hereditary diseases through genetic analysis alone. In addition, there is also the possibility that the disease is not hereditary but rather acquired.

JP 2021-114999 A

Anicom Household Animal White Paper 2021, Anicom Holdings Inc.

The present invention aims to provide a food recommendation system and a food recommendation method that can recommend foods suitable for improving the health and preventing disease in individual pet animals, even if it has not been determined whether the disease is hereditary.

The inventors discovered that the above problem could be solved by proposing food using at least insurance claim data, and thus completed the present invention.

That is, the present invention relates to the following [1] to [12].
[1] A food recommendation system comprising a recommendation means for recommending food suitable for a pet based on insurance claim data for the breed to which the pet belongs.
[2] A food recommendation system comprising a recommendation means for recommending food suitable for a pet based on insurance claim data on the breed and age of the pet.
[3] A food recommendation system according to [1] or [2], comprising a reception means for receiving individual information of a pet.
[4] The food recommendation system according to any one of [1] to [3], further comprising a determination means for determining the breed of the pet from a facial image of the pet.
[5] The determination means determines the breed of a pet animal from a facial image of the pet animal by using a trained model,
The food recommendation system of [4], wherein the trained model is trained using facial images of pet animals and the breed of the pet animals as training data, and the input is the facial image of the pet animal and the output is the breed of the pet animal.
[6] The food recommendation system of [1], wherein the insurance claim data is data regarding the insurance claim ratio by disease for each breed.
[7] The food recommendation system according to [1], wherein the recommendation means is for recommending the following food (A) or (B): (A) a food containing nutrients related to the prevention of a disease having a high insurance claim rate in the variety;
(B) Food that does not contain any nutrients that should be avoided in order to prevent diseases associated with a high rate of insurance claims in said breed, or that has low amounts of said nutrients.
[8] The food recommendation system described in [2], wherein the recommendation means uses data regarding the age of the pet and data regarding the most frequent age for diseases with a high insurance claim rate for the breed to which the pet belongs to recommend food appropriate to the age of the pet.
[9] The food recommendation system of [8], wherein the recommendation means recommends food that is useful for preventing the disease before the pet reaches an age equal to the modal age of the disease that has a high insurance claim rate for the breed to which the pet belongs.
[10] A food recommendation method for recommending food suitable for a pet based on insurance claim data for the breed to which the pet belongs.
[11] A food recommendation method for recommending food suitable for a pet based on insurance claim data on the breed and age of the pet.
[12] Food suggested by the food recommendation system in [1].

The present invention makes it possible to provide a food recommendation system and a food recommendation method that can recommend foods that are useful for improving the health and preventing diseases of individual pet animals.

1 is a configuration diagram showing one embodiment of a food recommendation system of the present invention. FIG. 13 is a diagram showing insurance claim rates etc. for toy poodles. FIG. 13 is a diagram showing insurance claim rates for Chihuahuas. FIG. 13 is a diagram showing insurance claim rates for miniature dachshunds. FIG. 13 is a diagram showing the insurance claim ratios for Shiba Inu dogs. FIG. 1 is a diagram showing diseases that each breed is susceptible to. FIG. 1 is a diagram showing ingredients that are effective against diseases. FIG. 1 is a diagram showing an example of a proposed hood.

<Food recommendation system>
The food recommendation system of the present invention is a food recommendation system that recommends food suitable for pets, and includes a recommendation means that recommends food suitable for pets based on insurance claim data for the breed to which the pet belongs. These comprehensive or specific aspects may be realized by a system, an apparatus, a method, an integrated circuit, a computer program, or a recording medium, or may be realized by any combination of the system, the apparatus, the method, the integrated circuit, the computer program, and the recording medium.

[Pet animals and breeds]
In the food recommendation system of the present invention, a user can input the breed of the target animal into the system, and thus food suitable for the animal can be recommended. Here, the target animal in the present invention is a dog or a cat. A breed is a unit of a group of organisms below the species level. For example, a breed of dog is also called a dog breed, and specific examples thereof include a toy poodle, a chihuahua, a miniature dachshund, a Shiba Inu, a Pomeranian, a Yorkshire terrier, a miniature schnauzer, a Shih Tzu, a French bulldog, a Papillon, a Maltese, a Labrador, a Dalmatian, and a Chow Chow. A breed of cat is also called a cat breed, and examples thereof include a Scottish Fold, an American Shorthair, a Norwegian Forest Cat, a Russian Blue, a British Shorthair, a Ragdoll, a Maine Coon, and a Persian. Examples of rabbits include a Netherland Dwarf, a Holland Lop, a Lop Ear, a Mini Rex, a Dwarf Lop, and an American Fuzzy Lop.

