CN116629916A - Restaurant demand prediction method and system - Google Patents
Restaurant demand prediction method and system Download PDFInfo
- Publication number
- CN116629916A CN116629916A CN202310560047.2A CN202310560047A CN116629916A CN 116629916 A CN116629916 A CN 116629916A CN 202310560047 A CN202310560047 A CN 202310560047A CN 116629916 A CN116629916 A CN 116629916A
- Authority
- CN
- China
- Prior art keywords
- restaurants
- batch
- recommended
- food
- dishes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 235000013305 food Nutrition 0.000 claims abstract description 239
- 239000000463 material Substances 0.000 claims abstract description 191
- 235000021186 dishes Nutrition 0.000 claims description 114
- 238000005516 engineering process Methods 0.000 claims description 9
- 235000013311 vegetables Nutrition 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 8
- 235000015067 sauces Nutrition 0.000 description 8
- 241000251468 Actinopterygii Species 0.000 description 7
- 241000272525 Anas platyrhynchos Species 0.000 description 7
- 235000015277 pork Nutrition 0.000 description 7
- 235000014347 soups Nutrition 0.000 description 5
- 230000011218 segmentation Effects 0.000 description 4
- 241000287828 Gallus gallus Species 0.000 description 3
- 239000004615 ingredient Substances 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 240000008067 Cucumis sativus Species 0.000 description 2
- 235000010799 Cucumis sativus var sativus Nutrition 0.000 description 2
- 244000000626 Daucus carota Species 0.000 description 2
- 235000002767 Daucus carota Nutrition 0.000 description 2
- 235000013409 condiments Nutrition 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 235000001674 Agaricus brunnescens Nutrition 0.000 description 1
- 235000006667 Aleurites moluccana Nutrition 0.000 description 1
- 235000002568 Capsicum frutescens Nutrition 0.000 description 1
- 240000004957 Castanea mollissima Species 0.000 description 1
- 235000018244 Castanea mollissima Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 244000061458 Solanum melongena Species 0.000 description 1
- 235000002597 Solanum melongena Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000004992 fission Effects 0.000 description 1
- 235000013312 flour Nutrition 0.000 description 1
- 235000015219 food category Nutrition 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 235000020997 lean meat Nutrition 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 235000013372 meat Nutrition 0.000 description 1
- 235000012149 noodles Nutrition 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/12—Hotels or restaurants
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Strategic Management (AREA)
- Computational Linguistics (AREA)
- Development Economics (AREA)
- Finance (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Tourism & Hospitality (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Databases & Information Systems (AREA)
- Marketing (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Animal Behavior & Ethology (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Human Resources & Organizations (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a restaurant demand prediction method and a restaurant demand prediction system, wherein the restaurant demand prediction method comprises the following steps: acquiring the food types of recommended dishes according to the names of the dishes in the recommended menu information of the restaurant, and the relative usage amount of each food; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales volume of the recommended dishes and the relative usage amount of each food material; determining a plurality of restaurants of the same type as the restaurants, and predicting predicted demand data of food materials in the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the known recommended dishes of the first batch of restaurants and the network comment data of the food materials; predicting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the relative duty ratio of each food material and the predicted demand data of the food materials in the recommended dishes of the second batch of restaurants; finding restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants. The application can rapidly develop B-end clients in the area.
Description
Technical Field
The application relates to the technical field of computers, in particular to a restaurant demand prediction method and a restaurant demand prediction system.
Background
Traditionally, food suppliers have passively accepted orders or recommended their goods by individually accessing enterprise-side customers to achieve more sales. Along with the coming of big data age, food material suppliers excavate the information of enterprise end customer through the open sea data, obtain the cuisine and the welcome degree that this B end customer sold the dish, estimate out the kind and the quantity of the required food material of this B end customer to make things convenient for the waiter to go on the gate and carry out accurate promotion. The B-end clients refer to enterprise clients with food material requirements around BU Store, and the units have two characteristics, namely, large purchase quantity and durable and stable purchase power. For example, restaurants are a typical class of B-side customers. The BU Store refers to a Store using the product, such as a supermarket, a sales center, etc., and can be extended to any sales unit providing and selling B2B services such as food, condiments, etc., for customers.
