TW201939424A - Method and system for recommending cuisine food according to personal emotion for analyzing the current emotion of a user and recommending a cuisine food complementary to the emotion of the user - Google Patents

Abstract

The present invention provides a method and a system for recommending cuisine food according to personal emotion, with which subjective and objective manners are applied to analyze emotions of a user and food. The subjective manner includes retrieving and analyzing dynamic posts published in a social medium by the user in order to acquire the category of the current emotion of the user. The objective manner includes analyzing the category of emotion of a food image. By using the subjective and objective manners to establish an emotion recipe, a cuisine food that is complementary to the emotion of the user is recommended according to the current feelings condition of the user.

Description

Method and system for recommending food according to personal mood

The present invention relates to a method and system for recommending cooking food, in particular to a method and system for analyzing a user's current emotional state before recommending a recipe complementary to the user's emotion.

In the modern era of Internet information, the types of items in people's lives are becoming more and more abundant. At this stage, various applications have similar product recommendation systems. The recommendation system usually analyzes the source data based on the historical data of the user or the relevant group of users and derives the final recommendation list. The product information in the recommendation list is provided to user.

Take food as an example. Currently, there are a large number of applications for meal recommendation types on the market. Generally, you search for food and distance to find a restaurant that meets the conditions, and then make recommendations based on the people's or blogger's food reviews and ratings. The advantage of this recommendation method is that the food notes with photos and pictures can let consumers understand the restaurant's internal information, menus, dishes, etc., and consumers have better reference materials when choosing.

However, the existing meal recommendation type application software only recommends restaurants based on historical records, and does not make full use of various available technologies to obtain user behavior and response information, and cannot well determine user demand intentions. In addition, under the influence of food safety factors, people are paying more and more attention to healthy eating. As it is impossible to determine the ingredients and cooking process when eating out, more and more people choose to cook by themselves, and the existing cooking recipes are mostly just cooking procedures The steps do not recommend cooking recipes that are complementary to the user ’s negative emotions based on the user ’s current physical and mental conditions (such as under negative emotions).

Therefore, how to provide a suitable cooking recipe according to the physical and mental conditions of the user, while the user cooks himself, effectively soothe his soul and improve the negative effects caused by negative emotions is a problem that needs to be solved at present.

The purpose of this creation is to provide a method of recommending food according to personal emotions. After analyzing the user's current emotional state, he recommends recipes that complement his emotions. While the user is cooking according to the recipe, it can effectively soothe his mind and improve the adverse effects caused by negative emotions.

In order to achieve the above purpose, this creation adopts the following technical means to achieve it. Among them, this creation provides a method for recommending food according to personal emotions, including the following steps: a providing a smart device, a server and a social network Media, the smart device includes an application program, which is in telecommunication connection with the server, the server includes a word segmentation module and a computing module, and the social network media has a plurality of dynamic information. b A user uses the application to log in to the social network media. c The user gives the application a permission, which allows the application to use the server to capture the user's dynamic information on the social network media. d The word segmentation module performs a word segmentation analysis on the dynamic information to form a corresponding one-word word list. e The computing module analyzes the vocabulary by using a text emotion recognition algorithm to obtain an emotion category of the dynamic information. f The computing module uses a recipe emotion recognition algorithm to find a recipe that matches the emotion category. gProvide the recipe to the user.

In an embodiment of the present invention, the text emotion recognition algorithm is shown in the following formula: ; among them, Is one of the words in the glossary, Is the emotion category, Is a function used to calculate The number of times it appears in that emotion category, For all words that appear in that mood category, for In mood category Probability of appearance.

In an embodiment of the present invention, after step e, the method further includes the following steps: e1 stores the analysis result of the text emotion recognition algorithm into a text emotion corpus, and continuously updates the text emotion corpus to increase the accuracy during analysis.

In an embodiment of the present invention, the recipe emotion recognition algorithm analyzes the food image to calculate the emotion of the food image. The analysis factors include the details of Colors, Texture, and Composition. Degree, low depth of field, dynamic range, etc., any combination of the above.

In an embodiment of the present invention, the step f further includes the following steps: f1 stores the calculation result of the recipe emotion recognition algorithm into an emotional recipe corpus, and continuously updates the emotional recipe corpus to increase the accuracy during analysis.

In an embodiment of the present invention, the step g further includes the following steps: g1 The user reports an emotional score of the recipe through the application.