[Type determination]
The present invention may also include a breed determination means for inputting a facial image (photograph) of an animal by a user and determining the breed from the facial image. The breed determination means is a means for determining the breed of an animal from the image of the animal input to the receiving means by using a trained model for breed determination. By determining the breed and using a trained model corresponding to the breed, it is possible to improve the accuracy of individual identification. Note that if the breed determination means is not provided, the specifications may be such that breed information is input separately from a user's terminal or the like.

The image may be a still image or a video. In this specification, the term "image" includes both still images and videos. In the case of a video, a facial image is a video in which a face appears. It is sufficient that the video contains a scene in which a face appears. A video is a collection of consecutive still images, and can be processed in the same way as a still image.

As a trained model for variety determination, artificial intelligence (AI) is preferable. Artificial intelligence (AI) is software or a system that mimics the intellectual tasks performed by the human brain using a computer, specifically, computer programs that understand natural language used by humans, perform logical inference, and learn from experience. AI can be either general-purpose or specialized, and can be any of deep neural networks, convolutional neural networks, etc., and publicly available software can be used.

To generate a trained model, artificial intelligence is trained using training data. Either machine learning or deep learning can be used for the learning, but deep learning is preferred. Deep learning is an advanced version of machine learning, and is characterized by its ability to automatically find features.

The learning method for generating the trained model is not particularly limited, and publicly available software can be used. For example, DIGITS (the Deep Learning GPU Training System) published by NVIDIA can be used. In addition, training may be performed using a known support vector machine method published in, for example, "Introduction to Support Vector Machines" (Kyoritsu Shuppan) or the like. As a learning dataset, a publicly available dataset such as ImageNet can be used.
The training data for learning is an image of an animal's face and the breed of that animal. It is preferable to use an image of the animal's entire face including the eyes, nose, mouth, ears, and facial contour as the training data.

[Insurance claims data]
The insurance claim data is data related to insurance claims for animals that have animal insurance. For example, data on what diseases or injuries have caused insurance claims in the past for a certain breed of an animal can be cited. Preferably, the data is the proportion of insurance claims by disease for each breed. It is more preferable to include the age at which insurance claims are most frequent for each breed and each disease, that is, the so-called mode age. It is also preferable to include the average annual medical expenses for each breed and each disease. Insurance claim data can be obtained, for example, from companies that provide animal insurance or veterinary clinics, and can also be obtained by conducting a survey of animal owners.

[Insurance Claims Data Analysis]
Insurance claim data analysis involves analyzing insurance claim data for animals insured under animal insurance to calculate at least the insurance claim ratio and medical expenses for each breed and disease. By clarifying the insurance claim ratio, it is possible to identify hereditary diseases that are not generally known and diseases caused by physical constitution (body size, leg length, etc.). In addition, by calculating medical expenses (for example, medical expenses for a year), it is possible to identify diseases that are a large financial burden for owners. In addition, diseases in this invention also include injuries that are likely to be suffered due to physical constitution, etc.

[disease]
The diseases in the present invention can be appropriately organized and classified by the system administrator, etc. For example, circulatory diseases include valvular disease, cardiomyopathy, arrhythmia, atrial/ventricular midwall defect, etc., respiratory diseases include rhinitis, pharyngitis, tracheitis, pneumonia, diaphragmatic hernia, asthma, etc., digestive diseases include esophagitis, esophageal obstruction, gastritis, bacterial enteritis, ascariasis, etc., liver/biliary tract/pancreatic diseases include hepatitis, cholangitis, pancreatitis, liver/biliary tract/pancreatic tumors, etc., urinary diseases include chronic kidney disease, kidney stones, cystitis, urolithiasis, etc., reproductive diseases include ovarian diseases, pyometra, vaginitis, prostatitis, balanitis, etc., neurological diseases include epilepsy, meningitis, hydrocephalus, brain tumors, etc., eye diseases include conjunctivitis, traumatic keratitis, cataracts, glaucoma, etc., ear diseases, Examples of diseases that may be caused by infection include bacterial otitis externa, allergic otitis externa, ear hematoma, otitis media, and otitis interna; dental and oral diseases include periodontal disease, dental root abscess, salivary gland disease, and stomatitis; musculoskeletal diseases include herniated disc, cruciate ligament injury/rupture, arthritis, fracture, sprain, and osteosarcoma; skin diseases include pyoderma, demodicosis, atopic dermatitis, melanocytoma/melanoma, and nail trauma; blood and hematopoietic diseases include anemia, bone marrow abnormalities, and lymph node swelling; endocrine diseases include diabetes, hyperthyroidism/hypothyroidism, and diabetes insipidus; and systemic diseases include anaphylaxis, poisoning, heat stroke, burns, trauma, dehydration, and other diseases.