The prior art mainly acquires basic information of enterprise clients in an area by manually searching open sea data, including comment and comment numbers, scores, recommended dishes, praise times and the like, and predicts food material requirements of B-end clients by means of manual experience analysis data, but the prior art cannot acquire accurate basic information of potential clients through massive data.
Disclosure of Invention
The application mainly aims to provide a restaurant demand prediction method and a restaurant demand prediction system, which are used for solving the problem that basic information of potential customers cannot be accurately acquired in the prior art.
According to one aspect of the present application, a method for predicting restaurant demand is provided, comprising: acquiring information of restaurants in a preset area, wherein the information of the restaurants at least comprises the following steps: restaurant type, recommended menu information, recommended dish sales information; acquiring the food type corresponding to the recommended dishes and the relative usage amount information of each food according to the names of the dishes in the recommended menu information of the restaurant; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the relative usage information of each food material; determining a plurality of restaurants of the same type as the restaurants, wherein the plurality of restaurants comprises a first batch of restaurants for which actual demand data of food materials of at least one recommended dish is known and a second batch of restaurants for which actual demand data of food materials of the recommended dish is unknown; predicting predicted demand data of food materials in the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the known recommended dishes of the first batch of restaurants and the network comment data of the food materials; predicting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the relative duty information of each food material and the predicted demand data of the food materials in the recommended dishes of the second batch of restaurants; searching the restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
Wherein the method further comprises: determining the menu type of the restaurant according to the recommended menu information of the restaurant; searching restaurants similar to the second-batch restaurants according to the menu type, the food material type and the food material forecast demand data of the recommended dishes of the second-batch restaurants.
Wherein the method further comprises: and constructing a knowledge graph according to the restaurant information, recommended dishes and food types of the second batch of restaurants, and acquiring food prediction demand data of the restaurants of the same type as the second batch of restaurants through the knowledge graph.
Wherein the method further comprises: constructing crossed tree-type features based on text similarity and text clustering technology according to the interrelation among the dishes, the dishes and the food materials of the second batch of restaurants, so as to calculate the approximation degree of the client and the potential B-end client; the node embedding algorithm based on graph theory takes the B-end client and the vegetable, dishes and food materials thereof as nodes, builds a knowledge graph, expresses the B-end client as a vector by utilizing the node embedding algorithm, and thus calculates the approximation degree of the B-end client and the potential B-end client.
Wherein the method further comprises: acquiring the actual demand data of the food materials of the second batch of restaurants, and correcting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the second batch of restaurants.
There is also provided in accordance with another aspect of the present application a restaurant demand prediction system, comprising: a restaurant information obtaining module, configured to obtain information of a restaurant in a predetermined area, where the information of the restaurant at least includes: restaurant type, recommended menu information, recommended dish sales information; the food material occupation ratio determining module is used for acquiring the type of the food material corresponding to the recommended dishes and the relative usage amount information of each food material according to the names of the dishes in the recommended menu information of the restaurant; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the relative usage information of each food material; a first food material forecast demand data forecast module for determining a plurality of restaurants of a same type as the restaurants, wherein the plurality of restaurants includes a first set of restaurants for which actual demand data for food materials of at least one recommended dish is known and a second set of restaurants for which actual demand data for food materials of the recommended dish is unknown; predicting predicted demand data of at least one food in the recommended dishes of the second batch of restaurants according to the actual demand data of the food of the known recommended dishes of the first batch of restaurants and the network comment data of the food; the second food material forecast demand data forecast module is used for forecasting forecast demand data of other food materials in the recommended dishes of the second batch of restaurants according to the relative duty ratio data of each food material in the recommended dishes of the restaurants and the forecast demand data of at least one food material in the recommended dishes of the second batch of restaurants; the searching module is used for searching the restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
Wherein, the look-up module is further configured to: determining the menu type of the restaurant according to the recommended menu information of the restaurant; searching restaurants similar to the second-batch restaurants according to the menu type, the food material type and the food material forecast demand data of the recommended dishes of the second-batch restaurants.
Wherein the system further comprises: and the third food material prediction demand data prediction module is used for constructing a knowledge graph according to the restaurant information, recommended dishes and food material types of the second batch of restaurants, and acquiring the food material prediction demand data of the restaurants of the same type as the second batch of restaurants through the knowledge graph.
