TWI640953B - Cloud recommendation method, computer program product and system for nutrient dosage - Google Patents

Abstract

A cloud recommendation method for nutrient usage, combined with a plurality of user terminals, a cloud database, and a control platform. The cloud database is connected to the management platform and provides a training sample. Each user terminal is installed with an application program. One user terminal loads the application and executes the following steps: sending a body type data to be tested to the control platform to calculate the body type data to be tested Enter into a classification model program. The classification model program uses the results of the nutrient usage of different body type data in the training sample as a score to classify each effective nutrient dosage according to each body type data. Then, the receiving control platform returns a recommendation result that is classified by the classification model program and meets the nutrient dosage of the body type data to be tested.

Description

Cloud recommendation method, computer program product and system for nutrient dosage

The invention relates to an information processing method and system for processing big data by using a cloud system, in particular to a cloud recommendation method and system for nutrient usage based on data of a plurality of user terminals.

Many people are very particular about healthy foods, drugs and drinking water, but busy life is always easy to forget to follow these intakes, it is difficult to persevere, and even ignore this important life details. Sometimes it's hard to know how much you need to get to a healthy standard without hurting your body. In addition, when you have a disease or other special reasons to adjust the dosage, you should also pay attention to how to take these healthy foods.

Therefore, there is a need for an intake that can be recommended by the user through the user's current body state (eg, weight value), and the intake can also be matched according to the user's own habits, such as the customary dosage, the previous special disease or constitution. To adjust the dosage to avoid poisoning or other special accidents.

It is an object of the present invention to provide a cloud recommendation method and system that addresses the lack of nutrient usage in the prior art.

In some implementations, the cloud recommendation method for the nutrient dosage of the present invention is used in conjunction with a plurality of user terminals, a cloud database, and a control platform. The cloud database links the management platform and provides a training sample. Each user terminal is installed with an application, wherein a user terminal loads the application and performs the following steps. The user terminal sends a body shape data to be tested to the control platform to input the body shape data to be input to a classification model program, and the classification model program is scored according to the effect of the nutrient dosage of different body type data in the training sample. This type of data classifies the amount of nutrients that are effective. Then, the user terminal receives the recommended result of the management platform to return a nutrient dosage that matches the body type data after being classified by the classification model program.

In some implementations, one of the user terminals loading the application and executing with the physiological sensor further comprises the steps of: inputting, to the control platform, a plurality of nutrient dosages corresponding to different timings, and timing The index is recorded in a user data corresponding to the cloud database; the physiological sensor is used to capture the plurality of digital physiological parameters of the user after taking the nutrient according to the timing, and the time data is recorded as an index in the cloud data. And corresponding to the user data of the library; and the effect of the management platform inputting the physiological parameters to the classification model program as a score to adjust the effective nutrient dosage corresponding to each body type data.

In some implementations, the classification model is classified using a vector space model, and the classification of the user terminal is evaluated by performing similarity with the training sample for recommendation.

In some implementations, the computer program product of the present invention can perform the method as described above after the computer is loaded into the application and executed.

In some implementations, the cloud recommendation system for the nutrient dosage of the present invention is used in conjunction with a plurality of user terminals and a cloud database.

The cloud recommendation system includes a statistical platform and a control platform. The statistical platform performs statistics and classification on the integrated data of each user terminal collected by the cloud database, a nutrient dosage and a taking effect as a training sample to obtain a classification model program, and the classification model program corresponds to each Body type data classify effective nutrient dosages. The control platform is electrically connected to the statistical platform, and one of the user terminals is queried by using a body shape data to be tested, and the body shape data to be tested is input to the classification model program, thereby generating a nutrient that conforms to the body shape data to be tested. The recommended result of the usage is provided to the user terminal.

In some implementations, each of the user terminals loading the application and executing with the physiological sensor further includes the steps of: receiving, by the user control terminal, the plurality of nutrient dosages corresponding to the different timings of the user terminals And recording, by using the time series as an index, in a user data of each user terminal corresponding to the cloud database; the control platform cooperates with the physiological sensor to capture the user's corresponding time sequence after taking the nutrient The plurality of digitized physiological parameters are recorded in the user data corresponding to the cloud database by using the time series as an index; and the control platform inputs the effects of the physiological parameters into the classification model program as a score to adjust the corresponding body shape. The data is classified as a productive nutrient dosage.

