Demand forecasting device, demand forecasting method, and program

Abstract

The initiative for controlling food production and reducing food losses by evaluating a relationship between the sales of each item of food and the atmospheric temperature is going on. However, considerations are not taken so far to include a case like a sudden event, such as a disaster, that would cause pervasive clamor among people, and therefore, this approach is insufficient for increasing sales or suppressing wasteful production.　The present invention is a demand prediction device for predicting the demand quantity of a commodity or a service, comprising: a feature quantity storage means for storing a feature quantity of intrinsic representation that indicates a human's psychological state; a feature quantity calculation means for calculating each feature quantity of a plurality of prescribed representations included in prescribed text data; an aggregation means for aggregating the number of specific representations among the plurality of prescribed representations, that pertain to the feature quantity of a prescribed degree of similarity to the feature quantity of the intrinsic representation; and a demand quantity prediction means for predicting said demand quantity that corresponds to the human's psychological state, on the basis of a correlation between performance data regarding demands and the number of specific representations aggregated by the aggregation means.

Description

[Technical field]

[0001] The present disclosure relates to a demand forecasting device, a demand forecasting method, and a program. [Background technology]

[0002] The Ministry of Agriculture, Forestry and Fisheries estimates that food waste in Japan is over 6 million tons per year, and the causes lie at various stages, from factories, distribution, retail, and the food and beverage industry to households. According to the FAO (Food and Agriculture Organization of the United Nations) report "Global Food Loss and Food Waste (2011)," "roughly one-third of the food produced for human consumption is lost or thrown away around the world, amounting to approximately 1.3 billion tons per year." There are various causes for food waste, such as the disposal of non-standard products at the production site when vegetables are different in size from regulations, and disposal due to a gap between supply and demand when there is overproduction.

[0003] In order to reduce food waste, a common approach is to use data obtained from sensors and POS data, and make predictions using techniques such as multiple regression analysis.

[0004] Also, efforts have been started to reduce food waste by controlling food production through evaluation of the relationship between the sales of each food item and temperature (see Non-Patent Document 1). This makes it possible to increase food production and improve sales based on the relationship with temperature, and to reduce food waste by decreasing food production. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Possibility study of reducing food waste by improving the accuracy of demand forecasts using AI at convenience stores March 2020 Japan Weather Association, Internet<URL:https: / / www.dei.or.jp / foodloss / pdf / 200317_2_06.pdf> Summary of the Invention [Problem to be solved by the invention]

[0006] However, conventional methods only take into account the relationship between the sales of each food (product) and temperature, and do not take into account circumstances that cause public uproar, such as sudden events such as disasters, and therefore are insufficient in increasing sales or reducing wasteful production.

[0007] The present invention has been made in consideration of the above points, and aims to increase sales of products or services, or to reduce as much as possible the production and purchase of unnecessary products or the provision of unnecessary services. [Means for solving the problem]

[0008] In order to solve the above problem, the invention according to claim 1 is a demand forecasting device for forecasting the demand for a product or service, comprising: a feature storage means for storing features of named entities that represent human psychological states; , which represents a person's psychological state a feature value calculation means for calculating each feature value of a plurality of predetermined expressions; and a counting means for counting the number of specific expressions related to features of a predetermined similarity to the feature values ​​of the named entities among the plurality of predetermined expressions; For a given product or service demand How it increases and a demand quantity prediction means for predicting the demand quantity according to a human psychological state based on a correlation between actual data regarding the demand and the number of specific expressions tallied by the talliation means. Effect of the Invention

[0009] As described above, the present invention has the effect of increasing sales of products or services, or suppressing wasteful production and purchase of products or provision of wasteful services as much as possible. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic diagram of a communication system. [Diagram 2] FIG. 2 is an electrical hardware configuration diagram of the demand prediction device. [Diagram 3] FIG. 2 is a diagram illustrating an electrical hardware configuration of a communication terminal. [Figure 4] FIG. 2 is a functional configuration diagram of a demand prediction device. [Diagram 5] 11 is a flowchart showing a process for storing features of a named entity. [Figure 6] 13 is a flowchart showing a process of predicting and outputting a demand amount according to a human psychological state. [Figure 7] This is a graph showing the frequency with which each expression was used on social media over a one-month period. [Figure 8] 1 is a graph showing sales figures for a specific product for one month as actual data on demand. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0012] [System configuration of the embodiment] First, an outline of the configuration of a communication system according to the present embodiment will be described with reference to Fig. 1. Fig. 1 is a schematic diagram of a communication system according to an embodiment of the present invention.