This creation also provides a system for recommending food according to personal emotions, including: a server, a smart device, and a social network media. The server includes a word segmentation module and a computing module. The smart device includes an application program, and the application program is in telecommunication connection with the server. The social network media has a plurality of dynamic information of a plurality of users. A user can use the application to log in to the social network media and authorize the application to use the server to capture the user's dynamic information on the social network media. The server uses the word segmentation mode The group performs a word segmentation analysis on the dynamic information to form a vocabulary, and the server uses the computing module to analyze the vocabulary using a text emotion recognition algorithm to obtain an emotional category of the dynamic information. The server uses The computing module uses a recipe emotion recognition algorithm to find a recipe matching the emotion category, and the server sends the recipe to the application.

In an embodiment of the present invention, the application program further includes a recipe generation module for generating a personalized recipe; the recipe generation module includes a video recording unit and a text editing unit. The video recording unit or the text editing unit creates a personalized recipe. The personalized recipe also includes a custom mood. The application sends the personalized recipe to the server for storage.

In an embodiment of the present invention, the server stores the personalized recipe in a personalized recipe database, and the computing module parses the personalized recipe and calculates it through the recipe emotion recognition algorithm, and then stores the computing result in a Corpus of mood recipes.

In an embodiment of the present invention, the server occasionally captures text information of any recipe through the Internet, and the server uses the word segmentation module to perform word segmentation analysis on the text information of any recipe to form a second Word list, the server analyzes the second word list using the text emotion recognition algorithm to obtain a second emotion category of the dynamic information, and the server stores the analysis result of the text emotion recognition algorithm into a text The emotional corpus, and the text emotional corpus is continuously updated to increase the accuracy of the analysis.

In order to achieve the above-mentioned purpose and effect, the technical means and structure used in this creation, here are some details about the features and functions of the preferred embodiment of this creation. I fully understand it, but it must be noted that these contents are not This constitutes the limitation of the present invention.

Please refer to FIG. 1 and FIG. 2 at the same time, which are a method flowchart and a system block diagram of a preferred embodiment of a method for recommending cooking food according to personal emotions. This creation provides a method of recommending food according to personal emotions, including the following steps:

Step 100: Provide a smart device 1, a server 2 and a social network media 3. The smart device 1 may be a common smartphone, tablet, or notebook, but is not limited thereto. The smart device 1 includes an application 11 which is connected to the server 2 by telecommunications. The application 11 is a cooking recipe application designed according to the theory of emotions. Cooking recipe function. The server 2 is a background server, which includes at least a word segmentation module 21, an operation module 22, and a plurality of databases. The social network media 3 may be Facebook, Twitter, etc. Social media such as Weibo, Weibo or IG (Instagram), the social network media 3 has multiple dynamic information posted by multiple users.

Step 110: A user uses the application 11 to log in to the social network media 3. The application 11 has a community login function and can be connected to social network media 3.

Step 120: The user gives the application 11 a permission to enable the server 2 to capture the user's dynamic information on the social network media 3. The permission enables the application 11 to use the server 2 to capture dynamic information posted by the user on the social network media 3. After logging in the social network media 3, the user can authorize the application 11 to capture the dynamic information of the user's post or comment on the social network media 3 on the same day or recently.

Step 130: Perform a hyphenation analysis on the dynamic information to form a corresponding one-word vocabulary. The word segmentation module 21 utilizes Chinese word segmentation tools commonly used on the Internet for word segmentation analysis, such as the CKIP word segmentation tool and the stuttered word segmentation tool. Hyphenation analysis can perform hyphenation on dynamic information, filter out meaningful words, and generate multiple word lists. Each word list corresponds to different dynamic information, which includes the word, number of words, part of speech, frequency of occurrence, and small icon of the dynamic information.

Step 140: Analyze the word list using a text emotion recognition algorithm to obtain an emotion category of the dynamic information. The arithmetic module 22 analyzes the word list using a text emotion recognition algorithm to obtain an emotional category of the dynamic information. The emotional category includes joy, joy, calm, surprise, like, peace of mind, depression, anger, sadness, Not limited to this. Preferably, the emotion category further includes a score or weight percentage to more accurately represent the attributes of the emotion category. The text sentiment recognition algorithm uses a simple Bayes classifier to construct the classification and avoids the zero probability problem through Laplace Smoothing. Among them, the text emotion recognition algorithm is shown in formula (1): ; (1) where Is one of the words in the glossary, Is the emotion category, Is a function used to calculate The number of times it appears in that emotion category, For all words that appear in that mood category, for In mood category Probability of appearance.

In an embodiment of the present invention, after step 140, step 141 is further included: the analysis result of the text emotion recognition algorithm is stored in a text emotion corpus 23, and the text emotion corpus 23 is continuously updated to increase the accuracy during analysis.