The third food material prediction demand data prediction module is further configured to: constructing crossed tree-type features based on text similarity and text clustering technology according to the interrelation among the dishes, the dishes and the food materials of the second batch of restaurants, so as to calculate the approximation degree of the client and the potential B-end client; the node embedding algorithm based on graph theory takes the B-end client and the vegetable, dishes and food materials thereof as nodes, builds a knowledge graph, expresses the B-end client as a vector by utilizing the node embedding algorithm, and thus calculates the approximation degree of the B-end client and the potential B-end client.
The second food material prediction demand data prediction module is further configured to: acquiring the actual demand data of the food materials of the second batch of restaurants, and correcting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the second batch of restaurants.
According to the embodiment of the application, the open sea data of shops around the BU Store are acquired and analyzed through a technical method, an AI model is constructed, and the requirements of potential B-end customers after the information is mined. Meanwhile, the real data obtained through investigation or actual transaction is used for fitting and iterating the demand model, so that the food material demands of the clients are more accurately analyzed. And finally, combining the basic information of the existing B-end client of the store and the predicted food demand information, arranging the accurate marketing of the salesmen to go on the gate, and realizing the aim of rapidly expanding the B-end client in the area.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a flow chart of a method of predicting restaurant demand in accordance with an embodiment of the present application;
FIG. 2 is a schematic diagram of a restaurant information link according to an embodiment of the present application;
FIGS. 3A and 3B are schematic diagrams of relationship map construction according to embodiments of the present application;
fig. 4 is a schematic diagram of food material parsing according to menu information according to an embodiment of the present application;
FIG. 5 is a schematic diagram of a search implemented using an LSH algorithm in accordance with an embodiment of the application;
FIGS. 6A and 6B are schematic diagrams of a method for performing a dish search using a similarity algorithm according to an embodiment of the present application;
FIG. 7 is a schematic diagram of identifying comment results using NER techniques provided in accordance with an embodiment of the present application;
FIG. 8 is a schematic diagram of store similarity provided in accordance with an embodiment of the present application;
FIG. 9 is a block diagram of a restaurant demand prediction system according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The following describes in detail the technical solutions provided by the embodiments of the present application with reference to the accompanying drawings.
FIG. 1 is a flowchart of a method of predicting restaurant demand according to an embodiment of the present application, as shown in FIG. 1, the method comprising the steps of:
step S102, acquiring information of restaurants in a preset area, wherein the information of the restaurants at least comprises the following steps: restaurant type, recommended menu information, recommended dish sales information.
In the embodiment of the application, information of one or more enterprise clients (B-end clients) in a preset area can be acquired in a multi-platform and multi-channel mode, wherein the B-end clients refer to enterprise clients with food material requirements around BU (Business Unit) Store, and the units have two characteristics, namely, large purchase amount and durable and stable purchase force. The BU Store may comprise supermarkets, sales centers, etc., and can be extended to any sales unit that provides and sells B2B services for food, condiments, etc. to customers. Among these, restaurants are a typical class of B-side customers.
In the embodiment of the application, taking a restaurant as an example, the same restaurant with different platforms and different channels can be linked through an entity linking technology in a knowledge graph, and various information for establishing a sound restaurant can include, but is not limited to: name, address, phone, recipe (type), score, comment and recommend dishes, menu (recipe), etc. The Knowledge Graph (knowledgegraph) is a Knowledge base that integrates data using a data model or topology of a Graph structure. In practical applications, the related information of the recommended dishes of the restaurant can be obtained through public data (such as a consumption comment website, a takeaway website and the like). As shown in fig. 2, the information of the restaurant a in an area is obtained through the platform a, the platform B and the platform C, and the same restaurant a in the platform a, the platform B and the platform C are linked together through the entity linking technology in the knowledge graph, so that the information of the restaurant a is summarized in data, and the data is high in quality and fully utilized.
Step S104, obtaining the food material types corresponding to the recommended dishes and the relative usage information of each food material according to the names of the dishes in the recommended dishes menu information of the restaurant, and obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the usage information of each food material.
Firstly, establishing an association relation between a dish name of a recommended dish and food materials used by the dish through the modes of open sea data mining, data processing, manual maintenance and the like, for example, establishing a relation map between the recommended dish name of a restaurant A and the food materials used. The open sea data refers to public data from public platforms, related websites and the like, such as consumption criticizing websites, takeaway websites and the like. As shown in fig. 3A and 3B, the (required) food categories corresponding to the recommended dish name "special cold noodles" of restaurant a include "carrot", "flour", "cucumber/cucumber", "bean sprouts/sprouts"; the recommended vegetable names of the restaurant A are that the food material types corresponding to the fish-flavored shredded pork include carrot, lean meat, mushroom and other staple foods. In addition, the association between the recommended dish name and the food used in the dish includes not only the type of food but also information on the amount of use or the relative amount of use of the food.