In some implementations, the classification model is classified using a vector space model, and the classification of the user terminal is evaluated by performing similarity with the training sample for recommendation.

The invention has at least the following effects: when the user opens the application to input the weight value, the classification model program can automatically recommend the amount of nutrients that should be ingested, and the control platform gives the preliminary recommended dosage to the user terminal according to the weight value. If you are not satisfied with the recommendation results, you can also feedback the system to adjust the dosage. The system can record the nutrient dosage of different users, in order to recommend the appropriate nutrient dosage.

1‧‧‧User terminal

10‧‧‧Processing module

100‧‧‧Cloud recommendation system

101‧‧‧Application

11‧‧‧Input module

12‧‧‧Memory Module

13‧‧‧Communication module

14‧‧‧Output module

201‧‧‧Classification model program

21‧‧‧Cloud database

22‧‧‧Statistical platform

23‧‧‧Control platform

3‧‧‧Communication network

5‧‧‧ Physiological sensor

S201~S203‧‧‧Steps

S301~S305‧‧‧Steps

Other features and effects of the present invention will be apparent from the following description of the drawings, wherein: Figure 1 is a system block diagram of an embodiment of a nutrient dosage cloud recommendation system of the present invention; A flowchart of a method for executing a control platform of an embodiment; FIG. 3 is a flowchart of a method for executing a user terminal of the embodiment.

Referring to FIG. 1, an embodiment of the cloud recommendation system 100 for nutrient usage of the present invention is used in conjunction with a plurality of user terminals 1 and a cloud database 21. The cloud recommendation system 100 includes a statistic platform 22 and a management platform 23, and the user terminals 1, the cloud database 21, the statistic platform 22, and the management platform 23 are electrically connected to a communication network 3, and the user terminals 1 and the cloud data are The library 21, the statistical platform 22, and the management platform 23 communicate information to each other through the communication network 3.

The statistical platform 22 performs statistics and classification on a training sample according to the integrated data, a nutrient dosage and a performance of each user terminal 1 collected by the cloud database 21 in a training stage to obtain a classification model program 201. Among them, the body type data may be a body mass index (BMI) index, waist circumference, age, special physical factors (high blood pressure, high cholesterol, high blood fat, etc.), weight or height, etc., which are related to describing the state of the body. Nutrients can be various elements required to maintain normal functioning of the body such as sugar, salt, and oil. The effect of administration may be, for example, the effect of taking a certain nutrient amount such that the physiological parameter of the body is a normal value or the physiological parameter is an abnormal value. The training sample may be published from each user terminal 1 on a web platform such as a website or social software.

The classification model program 201 classifies effective nutrient dosages for each body type data. In this embodiment, the classification model program 201 can be (but is not limited to) The vector space model (VSM) is used for classification, and the user terminal 1 is classified by calculating the similarity. The related nouns can refer to the bilingual vocabulary, academic nouns and dictionaries information network of the National Institute of Education. The classification model program 201 finds the effective nutrient dosage of the K user's body type data which is most similar to the user's body type data, and uses the K pen type data to correspond to the recommended value of the effective nutrient dosage.

The control platform 23 is mainly used to obtain the classification model program 201 trained by the statistical platform 22 for subsequent application in an application stage, including: receiving one of the user terminals 1 to perform a query on the body type data to be tested, and inputting the body type data to be tested. To the classification model program 201, a recommendation result for generating a nutrient dosage corresponding to the body type data to be tested is provided to the user terminal 1. For example, when it is known to query the user's weight (or height and other related information), find the daily nutrient usage data of the K users most similar to the size of the querying user, and then the amount of the daily nutrients of the K users can be used. The average value is recommended for the query user as a system recommendation result.