[0013] 1, a communication system 1 of this embodiment is constructed by a demand prediction device 3 and a communication terminal 5. The communication terminal 5 is managed and used by a user Y.

[0014] The demand prediction device 3 and the communication terminal 5 can communicate with each other via a communication network 100 such as the Internet. The connection form of the communication network 100 may be either wireless or wired.

[0015] The demand prediction device 3 is configured with one or more computers. When the demand prediction device 3 is configured with multiple computers, it may be referred to as a "demand prediction device" or a "future prediction system."

[0016] The demand forecasting device 3 is a device that forecasts the demand for a product or service. For example, the demand forecasting device 3 forecasts the future demand for the product or service, and outputs result data such as a graph showing the forecast result. Examples of the output method include transmitting the result data to the communication terminal 5, thereby displaying or printing a graph related to the result data on the communication terminal 5 side, displaying the graph on a display connected to the demand forecasting device 3, and printing the graph on a printer connected to the demand forecasting device 3.

[0017] The communication terminal 5 is a computer, and a notebook computer is shown as an example in Fig. 1. In Fig. 1, a user Y operates the communication terminal 5.

[0018] [Hardware configuration] <Hardware configuration of the demand forecasting device> Next, the electrical hardware configuration of the demand prediction device 3 will be described with reference to Fig. 2. Fig. 2 is a diagram showing the electrical hardware configuration of the demand prediction device.

[0019] As shown in FIG. 2, the demand forecasting device 3 is a computer and includes a CPU (Central Processing Unit) 301, a ROM (Read Only Memory) 302, a RAM (Random Access Memory) 303, a HD (Hard Disk) 304, a HDD (Hard Disk Drive) controller 305, an external device connection I / F (Interface) 308, a network I / F 309, a bus line 310, and a media I / F 314.

[0020] Of these, the CPU 301 controls the overall operation of the demand prediction device 3. The ROM 302 stores programs used to drive the CPU 301, such as an IPL (Initial Program Loader). The RAM 303 is used as a work area for the CPU 301.

[0021] The HD 304 stores various data such as programs. The HDD controller 305 controls reading and writing of various data from and to the HD 304 under the control of the CPU 301. Instead of the HD 304 and the HDD controller 305, a solid state drive (SSD) and an SSD controller may be installed.

[0022] The external device connection I / F 308 is an interface for connecting various external devices, such as a display, a speaker, a keyboard, a mouse, a USB (Universal Serial Bus) memory, and a printer.

[0023] The network I / F 309 is an interface for performing data communication via the communication network 100. The bus line 310 is an address bus, a data bus, or the like for electrically connecting each component such as the CPU 301 shown in FIG.

[0024] Furthermore, the media I / F 314 controls reading and writing (storing) of data from and to a recording medium 313 such as a flash memory. The recording medium 313 includes a DVD (Digital Versatile Disc) and a Blu-ray Disc (registered trademark).

[0025] <Hardware configuration of communication terminal> Next, the electrical hardware configuration of the communication terminal 5 will be described with reference to Fig. 3. Fig. 3 is a diagram showing the electrical hardware configuration of the communication terminal.

[0026] As shown in FIG. 3, the communication terminal 5 is a computer and is equipped with a CPU 501, a ROM 502, a RAM 503, a HD 504, a HDD controller 505, a display 506, an external device connection I / F (Interface) 508, a network I / F 509, a bus line 510, a pointing device 512, and a media I / F 514.

[0027] Of these, the CPU 501 controls the overall operation of the communication terminal 5. The ROM 502 stores programs, such as an IPL, used to drive the CPU 501. The RAM 503 is used as a work area for the CPU 501.

[0028] The HD 504 stores various data such as programs. The HDD controller 505 controls reading and writing of various data from and to the HD 504 under the control of the CPU 501. Instead of the HD 504 and the HDD controller 505, an SSD and an SSD controller may be installed.

[0029] The display 506 is a type of display means such as a liquid crystal or organic EL (Electro Luminescence) that displays various images. The external device connection I / F 508 is an interface for connecting various external devices. In this case, the external devices are a display, a speaker, a keyboard, a mouse, a USB memory, a printer, etc.