Then, the dish names of each recommended dish sold by the restaurant A can be obtained according to the obtained recommended menu information of the restaurant A, and the food type and the food usage information corresponding to the recommended dish are searched in the association relation between the recommended dish names and the food according to the recommended dish names, so that the usage information of each food required by the recommended dish of the restaurant A can be obtained. Wherein restaurant a's menu information may be obtained through open sea data, an external recipe website, or other means.
For each dish in the recommended dish menu information of restaurant A, the food materials it uses in the recipe can be searched for by dish name matching. Specifically, referring to fig. 4, recipe data is obtained from a recipe website, and for each dish in the enterprise-side customer menu, a search is performed using LSH (locality-sensitive hashing, locality sensitive hashing) algorithm, recalling a plurality of similar dishes. As shown in fig. 5, the "recipe" data includes shredded pork with a fish sauce, diced chicken with a chicken sauce and pot-back meat, and when the shredded pork with a fish sauce is selected, the shredded pork with a fish sauce and the dish name are searched by using an LSH (local sensitivity-hash) algorithm, so that similar dishes including shredded pork with a fish sauce, shredded pork with a chicken sauce and "shredded pork with a fish sauce" and "shredded eggplant with a fish sauce are obtained.
And then, word segmentation processing is carried out on the names of the dishes according to food materials, and removal of stop words is carried out. And then weighting the food materials by using a similarity algorithm, and searching out the dishes most similar to the target dishes in the food material structure according to the distinguishing degree of the food materials to the dishes and giving different weights to the word segmentation. Preferably, the similarity algorithm is a cosine similarity algorithm (Weighted Cosine Similarity).
Referring to fig. 6A, first, the stewed duck soup is first subjected to word segmentation in summer to obtain stewed duck, soup, summer and necessary, then the stop words are removed to obtain stewed duck, duck and duck soup, the food materials are weighted by using a similarity algorithm, weights are given to the word segmentation, and dishes which are most similar to the target dishes in the food material structure, such as Chinese chestnut roast duck, stewed mutton soup, duck soup, chopped chili fish heads and the like, are searched.
And labeling the similarity, taking the similarity as a feature training model, and obtaining a confidence level so as to measure the accuracy of the search result. In this embodiment, whether the similarity is similar is manually marked, and the similarity is used as a feature training model, so as to obtain Confidence Level (Confidence Level), that is, a measure of the accuracy of the search result. The confidence level can automatically locate data with inaccurate matching, and the lower confidence level can also warn the salesmen to remind them of proper manual correction to improve the accuracy.
For the situation that one dish is more than one way, considering that customers finish dining in different restaurants to post comments about the dish, comment data of the restaurants are used, the NER (Named-entity recognition, named entity identification) technology is used for identifying the dishes and food materials in the comments, so that the specific way of selling the dishes in a certain restaurant is found, and the result after NER is carried out on the comment data is shown in fig. 7. Wherein, the English label with the suffix-DIS indicates that the Chinese of the corresponding position is the selling dish for which the comment is aimed, and the English label with the suffix-ING indicates that the Chinese of the corresponding position is the food material for which the comment is pointed out. Preferably, the algorithm employed by the NER technique is a converter-based machine learning algorithm for NLP (Natural language processing ) pre-training, the model of which is BERT (bi-directional encoder representation using a transducer).
Step S106, determining a plurality of restaurants of the same type as the restaurants, wherein the plurality of restaurants includes a first batch of restaurants known with actual demand data of food materials of at least one recommended dish and a second batch of restaurants known with actual demand data of food materials of the recommended dish.
A plurality of restaurants of the same type as the restaurant a, among which restaurants of the same type as the restaurant a, including restaurants of which actual demand for food of one or more recommended dishes is known (first-batch restaurants), and restaurants of which actual demand for food of any recommended dish is unknown (second-batch restaurants), are determined according to the type of the restaurant.