The user terminal 1 has a processing module 10, an input module 11, a memory module 12, a communication module 13, and an output module 14. In this embodiment, the user terminal 1 can be an electronic device that can be used for Internet access, and the processing module 10 carries an application 101 to perform the cloud recommendation method for the nutrient dosage of the present invention. The processing module 10 is electrically connected to the input module 11, the memory module 12, the communication module 13, and the output module 14. The communication module 13 is used to connect the control platform 23 through the communication network 3. Input module 11 for user input application The function of the program 101 is mainly for the user to input the query data to the cloud recommendation system 100, and obtain the data returned by the cloud recommendation system 100 after the classification model program 201 is operated. The memory module 12 records the user input data and the returned data, and the output module 14 allows the user to view the recommendation result returned by the management platform 23.

In order to more objectively correct the recommendation result, each user terminal 1 can also cooperate with a physiological sensor 5 to perform the following steps: the management platform 23 receives the nutrient dosages of the user terminals 1 for inputting multiple times corresponding to different timings, and The time series is recorded in a user profile of each user terminal 1 corresponding to the cloud database 21 . The control platform 23 cooperates with the physiological sensor 5 to capture a plurality of digitized physiological parameters of the user after taking the nutrient corresponding to the time series, and records the time series as an index in each user data corresponding to the cloud database 21 . Then, the control platform 23 inputs the effects of the physiological parameters to the classification model program 201 as a score to adjust the effective nutrient dosages corresponding to the respective body type data.

Referring to Figure 2, and in conjunction with Figure 1, an embodiment of the execution procedure of the control platform 23 of the present invention is illustrated below. First, the management platform 23 receives the nutrient usage of the user terminal 1 for inputting a plurality of times corresponding to different timings, and records them in the cloud database 21 with the time series as an index (step S201). Then, the control platform 23 cooperates with the physiological sensor 5 to capture the plurality of digitized physiological parameters of the user after taking the nutrient corresponding to the time series, and records the time sequence as the index in the user data corresponding to the cloud database 21 (steps) S202); Finally, the effect of inputting the classification model program 201 into the physiological parameters is taken as The score is adjusted to adjust the amount of nutrient used corresponding to each of the body type data (step S203).

For example, the user wants to take the appropriate amount of sugar to control the blood sugar of the body. For example, the user terminal 1 inputs the sugar content of each meal and the blood glucose value measured by the physiological sensor 5 before and after the meal. The sensor 5 can be, for example, a non-invasive blood glucose meter based on metabolic heat and optical sensing, using a measurement technique such as Patent Publication No. EP 1656065 B1 to obtain the user's blood glucose level. Then, it is uploaded to the cloud database 21 as one of the training samples in the training stage, and the statistical platform 22 is provided according to a built-in criterion to delete the abnormal blood sugar value data and the corresponding sugar amount, and retain the normal blood sugar value data and The corresponding sugar dosage is used as a training sample adjustment classification model program 201, which may be referred to Table 1, but not limited thereto.

When the control platform 23 receives the body type data input by the next user for the recommendation query, the classification model program 201 obtains the amount of the sugar component that controls the blood sugar effect as the recommendation result, and gives the user the appropriate amount of sugar intake for the user. Reference Test value. Similar nutrients and physiological parameters, such as the relationship between salt and blood pressure values, can also be applied analogously, and are not limited by the contents described in this embodiment.

Referring to FIG. 3, with reference to FIG. 1, an embodiment of the execution program of the user terminal 1 of the present invention loaded into the application 101 is described as follows. Steps S301-303 belong to the training phase, and steps S304 and S305 belong to the application phase.

During the training phase, the user terminal 1 inputs a plurality of nutrient usages corresponding to different timings to the management platform 23, and records them in the cloud database 21 in time series (step S301). The physiological sensor 5 is used to capture a plurality of digitized physiological parameters after the user takes the nutrient corresponding to the time series, and is recorded in the cloud database 21 with the time series as an index (step S302). The management platform 23 inputs the effects of the physiological parameters to the classification model program 201 as scores to adjust the effective nutrient usages corresponding to the respective body type data (step S303).

In the application phase, the user terminal 1 sends a body shape data to be tested to the control platform 23 to input the body shape data to be tested into the classification model program 201 (step S304), and the classification model program 201 is based on different body types in the training sample. The effectiveness of the nutrient usage of the data was scored to classify the effective nutrient dosage for each of the body type data. Then, the user terminal 1 receives the recommendation result of the management platform 23 to return a nutrient dosage amount that matches the body shape data after classification by the classification model program 201 (step S305).