[0030] The network I / F 509 is an interface for performing data communication via the communication network 100. The bus line 510 is an address bus, a data bus, or the like for electrically connecting each component such as the CPU 501 shown in FIG.

[0031] In addition, the pointing device 512 is a type of input means for selecting and executing various instructions, selecting a processing target, moving a cursor, and the like. When the user Y uses a keyboard, the function of the pointing device 512 may be turned off. The media I / F 514 controls reading or writing (recording) of data to / from the recording media 513 such as flash memories. The recording media 513 includes DVDs, Blu-ray Discs (registered trademarks), and the like.

[0032] 〔Functional Configuration of Demand Prediction Device〕 Next, the functional configuration of the demand prediction device will be described with reference to FIG. 4. FIG. 4 is a functional configuration diagram of the demand prediction device according to an embodiment of the present invention.

[0033] In FIG. 4, the demand prediction device 3 includes an input unit 11, a text analysis unit 12, a named entity extraction unit 13, a predetermined expression extraction unit 14, a feature quantity calculation unit 15, a summarization unit 16, a demand quantity prediction unit 17, and an output unit 19. These units are functions realized by commands from the CPU 301 in FIG. 2 based on a program.

[0034] Furthermore, in the RAM 303 or HD 304 in FIG. 2, a text DB 21, a feature quantity storage unit 22, and a performance data DB 23 related to demand are constructed.

[0035] <Configuration of DB or Storage Unit> (Text DB) The text DB 21 stores and manages (specific) text data including a plurality of named entities representing the mental states of humans. Examples of named entities representing the mental states of humans include "tension / anxiety", "depression", "sadness", and "confusion". Note that the "expressions" in the present invention are character strings that objectify internal aspects such as psychological, emotional, and mental aspects as external, perceptual images, and include words and phrases.

[0036] (Feature Quantity Storage Unit) The feature amount storage unit 22 stores and manages features of a plurality of named entities that are included in the text data of the text DB 21 and that express human psychological states.

[0037] (Demand-related performance data DB) The demand-related performance data DB 23 stores and manages demand-related performance data collected from a POS (Point of Sale) system, etc. By analyzing regular purchase components and seasonal purchase components using a method such as factor decomposition, the influence of emotions related to events such as Christmas and Valentine's Day can be eliminated from the demand-related performance data.

[0038] <Each function configuration> Next, each functional configuration of the demand prediction device will be described with reference to FIG. 2 to FIG.

[0039] The input unit 11 accepts input of various (predetermined) text data from the social media 20 via the external device connection I / F 308, the network I / F 309, or the media I / F 314 in Fig. 2. When via the network I / F 309, the text data is input from the communication terminal 5 or the like via the communication network 100. Note that the social media 20 is a medium designed to spread through social interaction using a wide range of information transmission techniques in which anyone can participate. Examples of social media include blog sites, SNS (Social Networking Service) sites, video sites, microblog sites, and the like.

[0040] Furthermore, the input unit 11 may acquire text data not only from the social media 20 but also from media such as television, radio, electronic newspapers, electronic magazines, electronic government-issued documents, etc. In the case of television and radio, the input unit 11 converts voice data into text data using voice recognition technology.

[0041] The text analysis unit 12 performs morphological analysis on the text data acquired from the input unit 11 or the text DB 21. Morphological analysis is to divide a sentence written in a natural language into morphemes, which are the smallest units of language, and to determine the parts of speech and changes of each morpheme. Examples of morphological analysis engines include MeCab, JUMAN, and ChaSen.

[0042] The named entity extraction unit 13 extracts named entities expressing a human psychological state from text data by a method of named entity extraction. Named entity extraction (named entity recognition, named entity identification, named entity chunking, named entity extraction) is one of the natural language processing techniques using a computer, and is one field of information extraction. Named entity extraction extracts named entities (named entities) from sentences of text data, and classifies them into predefined named entity classifications such as proper nouns (personal names, organization names, place names, etc.), dates, time expressions, quantities, amounts, percentages, etc. In this embodiment, the named entity extraction unit 13 classifies named entities into "tension / anxiety", "depression", "sadness", and "confusion", etc.

[0043] The predetermined expression extraction unit 14 extracts a plurality of predetermined expressions from the text data after the morphological analysis. The predetermined expressions include expressions similar to each named entity, such as "excited," "what should I do," "unmotivated," and "panic."

[0044] The feature amount calculation unit 15 calculates the feature amounts of each of a plurality of predetermined expressions included in the text data acquired from social media, using an existing method such as Word2vec.