According to the embodiment of the application, the actual demand data of the food materials of at least one recommended dish of the restaurant can be obtained through network comment data such as different open sea platforms, actual transaction orders and the like, and in other embodiments, the actual demand data of the food materials of at least one recommended dish of the restaurant can also be obtained through channels such as online or offline user investigation data.
Step S108, predicting predicted demand data of at least one food in the recommended dishes of the second batch of restaurants according to the actual demand data of the food of the known recommended dishes of the first batch of restaurants and the network comment data of the food.
According to an embodiment of the application, a representative CBID (Candidate Benchmark for Ingredients Demand, food demand candidate benchmark) is selected in the second restaurant batch. The feature of selecting the CBID reference is that the relative demand for a certain food material or materials is highest in the same type of restaurant, or the feature of the CBID for a certain food material or materials is that the actual demand for that food material is higher than the average level of the actual demands for other food materials in the same type of restaurant. And then, calculating the actual demand of the restaurant food according to the CBID by adopting the following two algorithms.
For each food material, the restaurants are ordered in descending order according to the actual demand of CBID food materials, K largest restaurants are selected according to the order, and the food material demand is predicted according to a formula 1 and a formula 2, and the algorithm can be called as a Max-K Sampling for CBID algorithm in the application.
p m',j' ＝r m c j (2)
Wherein restaurants are divided into n groups of similar restaurants using a similarity algorithm. Within each group:
a m,j representing the actual demand of food m in restaurant j;
c j the comment number of restaurant j;
arrangement a from high to low m,j ，a m,j(i) The actual demand of the ith restaurant representing the highest actual demand; w (w) i The weight of i is represented, and the higher the ranking, the higher the weight.
r m Representing the ratio of the actual demand of the food material m to the number of comments;
p m',j' representing the predicted potential demand for food m 'in restaurant j'.
And (II) predicting the food material demand of the restaurant according to formulas 3, 4 and 5. Only restaurants with relatively high and stable demands are selected as CBIDs in the algorithm, so the algorithm can be called Average Pooling for CBID algorithm.
p m',j' ＝r m c j (5)
Stores are divided into n groups of similar restaurants using a similarity algorithm. The type of restaurants in each group is the same, and:
o represents the number of old customer restaurants;
a m,j representing the actual demand of restaurant j on food m;
c j the comment number of restaurant j;
avg m representing the average actual demand of the food material m;
r m representing the ratio of the actual demand of the food material m to the number of comments;
p m,j representing the predicted potential demand for food m in old customer restaurant j.
The actual requirements of one or more restaurants on one food material or one or more restaurants on various food materials can be calculated through the two algorithms. In addition, the GBDT algorithm can be used as an alternative algorithm for predicting food material prediction demand data. GBDT is an iterative decision tree algorithm consisting of a number of decision trees, all of which conclusions are accumulated to make the final answer. The GBDT algorithm takes the food material requirements or purchase information of the restaurant with real data as input characteristics, and trains a model for predicting the types and the quantity of the food material requirements. The more data the restaurant has real data, the higher the accuracy of the predictive model.
Step S110, predicting predicted demand data of other food materials in the recommended dishes of the second batch of restaurants according to the relative duty data of each food material and the predicted demand data of at least one food material in the recommended dishes of the second batch of restaurants.
In the present application, since the food material ratio information of the restaurants of the same type is the same or substantially the same, the relative ratio data of each food material in the recommended dishes of the restaurants obtained in step S104 is used as the relative ratio information of each food material in the recommended dishes of the second batch of restaurants. And (5) predicting the predicted demand information of other food materials in the recommended dishes of the second batch of restaurants by combining the predicted demand data of at least one food material in the recommended dishes of the second batch of restaurants obtained in the step S108. Thus, the predicted demand information of each food material in the recommended dishes of the second restaurant is obtained.
Step S112, finding restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
Referring to fig. 8, in the embodiment of the present application, a knowledge graph is constructed with shops, recommended dishes and food materials as entities and information of classifications and amounts of shops, recommended dishes and food materials as attributes, so that searching, querying and reasoning can be implemented using a graph knowledge base.
And searching all clients similar to the client at the B end by using a similarity algorithm, and realizing fission type expansion. When Store A becomes a B-end client of a BU Store, the BU Store calculates the stores most similar to Store A through a Store similarity algorithm, obtains the food material requirements of the similar stores through a knowledge graph, and expands the food material requirements into a new B-end client of the BU Store after sales promotion by a salesman.
