Information processing device

Abstract

Increase the amount of learning data about users based on their behavioral history. [Solution] The information processing device (10, 10a) includes an input means (111) for inputting a prompt to a first large-scale language model that includes historical information relating to the user's behavior history and facility information relating to one or more facilities; an acquisition means (112) for obtaining a response from the first large-scale language model indicating whether or not the user will visit one or more facilities; and a generation means (113) for generating training data for machine learning of a model different from the first large-scale language model based on the historical information and the response.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing apparatuses.

Background Art

[0002] As an apparatus of this kind, for example, an apparatus related to a recommendation service that recommends items based on a user's preferences using the user's usage history information has been proposed (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For example, a recommendation service may be provided using a learning model constructed by machine learning. By the way, when focusing on a single user, the number of items used by the single user is significantly smaller than the number of items provided. For this reason, even if the usage history information of a single user is used as learning data, there is a possibility that the characteristics of the single user are not reflected in the learning model.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide an information processing apparatus capable of increasing learning data regarding a single user.

Means for Solving the Problems

[0006] An information processing device according to one aspect of the present invention includes: an input means for inputting a prompt to a first large-scale language model, which includes historical information relating to a user's behavior history and facility information relating to one or more facilities; an acquisition means for obtaining a response from the first large-scale language model indicating whether or not the user will visit the one or more facilities; and a generation means for generating training data for machine learning of a model different from the first large-scale language model, based on the historical information and the response. [Brief explanation of the drawing]

[0007] [Figure 1] This is a block diagram showing the configuration of the information processing device according to the first embodiment. [Figure 2] This is a flowchart showing the operation of the information processing device according to the first embodiment. [Figure 3] This is a block diagram showing the configuration of the information processing device according to the second embodiment. [Figure 4] This is a flowchart showing the operation of the information processing device according to the second embodiment. [Modes for carrying out the invention]

[0008] <First Embodiment> A first embodiment of the information processing device will be described with reference to Figures 1 and 2. In Figure 1, the information processing device 10 includes an arithmetic unit 11, a storage device 12, a communication device 13, an input device 14, and an output device 15. The arithmetic unit 11, storage device 12, communication device 13, input device 14, and output device 15 are connected via a data bus 16. The information processing device 10 may be a personal computer, a tablet terminal, or a smartphone. The information processing device 10 can access the server 20, which provides services using a large-scale language model (LLM), via a network NW. In other words, the information processing device 10 and the server 20 are configured to communicate with each other via the network NW. The server 20 may be a cloud server. The network NW may be a wide-area network such as the Internet, or a narrow-area network such as a LAN (Local Area Network). The information processing device 10 and the server 20 may constitute an information processing system 1.

[0009] The arithmetic unit 11 may have a processor. The arithmetic unit 11 may have a single processor or multiple processors. In other words, the arithmetic unit 11 may have one or more processors. Furthermore, the processor may be a multi-core processor. If the arithmetic unit 11 has a single processor that is a multi-core processor, then logically, the arithmetic unit 11 can be said to have multiple processors.

[0010] The processor may be at least one of the following: CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), and TPU (Tensor Processing Unit).

[0011] The storage device 12 may be at least one of the following: RAM (Random Access Memory), ROM (Read Only Memory), hard disk drive, magneto-optical disk drive, SSD (Solid State Drive), and optical disk array. In other words, the storage device 12 may be implemented by a single device or by multiple devices.

[0012] The communication device 13 may be capable of communicating with devices outside the information processing device 10. The communication device 13 may use either wired or wireless communication.

[0013] The input device 14 is a device capable of receiving information input to the information processing device 10 from an external source. The input device 14 may include an operating device (e.g., keyboard, mouse, touch panel, etc.) that can be operated by the user of the information processing device 10. The input device 140 may include a microphone capable of collecting the user's voice. In this case, the user may perform voice input. The input device 14 may include a recording medium reader capable of reading information recorded on a recording medium that can be attached to and detached from the information processing device 10, such as a USB (Universal Serial Bus) memory. Furthermore, when information is input to the information processing device 10 via the communication device 13 (in other words, when the information processing device 10 acquires information via the communication device 13), the communication device 13 may function as an input device.

[0014] The output device 15 is a device capable of outputting information to the outside of the information processing device 10. The output device 15 has a display device 151 capable of outputting visual information such as characters and images as the above information. The output device 15 may also have a speaker capable of outputting auditory information such as sound as the above information. The output device 15 may also have a vibration motor capable of outputting tactile information such as vibration as the above information. The output device 15 may also have a printer. The output device 150 may be capable of outputting information to a recording medium that can be attached to and detached from the information processing device 10, such as a USB memory stick. When the information processing device 10 outputs information via the communication device 13, the communication device 13 may function as an output device.

[0015] The storage device 12 is capable of storing desired data. The storage device 12 may store the computer program CP that the arithmetic unit 11 will execute. The storage device 12 may temporarily store data that the arithmetic unit 11 will use temporarily when the arithmetic unit 11 is executing the computer program CP.