Step 150: Use a recipe emotion recognition algorithm to find a recipe matching the emotion category. After obtaining the current or recent emotion category of the user, the computing module 22 uses the recipe emotion recognition algorithm to perform calculations to find a recipe matching the emotion category.

In an embodiment of the present invention, after step 150, step 151 is further included: the calculation result of the recipe emotion recognition algorithm is stored in an emotional recipe corpus 24, and the emotional recipe corpus 24 is continuously updated to increase the accuracy during analysis. It is worth mentioning that the recipe emotion recognition algorithm can analyze the food image and calculate the emotion score of the food image. The analysis factors include the fineness of Colors, Texture, and Composition. , Low depth of field, dynamic range, etc., Content, etc.

Step 160: Provide the recipe to the user.

In an embodiment of the present invention, after step 160, step 161 is further included: the user feedbacks an emotional score of the recipe through the application 11. The application program 11 may further include a feedback unit. After the user accepts the recipe, the user can give a score to the recipe according to his own feelings. The feedback unit is used to input the emotion score to provide the server 2 with the subsequent calculation of the emotion category. Reference information.

Please refer to FIG. 3 more, which is a system architecture diagram of a method and system for recommending food according to personal emotions. This creation also provides a system for recommending food according to personal emotions, which is applicable to the method for recommending food according to personal emotions. A system for recommending food according to personal emotions includes: a smart device 1, a server 2 and a social network media 3.

Provide the aforementioned server 2, smart device 1, and social network media 3. The server 2 includes a word segmentation module 21 and a computing module 22. The smart device 1 includes an application program 11, and the application program 11 is in telecommunication connection with the server 2. The social network media 3 has a plurality of dynamic information of a user.

Any user can use the application 11 to log in to the social network media 3, and authorize the application 11 to use the server 2 to capture the dynamic information of the user on the social network media 3, the server 2 Use the word segmentation module 21 to perform a word segmentation analysis on the dynamic information to form a word list. The server 2 uses the operation module 22 to analyze the word list using a text emotion recognition algorithm to obtain the dynamic information. The server 2 uses the computing module 22 to find a recipe matching the emotion category by using a recipe emotion recognition algorithm, and the server 2 transmits the recipe to the application 11.

In an embodiment of the present invention, the application program 11 further includes a recipe generation module 110 for generating a personalized recipe; the recipe generation module includes a recording unit and a text editing unit. Recipe videos can be recorded through the recording unit, or a personalized recipe can be made by text input using the text editing unit. The personalized recipe also includes a custom mood, which is a user-defined mood category. The application program 11 sends the personalized recipe to the server 2. The server 2 parses and classifies the personalized recipe, stores the original personalized recipe in a personalized recipe database 25, and the operation module After parsing the personalized recipe and calculating it through the recipe emotion recognition algorithm, the calculation result is stored in an emotional recipe corpus 24.

In an embodiment of the present invention, the server 2 uses a search engine (such as a web crawler system) to irregularly capture text information of any recipe through the Internet. The server 2 uses the word segmentation module 21 to The text information of the arbitrary recipe is segmented to form a second word list, and the server 2 analyzes the second word list using the text emotion recognition algorithm to obtain a second emotion category of the dynamic information and the The server 2 stores the analysis result of the text emotion recognition algorithm in a text emotion corpus 23, and continuously analyzes any recipe on the Internet and updates the text emotion corpus 23 to increase the accuracy during analysis.

In an embodiment of the present invention, the application program 11 further includes a feedback unit. After the user accepts the recipe, he can give the recipe a rating based on his own feelings, input the emotional score through the feedback unit, and provide the server. 2 Reference information for subsequent calculation of emotion categories. The server 2 combines and calculates the emotional score input by the user and the emotional score of the recipe in the original recipe recipe corpus 24 to update the emotional category to which the recipe belongs.

It is worth mentioning that every time a recipe is updated a certain number of times due to the emotional score given by the user, the recipe will be placed in the search schedule. When the search engine grabs the recipe material on the Internet, it will first grab the recipe. Search for related recipe information in your schedule.

To sum up, it can be seen that the present invention provides a method and system for recommending food according to personal emotions, and captures a user's dynamic posts through a server for analysis to obtain the user's current emotion category. The algorithm then calculates and recommends recipes matching the emotion category to the user. While the user is cooking according to the recipe, it can effectively soothe his mind and improve the adverse effects caused by negative emotions.

After the above detailed description, it has been fully shown that this creation has a progressive nature for implementation, and it is a new creation that has not been seen before, which fully complies with the requirements of the invention patent, and has filed an application in accordance with the law. However, the above is only a preferred embodiment of the present invention. When it cannot be used to limit the scope of the implementation of the invention, that is, equal changes and modifications made in accordance with the scope of the invention patent should all fall within the scope of the invention patent. .