Specifically, the calculation of store similarity involves two main algorithms:
1. tree Features Based on Text (text-based tree feature):
considering the relation of interconnection among cuisine (cuisine), recommended Dishes (recommended Dishes) and food materials (Ingredients), and constructing crossed tree-type features based on text similarity and text clustering technology, thereby calculating the similarity between restaurants.
2. Node Embedding Based on Graph (graph theory based node embedding):
the method comprises the steps of constructing a knowledge graph by taking a restaurant (restaurant), a menu (Cusine), dishes (recommended Dishes) and food materials (Ingredients) as nodes, and expressing the restaurant as a vector by using a Node Embedding algorithm so as to calculate the similarity between the restaurants.
By matching the requirements of the dominant product of BUstore and the B-end client, the knowledge graph can quickly find all shops needing the food by taking the food as an entrance, so that the B-end client can find the B-end client with the requirements most conforming to the dominant product from the requirement matching angle, thereby achieving the purpose of quickly expanding users.
Referring to fig. 9, there is also provided a restaurant demand prediction system according to an embodiment of the present application, including:
a restaurant information acquiring module 91, configured to acquire information of restaurants in a predetermined area, where the information of restaurants includes at least: restaurant type, recommended menu information, recommended dish sales information;
the food material occupation ratio determining module 92 is configured to obtain a type of food material corresponding to the recommended dishes and information of a relative usage amount of each food material according to a name of the dish in the recommended menu information of the restaurant; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the relative usage information of each food material;
a first food material forecast demand data forecast module 93 for determining a plurality of restaurants of the same type as the restaurant, wherein the plurality of restaurants includes a first set of restaurants for which actual demand data for food materials of at least one recommended dish is known and a second set of restaurants for which actual demand data for food materials of the recommended dish is unknown; predicting predicted demand data of at least one food in the recommended dishes of the second batch of restaurants according to the actual demand data of the food of the known recommended dishes of the first batch of restaurants and the network comment data of the food;
a second food material forecast demand data forecast module 94, configured to forecast demand data for other food materials in the recommended dishes of the second batch of restaurants based on the relative duty data of each food material in the recommended dishes of the restaurants and the forecast demand data for at least one food material in the recommended dishes of the second batch of restaurants;
the searching module 95 is configured to search for restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
Further, the searching module 95 is further configured to:
determining the menu type of the restaurant according to the recommended menu information of the restaurant;
searching restaurants similar to the second-batch restaurants according to the menu type, the food material type and the food material forecast demand data of the recommended dishes of the second-batch restaurants.
Further, the prediction system further includes:
and a third food material forecast demand data forecast module (not shown) configured to construct a knowledge graph according to the restaurant information, recommended dishes, and food material types of the second batch of restaurants, and obtain food material forecast demand data of the same type of restaurants as the second batch of restaurants through the knowledge graph.
Further, the third food material prediction demand data prediction module is further configured to:
constructing crossed tree-type features based on text similarity and text clustering technology according to the interrelation among the dishes, the dishes and the food materials of the second batch of restaurants, so as to calculate the approximation degree of the client and the potential B-end client;
the node embedding algorithm based on graph theory takes the B-end client and the vegetable, dishes and food materials thereof as nodes, builds a knowledge graph, expresses the B-end client as a vector by utilizing the node embedding algorithm, and thus calculates the approximation degree of the B-end client and the potential B-end client.
Further, the second food material forecast demand data forecast module 94 is further configured to: acquiring the actual demand data of the food materials of the second batch of restaurants, and correcting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the second batch of restaurants.
The operation steps of the method of the application correspond to the structural features of the system, and can be referred to each other, and will not be described in detail.
In summary, the embodiment of the application firstly acquires restaurant information of clients at the enterprise end in the area by adopting a multi-platform multi-channel mode, links the same restaurant with different platforms and different channels, complements the restaurant information, has high data quality and is fully utilized; then analyzing food materials required by the restaurant through the recommended dishes, constructing a relation map, acquiring menu information of the enterprise client in the open sea data, and analyzing the food materials used by each dish in the menu information, so as to accurately analyze the food materials required by the restaurant; then, a demand pre-estimating algorithm is adopted to predict the food demand type and the food usage amount of the enterprise client, so as to obtain the predicted food demand amount; and finally, correcting the predicted values of the food material demand types and the food material usage amounts of the clients at the enterprise end by using real data, fitting and iterating the demand model, so that the accuracy of a food material demand prediction algorithm can be improved in a self-iterating way, and the accuracy is higher as the real data are more.