Take the weight information of the body data system as an example to illustrate, when the control platform After receiving the weight information, the weight information is uploaded to the classification model program 201 to find out the amount of nutrients taken by the user who is currently weighing, and output the preliminary recommendation result. According to the recommendation result, the user will give a satisfactory reply. If the user is satisfied with the result, the classification model program 201 will regard this value as the final recommendation result; if the user is not satisfied with the result, another adjustment can be made for the nutrient dosage, and the input is made. After the nutrient dosage to be adjusted is uploaded to the cloud database 21, the classification model program 201 will refer to the updated data value of the cloud database 21, and when the next user with the approximate body condition needs to be recommended, the classification model program 201 will be based on The adjustment record is used to adjust the recommended nutrient dosage, and a more appropriate recommendation result is given.

In summary, the present invention has at least the following effects: when the user opens the application 101 to input the weight value, the classification model program 201 can automatically recommend the amount of nutrients that should be ingested, and the control platform 23 gives the preliminary recommended amount according to the weight value. Go to user terminal 1. Assuming that the recommendation result is not satisfactory, it is also possible to feed back and record the nutrient dosage of different users, so as to recommend a suitable nutrient dosage for subsequent users, it is indeed possible to achieve the object of the present invention.

However, the above is only the embodiment of the present invention, and the scope of the invention is not limited thereto, and all the equivalent equivalent changes and modifications according to the scope of the patent application and the patent specification of the present invention are still The scope of the invention is covered.

Claims (5)

A cloud recommendation method for nutrient usage, which is used in conjunction with a plurality of user terminals, a cloud database, and a control platform, the cloud database is coupled to the management platform and provides a training sample, and each user terminal is installed with an application, wherein one user terminal Loading the application and cooperating with a physiological sensor and performing the steps of: sending a body shape data to be tested to the control platform to input the body shape data to be input into a classification model program, the classification model program is based on the The effect of the nutrient usage of different body type data in the training sample is used as a score to classify the effective nutrient dosage corresponding to each body type data; receiving the control platform returns a nutrient that matches the body type data after classification by the classification model program The recommended result of the dosage; inputting a plurality of nutrient dosages corresponding to different times at the control platform, and recording the chronological index into a user data corresponding to the cloud database; and using the physiological sensor to capture the The user's multiple digitized physiology after taking nutrients for the time The number is chronologically recorded in the user data corresponding to the cloud database; and the management platform is configured to input the effects of the physiological parameters to the classification model program as a score, and the corresponding data of the body type is adjusted and classified. The effective amount of nutrients used. The cloud recommendation method of the nutrient dosage according to claim 1, wherein the classification model program performs classification by using a vector space model, and the classification of the user terminal is evaluated by performing similarity with the training sample for recommendation. A computer program product, when the computer is loaded into the application and executed, the method as claimed in any one of claims 1 to 2 can be completed. A cloud recommendation system for nutrient usage, which is used in conjunction with a plurality of user terminals and a cloud database, each user terminal loading the application and supporting a physiological sensor, the cloud recommendation system comprising: a statistical platform, according to the cloud data The integrated data collected by the library, the nutrient dosage and the effect of taking a dose are used as a training sample for statistics and classification to obtain a classification model program, and the classification model program is effective for classifying the body type data. a nutrient dosage; and a control platform electrically connected to the statistical platform, receiving one of the user terminals to query a body type data to be tested, and inputting the body type data to be tested into the classification model program, thereby generating a compliance The recommended result of the nutrient dosage of the body type data is provided to the user terminal; the control platform further receives the nutrient dosages of the user terminals input corresponding to different times, and records in the chronological order as the index corresponding to the cloud database. In a user profile of each user terminal; the management platform cooperates The physiological sensor captures a plurality of digitized physiological parameters of the user after taking the nutrient for the time, and records the chronological index into each user data corresponding to the cloud database; and the management platform refers to the classification The model program inputs the effects of the physiological parameters as a score to adjust the amount of effective nutrient used for each of the body type data. The cloud recommendation system for nutrient usage according to claim 4, wherein the classification model program is classified by a vector space model, and the classification of the user terminal is evaluated by performing similarity with the training sample for recommendation.