[0016] Furthermore, the computer program CP may be recorded on a non-temporary recording medium that is readable by a computer. In this case, the computer program CP may be stored in the storage device 12 by reading the recording medium using a recording medium reading device (not shown) provided by the information processing device 10. Furthermore, at least one of the following may be used as the recording medium: an optical disc, a magnetic medium, a magneto-optical disc, a semiconductor memory, and any other medium capable of storing a program. Furthermore, the computer program CP may be obtained from an external device (not shown) of the information processing device 10 via a communication device 13. In other words, the computer program CP may be downloaded from an external device to the storage device 12 of the information processing device 10.

[0017] The arithmetic unit 11 (for example, a processor) may execute the processing that the information processing device 10 should perform together with the memory device 12 in which the computer program CP is stored (in other words, together with the memory device 12 and the computer program CP stored in the memory device 12). For example, by the arithmetic unit 11 executing the computer program CP, a logical functional block for executing the processing that the information processing device 10 should perform may be realized within the arithmetic unit 11 (for example, within the processor).

[0018] In this embodiment, the arithmetic unit 11 implements an input unit 111, an acquisition unit 112, and a generation unit 113 as functional blocks. At least one of the input unit 111, the acquisition unit 112, and the generation unit 113 may be implemented as a physical processing circuit. Alternatively, at least one of the input unit 111, the acquisition unit 112, and the generation unit 113 may be implemented in a form that combines a logical functional block and a physical processing circuit.

[0019] For example, there is a recommendation service that recommends restaurants based on a user's preferences. When attempting to provide a recommendation service using a learning model constructed by machine learning, there are the following problems. When focusing on a single user, the number of restaurants visited by the single user is significantly smaller compared to the total number of restaurants targeted by the recommendation service. Therefore, even if the historical information of a single user is used as learning data, there is a possibility that the learning model cannot recommend restaurants that match the preferences of the single user. The information processing apparatus 10 according to the present embodiment performs the operations described below in order to increase the learning data related to a single user.

[0020] The operation of the information processing apparatus 10 will be described with reference to the flowchart of FIG. 2. In FIG. 2, an input unit 111 of the information processing apparatus 10 transmits a prompt including historical information regarding the behavior history of a single user and facility information regarding one or more facilities to the server 20 via the network NW (step S101). As a result, the above prompt is input into the large language model.

[0021] The behavior history is a concept including at least one of an access history, a purchase history, a usage history, and a browsing history. The behavior history of a single user may be stored in the storage device 12 or may be stored in a device different from the information processing apparatus 10. The facility may be an indoor facility or an outdoor facility. For example, the prompt input into the large language model may include a question such as "Does a single user visit XX (facility name)?". For example, the prompt input into the large language model may include a question such as "What facilities does a single user visit when going to YY (region name)?". Note that the prompt input into the large language model may include words specifying the day of the week or the time zone.

[0022] When the large language model outputs an answer to a prompt (for example, an answer indicating whether a user visits one or more facilities), the server 20 transmits the answer to the information processing device 10 via the network NW. As a result, the acquisition unit 112 of the information processing device 10 acquires the answer of the large language model (step S102).

[0023] Next, the generation unit 113 of the information processing device 10 generates learning data (that is, learning data for machine learning of a model different from the large language model) based on the history information of one user and the answer of the large language model (step S103).

[0024] (Technical effect) The prompt input to the large language model includes history information regarding the action history of one user. Therefore, it can be expected that the answer of the large language model is based on the action history of one user. As a result, based on the answer of the large language model, the action history of one user can be artificially generated. By the generation unit 113 of the information processing device 10 generating learning data based on the actual action history of one user and the answer of the large language model, the learning data regarding one user can be significantly increased.

[0025] As described above, by performing machine learning using the generated training data, it is possible to construct a learning model that can make recommendations tailored to the preferences of a single user. Furthermore, recommendation services using such learning models are more efficient than recommendation services using large-scale language models in terms of response time and the cost required for model operation. In addition, when the training data based on the history information of a single user is relatively small, a method has been proposed to generate training data based on the history information of a single user and the history information of other users similar to that user. However, if the history information of a single user is smaller than the history information of other users, the accuracy of the constructed learning model may be relatively low. In contrast, in this embodiment, since the history information of other users is not used, it is possible to suppress the relatively low accuracy of the constructed learning model.

[0026] <Second Embodiment> A second embodiment of the information processing device will be described with reference to Figures 3 and 4. Note that, in the description of the second embodiment, explanations that overlap with the description of the first embodiment will be omitted as appropriate. Common parts in the drawings are denoted by the same reference numerals. In Figure 3, the information processing device 10a includes an arithmetic unit 11, a storage device 12, a communication device 13, an input device 14, and an output device 15. The information processing device 10a can access servers 20 and 30 that provide services using a large-scale language model via a network NW. Note that server 30 may be a cloud server. Note that the information processing device 10a and servers 20 and 30 may constitute an information processing system 2.