[0045] The counting unit 16 counts the number of specific expressions related to features of a predetermined similarity to the features of the named entity among a plurality of predetermined expressions. For example, the counting unit 16 counts the number of specific expressions "heart-pounding" related to features in the same similarity range as the features of the named entity "nervous" among a plurality of predetermined expressions (such as "heart-pounding," "what should I do," "unmotivated," and "panic"). In this case, the feature calculation unit 15 may calculate the similarity in vector representation using the following (Equation 1). For example, when the similarity is 0.9 or more, the two expressions (x, y) are similar.

[0046]

number

[0047] The output unit 19 outputs result data such as graphs showing the prediction results by the demand amount prediction unit 17 to the outside of the demand prediction device 3. As described above, examples of the output method include transmitting the result data to the communication terminal 5 to display or print a graph or the like related to the result data on the communication terminal 5 side, displaying the graph or the like on a display connected to the demand prediction device 3, or printing the graph or the like on a printer or the like connected to the demand prediction device 3.

[0048] [Processing or Operation of the Embodiment] Next, the processing or operation of this embodiment will be described in detail with reference to Figs. 5 to 10. <Storage of Features of Named Entities> The process for storing the feature amounts of named entities will be described with reference to Fig. 5. Fig. 5 is a flowchart showing the process for storing the feature amounts of named entities.

[0049] First, as shown in FIG. 5, the text analysis unit 12 acquires text data from the text DB 21 and performs a morphological analysis (S11).

[0050] Next, the named entity extraction unit 13 extracts named entities representing human psychological states from the text data after the morphological analysis (S12).

[0051] Next, the feature calculation unit 15 calculates the feature of the named entity (S13).

[0052] Finally, the feature storage unit 22 stores the features of the named entity (S14).

[0053] This completes the process for storing the features of a named entity.

[0054] <Demand prediction according to human psychological state> Next, a process of predicting and outputting a demand amount according to a human psychological state will be described with reference to Fig. 6. Fig. 6 is a flowchart showing a process of predicting and outputting a demand amount according to a human psychological state.

[0055] 6, the input unit 11 acquires text data from the social media 20 (S21). Then, the text analysis unit 12 performs a morphological analysis of the text data (S22).

[0056] Next, the predetermined expression extraction unit 14 extracts a plurality of predetermined expressions from the text data after the morphological analysis (S23).Then, the feature amount calculation unit 15 calculates the feature amount of each of the plurality of predetermined expressions (S24).

[0057] Next, the counting unit 16 counts the number of specific expressions related to the feature amount of a predetermined similarity to the feature amount of the named entity, among the plurality of predetermined expressions (S25).

[0058] Next, the demand prediction unit 17 predicts the demand according to the human psychological state based on the correlation between the demand-related performance data acquired from the demand-related performance data DB 23 and the number of specific expressions tabulated by the tabulation unit 16 (S26).

[0059] Here, the process by the demand prediction unit 17 will be described with reference to Fig. 7 and Fig. 8. Fig. 7 is a graph showing the frequency of each expression used in social media for one month. Fig. 8 is a graph showing the sales amount of a specific product for one month as actual data on demand.

[0060] As shown in FIG. 7, the frequency with which specific expressions (such as "tension and anxiety") were used on the social media 20 from the 14th to the 20th, and the sales of a specific product increased sharply around the peak on the 16th, as shown in FIG. 8. Therefore, the demand prediction unit 17 predicts the demand for a specific product according to a human psychological state based on these correlations. As a result, if specific expressions such as "tension and anxiety," "sadness," "confusion," and "depression" start to be used on the social media 20 in the order of frequency, it can be predicted that the demand for the specific product will increase sharply in a few days. Therefore, producers can increase the production volume of a specific product in advance, and sellers can purchase the specific product in advance, to deal with the sudden increase in demand in advance.

[0061] Finally, the output unit 19 outputs the result data indicating the predicted demand amount (S27).

[0062] [Effects of the embodiment] As described above, according to this embodiment, when it is possible to predict in advance that demand will increase, the supply can be increased, thereby increasing the production and purchase of goods, or increasing the number of people providing services, thereby increasing sales. Conversely, when it is possible to predict in advance that demand will not increase (or will decrease), the supply can be suppressed, thereby avoiding wasteful production and purchase of goods, or reducing the number of people providing services, thereby suppressing wasteful expenses. As a result, for example, when the product is a food with a short expiration date, the store manager can increase sales by purchasing in advance a larger amount of food whose demand is likely to increase. Similarly, when the product is a food with a short expiration date, the store manager can solve the problem of unsold food having to be discarded by purchasing in advance a smaller amount of food whose demand is unlikely to increase (or decrease).