[Food suggestions]
The food recommendation system of the present invention, when the breed of the target animal is input (the breed may be determined from a facial image, as described later), suggests food that is useful for preventing a disease that the target animal is likely to suffer from, based on the insurance claim data for that breed. For example, when the insurance claim data includes data on the insurance claim rate for each disease for each breed, food that is useful for preventing a disease with a high insurance claim rate is suggested. A disease with a high insurance claim rate for a certain breed means that each animal belonging to that breed is also likely to suffer from the same disease, and prevention of that disease is required. For example, when the insurance claim rate for fractures is high for the breed to which the target animal belongs, food that is useful for preventing fractures, such as food with a high calcium content, is suggested to the target animal. The correspondence between the disease and the nutrients and foods that are useful for its prevention may be stored in a database in advance, or information on nutrients that are useful for preventing a certain disease may be obtained by accessing an external server, an academic database, or the Internet.
In another aspect, the food recommendation system of the present invention identifies diseases to which the target animal is susceptible by analyzing insurance claim data, and proposes food that contains nutrients effective against the disease or that has nutrients that cause the disease removed. That is, (A) food that contains nutrients related to or useful for preventing a disease with a high insurance claim rate in a certain breed, or (B) food that does not contain nutrients that are desirable to avoid in order to prevent a disease with a high insurance claim rate in a certain breed, or food that has a low content of the nutrients. For example, if data is obtained from the insurance claim data analysis that indicates a susceptibility to eye diseases, food with added astaxanthin, lutein, tocotrienol, bilberry extract, etc. is proposed. Also, if data is obtained from the insurance claim data that indicates a susceptibility to diabetes, food with reduced sugar is proposed. Note that the food referred to in the present invention is not limited to pet food, but also includes supplements, nutritional drinks, etc.

In addition, nutrition can be adjusted based on individual information such as age, sex, weight, spay/neuter history, genetic test results, and medical history of the animal, in addition to breed. For example, even if the target animal is a breed that is not prone to eye diseases, if the animal has a history of eye diseases, the nutrients can be adjusted to take eye diseases into consideration. Furthermore, if the animal has a history of a certain disease, food can be suggested that reflects the tendency that the animal is more likely to suffer from another specific disease.
Generally, as an animal ages, the diseases to which the animal is susceptible may change. Therefore, the food recommendation system of the present invention uses the age of the target animal to suggest food that is more suitable for the individual animal. In addition, when the insurance claim data includes the mode age for each breed and each disease, that is, the age at which an animal of a certain breed makes the highest number or rate of insurance claims due to a certain disease, in other words, the age at which an animal of a certain breed is most susceptible to a certain disease, it is possible to suggest food with high accuracy that reflects the age by using the age of the target animal and the mode age in the insurance claim data. Preferably, food useful for preventing the disease is suggested before the target animal reaches the same age as the mode age of a disease with a high insurance claim rate for the breed to which the animal belongs. The reason for suggesting the same age as the mode age is that if the target animal reaches the mode age, prevention may not be in time. More preferably, food useful for preventing the disease is suggested around the time the target animal reaches half the mode age.

[output]
The format of the output of the proposal by the proposal means of the present invention is not particularly limited, and for example, a method of outputting a message including the content of the food proposal, such as "Try to get a lot of vitamin B12" or "Try to eat foods with low fat content" on the screen of a terminal such as a personal computer or a smartphone, may be used. Also, specific food ingredients and recipes may be proposed. An example of the case where ingredients are proposed is the output of a message such as "Please feed the patient a diet that contains a lot of pork" or "Please feed the patient a diet that contains a lot of chicken." When a specific recipe is proposed, the most suitable recipe may be picked up and proposed from recipes stored in a database in advance. Also, a configuration may be used in which commercially available foods are stored in a database or the like, and the product name of the most suitable commercially available food is selected from among them and proposed.

Hereinafter, one embodiment of a food recommendation system of the present invention will be described with reference to FIG.
1, terminal 40 is a terminal used by a person (user) who wants to use the food recommendation system, and is connected to server 10 via the Internet or the like. Examples of terminal 40 include a personal computer, a smartphone, and a tablet terminal. Terminal 40 includes a processing unit such as a CPU, a storage unit such as a hard disk, a ROM, or a RAM, a display unit such as a liquid crystal panel, an input unit such as a mouse, a keyboard, or a touch panel, and a communication unit such as a network adapter.
The user accesses the server 10 from the terminal 40 to input and transmit information such as the breed and, if necessary, the age, weight, medical history, etc. of the animal. In addition, the user can receive food suggestions by having the terminal 40 access the server.

In this embodiment, the server 10 is configured by a computer, but may be any device as long as it has the functions according to the present invention.
The storage unit is composed of, for example, a ROM, a RAM, a hard disk, etc. The storage unit stores information processing programs for operating each part of the server, and in particular, software for an insurance claim data analysis means (not shown) and a proposal means 11, data such as insurance claim data 12, etc.