1‧‧‧ smart device

11‧‧‧ Apps

2‧‧‧Server

21‧‧‧ Hyphenation Module

22‧‧‧ Computing Module

23‧‧‧Text Emotion Corpus

24‧‧‧ Emotional Recipe Corpus

25‧‧‧ Personalized recipe library

3‧‧‧ Social Network Media

100 ～ 161‧‧‧step

Figure 1 is a flowchart of a method and system for recommending food according to personal emotions; Figure 2 is a system block diagram of a method and system for recommending food according to personal emotions; System architecture diagram of food method and system.

Claims (10)

A method for recommending food according to personal emotions includes the following steps: a providing a smart device, a server, and a social network media, the smart device includes an application program, and the application program is connected with the server by telecommunications, The server includes a word segmentation module and a computing module, the social network media has a plurality of dynamic information; b a user uses the application to log in to the social network media; c the user gives the application a Permission, which allows the application to use the server to capture the user's dynamic information on the social network media; d the word segmentation module performs a word segmentation analysis on the dynamic information to form a corresponding word Table, the word list has a plurality of words; e the calculation module analyzes the word list using a text emotion recognition algorithm to obtain an emotion category of the dynamic information; f the calculation module uses a recipe emotion recognition algorithm Cannot find a recipe that matches the emotion category; g provides the recipe to the user. The method for recommending food according to personal emotions as described in the first patent application scope, wherein the text emotion recognition algorithm is shown in the following formula: ; among them, Is one of the words in the glossary, Is the emotion category, Is a function used to calculate The number of times it appears in that emotion category, For all words that appear in that mood category, for In mood category Probability of appearance. As described in item 1 of the scope of the patent application, the method of recommending food according to personal emotions, wherein the recipe emotion recognition algorithm analyzes the food image to calculate the emotion of the food image. The analysis factors include colors, Texture (Texture), Composition (Composition) meticulous degree, low depth of field, dynamic range, etc., (Content) any combination of the above. As described in item 1 of the scope of patent application, the method for recommending food according to personal emotions, wherein the step e further includes the following steps: e1 stores the analysis result of the text emotion recognition algorithm into a text emotion corpus, and continuously updates the Text sentiment corpus to increase accuracy during analysis. The method for recommending cooking food according to individual emotions as described in the first patent application scope, wherein the step f further includes the following steps: f1 stores the calculation result of the recipe emotion recognition algorithm into an emotional recipe corpus, and continuously updates the A corpus of emotional recipes to increase the accuracy of analysis. As described in item 1 of the scope of patent application, the method for recommending food according to personal emotions, wherein step g further includes the following steps: g1 The user feedbacks an emotional score of the recipe through the application. A system for recommending food according to personal emotions includes: a server including a word segmentation module and a computing module; a smart device including an application program, the application program being connected to the server by telecommunications; a social network Network media with multiple dynamic information for multiple users; where a user can use the application to log in to the social network media and authorize the application to use the server to crawl the user on the social network Dynamic information on the media, the server uses the word segmentation module to perform a word segmentation analysis on the dynamic information to form a word list, and the server uses the computing module to analyze the word using a text emotion recognition algorithm Table to obtain an emotion category of the dynamic information, the server uses the computing module to use a recipe emotion recognition algorithm to find a recipe matching the emotion category, and the server sends the recipe to the application. As described in item 7 of the scope of the patent application, a system for recommending food according to personal emotions, wherein the application program further includes a recipe generation module for generating a personalized recipe; the recipe generation module includes a A video recording unit and a text editing unit. Users can record through the video recording unit or use the text editing unit to make a personalized recipe. The personalized recipe also includes a custom mood. The application sends the personalized recipe to the server. Save. The system for recommending food according to personal emotions as described in item 8 of the scope of patent application, wherein the server stores the personalized recipe into a personalized recipe database, and the computing module parses the personalized recipe and uses the recipe mood After the recognition algorithm has performed the operation, the operation result is stored in an emotional recipe corpus. As described in item 7 of the scope of the patent application, a system for recommending food according to personal emotions, wherein the server occasionally captures text information of any recipe through the Internet, and the server uses the word segmentation module to The text information of the recipe is segmented to form a second vocabulary, and the server analyzes the second vocabulary using the text emotion recognition algorithm to obtain a second emotional category of the dynamic information, and the server will The analysis result of the text emotion recognition algorithm is stored in a text emotion corpus, and the text emotion corpus is continuously updated to increase the accuracy during analysis.