[0063] 〔supplement〕 The present invention is not limited to the above-described embodiment, and may have the following configurations or processes (operations). (1) The demand forecasting device 3 can be realized by a computer and a program. This program can be recorded on a recording medium or provided via the communication network 100. (2) In the communication between the demand forecasting device 3 and the communication terminal 5, other devices (such as a server or a router) may relay data. For example, for the sake of simplicity, this specification describes that the input unit 11 of the demand forecasting device 3 transmits the result data to the communication terminal 5, but this transmission process is intended to include cases where other devices relay the result data. (4) In the above embodiment, a notebook computer is shown as an example of the communications terminal 5, but this is not limited to this and may be, for example, a desktop computer, a tablet terminal, a smartphone, a smart watch, a car navigation device, a refrigerator, a microwave oven, etc. (5) Each of the CPUs 301 and 501 may be multiple, rather than just single. (6) At least one of the calculations performed by the demand prediction unit 17 may use a neural network. [Explanation of symbols]

[0064] 1. Communication Systems 3. Demand forecasting device 5. Communication terminals 11 Input unit (an example of an input means) 12 Text analysis unit (an example of a text analysis means) 13 Named entity extraction unit (an example of a named entity extraction means) 14. Predetermined expression extraction unit (an example of a predetermined expression extraction means) 15 Feature calculation unit (an example of feature calculation means) 16 Counting unit (an example of a counting means) 17 Demand forecasting unit (an example of a demand forecasting means) 19 Output unit (an example of an output means) 21 Text Database 22 Feature storage unit (an example of feature storage means) 23 Demand-related performance data DB

Claims

1. A demand forecasting device for forecasting a demand for a product or service, A feature storage means for storing features of named entities representing a human psychological state; a feature value calculation means for calculating feature values ​​of a plurality of predetermined expressions, which are included in the predetermined text data and which express a human psychological state; a counting means for counting the number of specific expressions related to a feature of a predetermined similarity to the feature of the named entity among the plurality of predetermined expressions; a demand prediction means for predicting the demand according to a human psychological state based on a correlation between performance data on the increase in demand for a predetermined product or service and the number of specific expressions counted by the counting means; A demand forecasting device comprising:

2. The counting means counts the number of the multiple specific expressions, the demand prediction means predicts the demand according to the psychological state of the person based on a correlation between the performance data and the order of occurrence of the plurality of specific expressions; The demand forecasting device according to claim 1 .

3. The demand forecasting device according to claim 1 or 2, a named entity extraction means for extracting the named entity representing a human psychological state from specific text data, The feature calculation means calculates features of the named entities extracted by the named entity extraction means, The feature amount storage means stores the feature amount of the named entity calculated by the feature amount calculation means. A demand forecasting device comprising:

4. The demand forecasting device according to claim 1 or 2, A predetermined expression extraction means for extracting the plurality of predetermined expressions from the predetermined text data acquired from social media, The demand forecasting device is characterized in that the feature calculation means calculates each feature of the plurality of specified expressions extracted by the specified expression extraction means.

5. The demand forecasting device according to claim 4, a text analysis means for performing morphological analysis on the predetermined text data, The demand forecasting device, wherein the predetermined expression extraction means extracts the plurality of predetermined expressions from text data after morphological analysis by the text analysis means.

6. A demand forecasting method executed by a demand forecasting device that forecasts a demand for a product or service, comprising: the demand prediction device has a feature storage means for storing features of named entities representing human psychological states, The demand prediction device includes: a feature amount calculation step of calculating feature amounts of each of a plurality of predetermined expressions which are included in the predetermined text data and which express a human psychological state; a counting step of counting the number of specific expressions related to a feature of a predetermined similarity to the feature of the named entity among the plurality of predetermined expressions; a demand prediction step of predicting the demand according to a human psychological state based on a correlation between performance data on the increase in demand for a predetermined product or service and the number of specific expressions counted in the counting step; A demand forecasting method comprising the steps of:

7. A program causing a computer to execute the method according to claim 6.

Images