[0027] In this embodiment, the arithmetic unit 11 implements an input unit 111, an acquisition unit 112, a generation unit 113, a determination unit 114, and an adjustment unit 115 as functional blocks. At least one of the input unit 111, acquisition unit 112, generation unit 113, determination unit 114, and adjustment unit 115 may be implemented as a physical processing circuit. Alternatively, at least one of the input unit 111, acquisition unit 112, generation unit 113, determination unit 114, and adjustment unit 115 may be implemented in a form that combines a logical functional block and a physical processing circuit.

[0028] The operation of the information processing device 10a will be explained with reference to the flowchart in Figure 4. After the processing of step S102 described above, the determination unit 114 of the information processing device 10a determines whether the response of the server 20's large-scale language model is valid or not (step S201). In the processing of step S201, the determination unit 114 may transmit the prompt input to the server 20's large-scale language model in the processing of step S101 and the response obtained in the processing of step S102 to the server 30 via the network NW. As a result, the above prompt and response may be input to the server 30's large-scale language model. Subsequently, the determination unit 114 may determine whether the response of the server 20's large-scale language model is valid or not based on the response from the server 30's large-scale language model. In other words, the determination unit 114 may use the server 30's large-scale language model to determine whether the response of the server 20's large-scale language model is valid or not. This method is called LLM-as-a-Judge.

[0029] If, in step S201, the response of the server 20's large-scale language model is determined to be valid (step S201: Yes), the process of step S103 described above is performed. In this case, the generation unit 113 generates training data based on the history information of a user and the response that the determination unit 114 determined to be valid.

[0030] If, in the process of step S201, it is determined that the response from the server 20's large-scale language model is not valid (step S201: No), the adjustment unit 115 of the information processing device 10a may adjust the prompt input to the server 20's large-scale language model in the process of step S101 (step S202). After that, the process of step S101 may be performed. The adjustment unit 115 may also adjust the prompt using, for example, the server 30's large-scale language model.

[0031] (Technical effects) In this embodiment, the determination unit 114 determines whether the response of the server 20's large-scale language model is valid or not. If the server 20's large-scale language model is determined to be valid, the generation unit 113 generates training data. Therefore, according to this embodiment, training data can be generated based on the large-scale language model's responses, which have relatively high accuracy.

[0032] Various aspects of the invention derived from the embodiments described above are described below.

[0033] An information processing device according to one aspect of the invention includes: an input means for inputting a prompt to a first large-scale language model, which includes historical information relating to a user's behavior history and facility information relating to one or more facilities; an acquisition means for obtaining a response from the first large-scale language model indicating whether or not the user will visit the one or more facilities; and a generation means for generating training data for machine learning of a model different from the first large-scale language model, based on the historical information and the response.

[0034] In the above embodiment, the "input unit 111" corresponds to an example of an "input means," the "acquisition unit 112" corresponds to an example of an "acquisition means," and the "generation unit 113" corresponds to an example of a "generation means." In the above embodiment of the information processing device, even if a user's behavioral history is relatively small, the amount of learning data about that user based on that behavioral history can be increased.

[0035] The information processing device according to the above embodiment may include a determination means for determining whether the answer is valid or not. In the above embodiment, the "determination unit 114" corresponds to an example of the "determination means". In this case, the determination means may determine whether the answer is valid or not using a second large-scale language model different from the first large-scale language model. Furthermore, the generation means may generate the learning data based on the history information and the answer that the determination means has determined to be valid.

[0036] The information processing device according to the above embodiment may include an adjustment means for adjusting the prompt when the determination means determines that the response is not valid. In the above embodiment, the "adjustment unit 115" corresponds to an example of the "adjustment means".

[0037] The present invention is not limited to the embodiments described above, and can be modified as appropriate without contradicting the gist or idea of ​​the invention as can be read from the claims and specification as a whole. Information processing devices that involve such modifications are also included within the technical scope of the present invention. [Explanation of Symbols]

[0038] 10, 10a... Information processing device, 111... Input unit, 112... Acquisition unit, 113... Generation unit, 114... Determination unit, 115... Adjustment unit

Claims

1. An input means for inputting a prompt containing historical information about the user's behavior history and facility information about one or more facilities into a first large-scale language model, An acquisition means for obtaining a response from the first large-scale language model indicating whether the user will visit the one or more facilities, A generation means that generates training data for machine learning of a model different from the first large-scale language model, based on the historical information and the response, An information processing device equipped with the following features.

2. The system includes a determination means for determining whether the aforementioned answer is valid or not. The information processing apparatus according to claim 1.

3. The determination means uses a second large-scale language model, which is different from the first large-scale language model, to determine whether the response is valid or not. The information processing apparatus according to claim 2.

4. The generation means generates the learning data based on the historical information and the answers that the determination means has determined to be valid. The information processing apparatus according to claim 2.

5. The determination means includes an adjustment means for adjusting the prompt if it determines that the response is not valid. The information processing apparatus according to claim 2.

Images