The insurance claim data analysis means is software that receives the breed of the target animal input by the user and outputs diseases that the animal is likely to suffer from depending on the breed, and is executed by the processing calculation unit. The analysis means may be a trained model. Such a trained model may include, for example, XGBoost, CatBoost, LightGBM, or a deep neural network or a convolutional neural network. The analysis means may be configured to output optimal insurance claim data using the age, weight, medical history, etc. of the animal input by the user in addition to the breed of the animal.

The suggestion means 11 is software that receives as input a disease to which the animal is susceptible, determines a food corresponding to the disease, and outputs the software, which is executed by a processing calculation unit. The suggestion means 11 may be a trained model. Such a trained model may include, for example, XGBoost, CatBoost, LightGBM, or a deep neural network or a convolutional neural network.

In this embodiment, the suggestion means and reception means are stored in a server and are connected to the user's terminal via a connection means such as the Internet or a LAN, but the present invention is not limited to this, and may be in a form in which the suggestion means, reception means, and interface unit are stored in a single server or device, or in a form that does not require a terminal 40, etc.

The processing and calculation unit 20 calculates the food recommendation results using the insurance claim data analysis means and recommendation means 11 stored in the memory unit.

The interface unit (communication unit) 30 includes a receiving means 31 and an output means 32, and receives information about the animal's breed and other information from the user's terminal, and outputs and transmits food suggestions to the user's terminal.

<How to suggest food>
The food recommendation method of the present invention is a food recommendation method that suggests food suitable for animals, and involves analyzing insurance claim data to identify diseases and injuries to which the target animal is prone, and suggesting food that contains nutrients effective against the disease, or that has removed nutrients that cause the disease.

Figures 2 to 5 show data extracted from Non-Patent Document 1 on the insurance claim rate, annual medical expenses, and most common age for each disease for toy poodles, chihuahuas, miniature dachshunds, and Shiba Inu (Non-Patent Document 1, pages 20 to 22, 24).

Based on the above data, we constructed software (insurance claim data analysis means) that outputs the diseases that a dog is likely to suffer from when a breed is input (Figure 6). Specifically, when the breed "Toy Poodle" is input, the diseases that a dog is likely to suffer from are "eye diseases" and "digestive diseases." Cataracts (ranked 3rd) and eye discharge (ranked 9th) are listed as eye diseases for "Toy Poodles" (Figure 2), but cataracts pose a greater burden on the dog and in terms of medical expenses, and the most common age for dogs to suffer from cataracts is 10 years old. Therefore, in this embodiment, we have set the specifications to output "eye diseases" as a disease that a dog is likely to suffer from when it reaches half the age (5 years old) of the most common age for dogs to suffer from cataracts.

Next, a list of ingredients effective for each disease was prepared (Figure 7). Based on this list, software (proposed means) was constructed that outputs ingredients to be added to food in response to input of the disease to which the animal is prone (Figure 7). Specifically, when the breed "toy poodle" was input, astaxanthin and lutein, which are effective for eye diseases, and astaxanthin and cranberry powder, which are effective for digestive diseases, were output.

Claims (6)

A food recommendation system comprising a recommendation means for recommending food suitable for a pet based on insurance claim data of the breed and age of the pet,
The insurance claim data includes data on the proportion of insurance claims by disease by breed,
The suggestion means suggests food useful for preventing a disease before the pet reaches an age equal to a mode age of a disease having a high insurance claim rate in a breed to which the pet belongs.
Food suggestion system. 2. The food recommendation system according to claim 1, further comprising a reception means for receiving individual information of the pet animal. The food recommendation system of claim 1 further comprises a determination means for determining the breed of a pet animal from a facial image of the pet animal. The determination means determines the breed of the pet animal from a facial image of the pet animal by using a trained model,
The food recommendation system of claim 3, wherein the trained model is trained using facial images of pet animals and the breed of the pet animals as training data, and the input is the facial image of the pet animal and the output is the breed of the pet animal. 2. The food recommendation system according to claim 1, wherein the recommendation means is configured to recommend the following foods (A) or (B): (A) foods that contain nutrients related to the prevention of diseases with a high insurance claim rate for the variety;
(B) Food that does not contain any nutrients that are desirably avoided in order to prevent diseases associated with a high rate of insurance claims in said breed, or that has low amounts of said nutrients. A food recommendation method for recommending food suitable for a pet animal, the method comprising:
The insurance claim data includes data on the proportion of insurance claims by disease by breed,
The computer suggests food that is useful for preventing a disease before the pet reaches an age equal to a mode age of the disease having a high insurance claim rate in the breed to which the pet belongs.
How to suggest food.