Although the present disclosure has been described in detail with reference to particular embodiments thereof, those skilled in the art will appreciate that various changes and modifications can be made therein without departing from the spirit and scope of the embodiments. It is therefore intended that the present application cover the modifications and variations of this application provided they come within the spirit and scope of the appended claims and their equivalents.
Furthermore, the features disclosed in the foregoing description, or the following claims, or the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for attaining the disclosed result, as appropriate, may, separately, or in any combination of such features, be utilised for realising the application in diverse forms thereof. In particular, one or more features of any one of the embodiments described herein may be combined with one or more features of any other of the embodiments described herein.
Protection may also be sought for any feature disclosed in any one or more of the publications cited in connection with the present application and/or incorporated by reference.
Claims (10)
1. A method for predicting restaurant demand, comprising:
acquiring information of restaurants in a preset area, wherein the information of the restaurants at least comprises the following steps: restaurant type, recommended menu information, recommended dish sales information;
acquiring the food type corresponding to the recommended dishes and the relative usage amount information of each food according to the names of the dishes in the recommended menu information of the restaurant; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the relative usage information of each food material;
determining a plurality of restaurants of the same type as the restaurants, wherein the plurality of restaurants comprises a first batch of restaurants for which actual demand data of food materials of at least one recommended dish is known and a second batch of restaurants for which actual demand data of food materials of the recommended dish is unknown;
predicting predicted demand data of at least one food in the recommended dishes of the second batch of restaurants according to the actual demand data of the food of the known recommended dishes of the first batch of restaurants and the network comment data of the food;
predicting predicted demand data of other food materials in the recommended dishes of the second batch of restaurants according to the relative duty data of each food material in the recommended dishes of the restaurants and the predicted demand data of at least one food material in the recommended dishes of the second batch of restaurants;
searching the restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
2. The method as recited in claim 1, further comprising:
determining the menu type of the restaurant according to the recommended menu information of the restaurant;
searching restaurants similar to the second-batch restaurants according to the menu type, the food material type and the food material forecast demand data of the recommended dishes of the second-batch restaurants.
3. The method as recited in claim 2, further comprising:
and constructing a knowledge graph according to the restaurant information, recommended dishes and food types of the second batch of restaurants, and acquiring food prediction demand data of the restaurants of the same type as the second batch of restaurants through the knowledge graph.
4. A method according to claim 3, further comprising:
constructing crossed tree-type features based on text similarity and text clustering technology according to the interrelation among the dishes, the dishes and the food materials of the second batch of restaurants, so as to calculate the approximation degree of the client and the potential B-end client;
the node embedding algorithm based on graph theory takes the B-end client and the vegetable, dishes and food materials thereof as nodes, builds a knowledge graph, expresses the B-end client as a vector by utilizing the node embedding algorithm, and thus calculates the approximation degree of the B-end client and the potential B-end client.
5. The method as recited in claim 1, further comprising:
acquiring the actual demand data of the food materials of the second batch of restaurants, and correcting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the second batch of restaurants.
6. A restaurant demand prediction system, comprising:
a restaurant information obtaining module, configured to obtain information of a restaurant in a predetermined area, where the information of the restaurant at least includes: restaurant type, recommended menu information, recommended dish sales information;
the food material occupation ratio determining module is used for acquiring the type of the food material corresponding to the recommended dishes and the relative usage amount information of each food material according to the names of the dishes in the recommended menu information of the restaurant; obtaining the relative duty ratio data of each food material in the recommended dishes of the restaurant according to the sales information of the recommended dishes and the relative usage information of each food material;
a first food material forecast demand data forecast module for determining a plurality of restaurants of a same type as the restaurants, wherein the plurality of restaurants includes a first set of restaurants for which actual demand data for food materials of at least one recommended dish is known and a second set of restaurants for which actual demand data for food materials of the recommended dish is unknown; predicting predicted demand data of at least one food in the recommended dishes of the second batch of restaurants according to the actual demand data of the food of the known recommended dishes of the first batch of restaurants and the network comment data of the food;
the second food material forecast demand data forecast module is used for forecasting forecast demand data of other food materials in the recommended dishes of the second batch of restaurants according to the relative duty ratio data of each food material in the recommended dishes of the restaurants and the forecast demand data of at least one food material in the recommended dishes of the second batch of restaurants;
the searching module is used for searching the restaurants of the same type as the second batch of restaurants according to the food material types and the food material forecast demand data of the recommended dishes of the second batch of restaurants.
7. The system of claim 6, wherein the lookup module is further configured to:
determining the menu type of the restaurant according to the recommended menu information of the restaurant;
searching restaurants similar to the second-batch restaurants according to the menu type, the food material type and the food material forecast demand data of the recommended dishes of the second-batch restaurants.
8. The system of claim 7, further comprising:
and the third food material prediction demand data prediction module is used for constructing a knowledge graph according to the restaurant information, recommended dishes and food material types of the second batch of restaurants, and acquiring the food material prediction demand data of the restaurants of the same type as the second batch of restaurants through the knowledge graph.
9. The system of claim 8, wherein the third food material forecast demand data forecast module is further configured to:
constructing crossed tree-type features based on text similarity and text clustering technology according to the interrelation among the dishes, the dishes and the food materials of the second batch of restaurants, so as to calculate the approximation degree of the client and the potential B-end client;
the node embedding algorithm based on graph theory takes the B-end client and the vegetable, dishes and food materials thereof as nodes, builds a knowledge graph, expresses the B-end client as a vector by utilizing the node embedding algorithm, and thus calculates the approximation degree of the B-end client and the potential B-end client.
10. The system of claim 6, wherein the second food material forecast demand data forecast module is further configured to: acquiring the actual demand data of the food materials of the second batch of restaurants, and correcting the predicted demand data of the food materials of the recommended dishes of the second batch of restaurants according to the actual demand data of the food materials of the second batch of restaurants.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310560047.2A CN116629916A (en) | 2023-05-18 | 2023-05-18 | Restaurant demand prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310560047.2A CN116629916A (en) | 2023-05-18 | 2023-05-18 | Restaurant demand prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116629916A true CN116629916A (en) | 2023-08-22 |
Family
ID=87601894
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310560047.2A Pending CN116629916A (en) | 2023-05-18 | 2023-05-18 | Restaurant demand prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116629916A (en) |
-
2023
- 2023-05-18 CN CN202310560047.2A patent/CN116629916A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5962926B2 (en) | Recommender system, recommendation method, and program | |
| CA2655196C (en) | System and method for generating a display of tags | |
| Mahajan et al. | Parameter estimation in marketing models in the presence of multicollinearity: An application of ridge regression | |
| CN112182412A (en) | Method, computing device, and computer storage medium for recommending physical examination items | |
| CN109389442A (en) | Method of Commodity Recommendation and device, storage medium and electric terminal | |
| CN111400507B (en) | Entity matching method and device | |
| CN112200601B (en) | Item recommendation method, device and readable storage medium | |
| Scholz et al. | Using PageRank for non-personalized default rankings in dynamic markets | |
| KR20170129529A (en) | Method and device for predicting future number of customers coming into the store based on pattern information of floating population | |
| CN110750697B (en) | Merchant classification method, device, equipment and storage medium | |
| CN110851571B (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
| US20210272178A1 (en) | Determining item relevancy | |
| CN112711713B (en) | Interest point recommendation and display method and device, computer equipment and storage medium | |
| KR20210032691A (en) | Method and apparatus of recommending goods based on network | |
| CN110929764A (en) | Picture auditing method and device, electronic equipment and storage medium | |
| CN115712780A (en) | Information pushing method and device based on cloud computing and big data | |
| JP7565926B2 (en) | Encoding Text Data for Personalized Inventory Management | |
| WO2021183999A1 (en) | Value based commodity selection | |
| Li et al. | Learning user preferences across multiple aspects for merchant recommendation | |
| CN116542738A (en) | Information pushing method based on electronic commerce big data | |
| CN112818088A (en) | Commodity search data processing method, commodity search data processing device, commodity search equipment and storage medium | |
| KR20220026255A (en) | Recommendation System for Health Supplement by Using Big Data | |
| JP6979902B2 (en) | Information processing equipment, information processing methods, and programs | |
| CN116629916A (en) | Restaurant demand prediction method and system | |
| CN116340383A (en) | Searching method, device, medium and equipment based on correlation of query and result |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |