Information processing device and information processing method

The information processing device enhances recommendation accuracy by extracting and combining personal and service characteristic vectors from time-series data, addressing data sparsity and cold start issues in cross-domain scenarios.

WO2026239537A1PCT designated stage Publication Date: 2026-11-19NT T INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2025/017603
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-05-14
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Conventional cross-domain recommendation systems face challenges in improving recommendation accuracy due to data sparsity and cold start problems, particularly when user item attribute information is heterogeneous across domains.

Method used

An information processing device that collects time-series data from users across services, extracts personal and service characteristic vectors, and combines them to enhance recommendation accuracy by considering both user preferences and service tendencies.

Benefits of technology

Improves recommendation accuracy for new users or services with limited data, and across domains with varying user-item information, by leveraging service and personal characteristic vectors, including those of similar users.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2025017603_19112026_PF_FP_ABST
    Figure JP2025017603_19112026_PF_FP_ABST
Patent Text Reader

Abstract

An information processing device (10) collects time-series data of each user using each service, extracts, for each user, a personal characteristic vector indicating the characteristics of the user from the collected time-series data, and extracts, for each service, a service characteristic vector indicating the characteristics of users using the service from the collected time-series data. The information processing device (10) outputs, as information to be recommended to a recommendation target user using a service, the personal characteristic vector of the recommendation target user and the service characteristic vector related to the service.
Need to check novelty before this filing date? Find Prior Art

Description

Information Processing Apparatus and Information Processing Method

[0001] The present invention relates to an information processing apparatus and an information processing method.

[0002] Conventionally, recommendation services are widely used on EC (Electronic Commerce) sites and Web services as a method of proposing products and services that match user interests. In such recommendation services, latent needs are grasped by analyzing and learning user browsing, behavior, purchase history, and the like, and recommendation display is performed for each individual user to promote user purchasing behavior. In addition, in order to provide accurate recommendations to users, recommendation services demand a technology that collects user data and understands customers down to insights that are difficult to manifest, such as personality and values.

[0003] General recommendation systems may suffer from degraded recommendation performance due to data sparsity (few items rated by users) and cold start problems (insufficient information on new users or items). Accordingly, there is a cross-domain recommendation system (hereinafter referred to as "CDR") as a recommendation system that utilizes information from a "source domain" with abundant data to improve the recommendation accuracy of a "target domain" with little data. CDR performs recommendations by learning the relevance at the content level, item level, and user level from data. Here, the "domain" in the recommendation system refers to the type or category of data. For example, "books" and "movies" are examples of domains.

[0004] Feng Zhu, Yan Wang, Chaochao Chen, Jun Zhou, Longfei Li, Guanfeng Liu. "Cross-Domain Recommendation: Challenges, Progress, and Prospects", 2021 [online], [retrieved on May 9, 2025], Internet <https: / / arxiv.org / pdf / 2103.01696>

[0005] However, conventional technologies had the problem that they could not always improve recommendation accuracy. For example, most CD-Rs assume that user item attribute information is homogeneous. Here, homogeneous user item attribute information means, for example, that service A and service B have a common domain X. To explain with a concrete example, for example, both Amazon® and Kindle® services have a common domain "books," and their attribute information can be said to be homogeneous. However, in reality, the information on user items that can be used between domains often differs greatly, and there were cases where high recommendation accuracy could not be achieved.

[0006] To solve the above-mentioned problems and achieve the objective, the information processing device of the present invention is characterized by comprising: a collection unit that collects time-series data of each user using each service; a personal characteristic extraction unit that extracts a personal characteristic vector indicating the characteristics of each user from the time-series data collected by the collection unit; a service characteristic extraction unit that extracts a service characteristic vector indicating the characteristics of users using each service from the time-series data collected by the collection unit; and an output unit that outputs the personal characteristic vector of the recommended user and the service characteristic vector related to the service as information for recommending to a recommended user who uses the service.

[0007] According to the present invention, it is possible to improve recommendation accuracy.

[0008] Figure 1 is a diagram illustrating an example configuration of a recommendation system according to the first embodiment. Figure 2 is a block diagram illustrating the configuration of the information processing device according to the first embodiment. Figure 3 is a diagram illustrating the learning process of the behavioral sequence generation model. Figure 4 is a diagram illustrating the extraction process of individual characteristic vectors. Figure 5 is a diagram illustrating the extraction process of service characteristic vectors. Figure 6 is a diagram illustrating an example of recommendation output. Figure 7 is a diagram illustrating an example of processing of the information processing device according to the first embodiment. Figure 8 is a flowchart illustrating the processing flow of the information processing device according to the first embodiment. Figure 9 is a diagram illustrating an example configuration of a recommendation system according to the second embodiment. Figure 10 is a block diagram illustrating an example of the configuration of the information processing device according to the third embodiment. Figure 11 is a diagram illustrating an example of processing of the information processing device according to the third embodiment. Figure 12 is a diagram illustrating an example of processing of the information processing device according to another embodiment. Figure 13 is a diagram illustrating an example of a computer that executes an information processing program.

[0009] Hereinafter, the first to fourth embodiments of the present invention will be described in detail in order with reference to the drawings. However, the present invention is not limited to these embodiments. Furthermore, in the drawings, the same parts are denoted by the same reference numerals.

[0010] [First Embodiment] Figure 1 is a diagram showing an example configuration of a recommendation system according to the first embodiment. As shown in Figure 1, the recommendation system 1 according to the first embodiment has an information processing device 10 and a plurality of servers 20X, 20Y that provide services. Furthermore, the configuration shown in Figure 1 is merely an example, and the specific configuration and the number of each device are not particularly limited. Note that each device shown in Figure 1 may communicate via any communication network such as the Internet, LAN (Local Area Network), or VPN (Virtual Private Network).

[0011] The information processing device 10 collects time-series data of each user using each service from each server 20X, 20Y. Here, time-series data can be anything, but for example, it includes data such as the user's behavior history (purchase history, GPS information, application usage logs) and conversation information. The information processing device 10 stores the collected time-series data in the storage unit 13.

[0012] The information processing device 10 then extracts a personal characteristic vector representing the characteristics of each user from the collected time-series data and stores it in the storage unit 13. In other words, the information processing device 10 extracts the user's personal characteristic vector (for example, user-specific preferences) from service-independent time-series data.

[0013] Next, the information processing device 10 extracts a service characteristic vector from the collected time-series data, representing the characteristics of users who utilize each service, and stores it in the storage unit 13. In other words, the information processing device 10 extracts service characteristics (for example, service-specific trends) from the time-series data of users utilizing the service.

[0014] Subsequently, the information processing device 10 outputs to servers 20X and 20Y the personal characteristic vector of the recommended user and the service characteristic vector related to the service, as information for recommending to the recommended user who will use the service. In this way, the information processing device 10 makes it possible to make recommendations by considering both the user's preferences and the service's tendencies by combining a single characteristic vector.

[0015] Servers 20X and 20Y are server devices that provide services. Servers X and 20Y each have a time-series behavior recognition unit 21X and 21Y and a recommendation engine 22X and 22Y, respectively. The time-series behavior recognition units 21X and 21Y format the data obtained from the service into time-series data and notify the information processing device 10. The recommendation engines 22X and 22Y output recommendation information when they receive a user characteristic vector and a service characteristic vector as input. The recommendation engine can be of any type and is not limited thereto. In the following description, when servers 20X and 20Y are not specifically distinguished, they will be referred to as server 20.

[0016] Thus, the information processing device 10 of the recommendation system 1 according to the first embodiment extracts service characteristic vectors from time-series data of users using the service, and also extracts personal characteristic vectors of users from service-independent time-series data. By utilizing these, it is possible to improve recommendation accuracy even in cross-domain settings where the information on user items available across domains differs significantly.

[0017] [Configuration of the Information Processing Device] Figure 2 is a block diagram illustrating the configuration of the information processing device of this embodiment. As illustrated in Figure 2, the information processing device 10 of this embodiment includes a communication processing unit 11, a control unit 12, and a storage unit 13.

[0018] The communication processing unit 11 is implemented using a NIC (Network Interface Card) or the like, and controls communication via telecommunication lines such as a LAN (Local Area Network) or the Internet.

[0019] The storage unit 13 stores data and programs necessary for various processes performed by the control unit 12. For example, the storage unit 13 may be a semiconductor memory element such as RAM (Random Access Memory) or flash memory, or a storage device such as a hard disk or optical disc. As illustrated in Figure 2, the storage unit 13 includes a time-series data storage unit 13a, a personal characteristics storage unit 13b, a service characteristics storage unit 13c, and a similarity storage unit 13d.

[0020] The time-series data storage unit 13a stores time-series data for each user using each service. For example, the time-series data storage unit 13a stores data including each user's activity history (purchase history, GPS information, app usage logs) and conversation information as time-series data. To explain using purchase history as an example, the time-series data storage unit 13a stores, for example, user A's past purchase history, associating information identifying the service used at the time of purchase, information identifying the purchased product, and the date and time of purchase.

[0021] The personal characteristics storage unit 13b stores a personal characteristics vector for each user, representing the characteristics of each user. For example, the personal characteristics storage unit 13b stores a numerical vector representing the internal characteristics of an individual user as the personal characteristics vector. The service characteristics storage unit 13c stores a service characteristics vector for each service, representing the characteristics of users who use each service. The personal characteristics vector and service characteristics vector can be anything, but for example, a 768-dimensional feature vector is used.

[0022] The similarity storage unit 13d stores the similarity of the personal characteristic vectors between users. For example, the similarity storage unit 13d stores a value between 0 and 1 as the similarity. To explain with a specific example, if there are users A to D, the similarity storage unit 13d stores the similarity between user A and user B as "0.5", between user A and user C as "0.7", between user A and user D as "0.2", between user B and user C as "0.9", between user B and user D as "0.1", and between user B and user D as "0.5".

[0023] The control unit 12 has an internal memory for storing programs that define various processing procedures and required data, and executes various processes using these. For example, the control unit 12 has a collection unit 12a, a personal characteristics extraction unit 12b, a similarity estimation unit 12c, a service characteristics extraction unit 12d, and an output unit 12e. Here, the control unit 12 is an electronic circuit such as a CPU (Central Processing Unit) or MPU (Micro Processing Unit), or an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array).

[0024] The collection unit 12a collects time-series data for each user using each service. For example, the collection unit 12a collects time-series data for each user from the server 20 of each service at predetermined time intervals and stores the collected time-series data in the time-series data storage unit 13a. The timing at which the collection unit 12a collects the time-series data is not limited to any specific timing.

[0025] The personal characteristic extraction unit 12b extracts a personal characteristic vector representing the characteristics of each user from the time-series data collected by the collection unit 12a. For example, the personal characteristic extraction unit 12b first reads time-series data of many users from the time-series data storage unit 13a and learns a behavioral series generation model. This behavioral series generation model, given behavior at time t as input, generates general behavioral tendencies at times t-1 and t+1. Subsequently, the personal characteristic extraction unit 12b uses the behavioral series generation model to learn a user embedding layer from each person's time-series data to bridge the "gap" between individual behavior and general behavioral tendencies. After learning, the user embedding layer is extracted and stored in the personal characteristic storage unit 13b as a personal characteristic vector.

[0026] Here, we will explain the learning process of the behavioral sequence generation model using Figure 3. Figure 3 is a diagram illustrating the learning process of the behavioral sequence generation model. As illustrated in Figure 3, the behavioral sequence generation model receives behavioral data (time-series data) from all users as training data (see (1-1) in Figure 3) and learns general behavioral tendencies (see (1-2) in Figure 3).

[0027] Next, the process of extracting individual characteristic vectors will be explained using Figure 4. Figure 4 is a diagram illustrating the process of extracting individual characteristic vectors. As illustrated in Figure 4, the individual characteristic extraction unit 12b inputs the behavioral data of a specific user (in the example of Figure 4, "Person A") as training data into the behavioral sequence generation model (see (2-1) in Figure 4), and learns the user embedding layer by learning the "deviation" from common behavioral tendencies as individual characteristics (see (2-2) in Figure 4). Then, the individual characteristic extraction unit 12b takes Person A's time-series data as input and extracts Person A's individual characteristic vector using the behavioral sequence generation model that includes the learned user embedding layer (see (2-3) in Figure 4).

[0028] The similarity estimation unit 12c compares the personal characteristic vectors of each user extracted by the personal characteristic extraction unit 12b and estimates the similarity. For example, the similarity estimation unit 12c reads the personal characteristic vectors of each user from the personal characteristic storage unit 13b, compares the personal characteristic vectors of two users, estimates the similarity, and stores the estimated similarity in the similarity storage unit 13d. The similarity estimation unit 12c then repeatedly performs the process of estimating similarity for all combinations of users. Note that the method for estimating similarity is not limited and any method may be used.

[0029] The service characteristic extraction unit 12d extracts service characteristic vectors for each service from the time-series data collected by the collection unit 12a, which represent the characteristics of users who use that service. For example, the procedure for learning the behavioral sequence generation model of the service characteristic extraction unit 12d is the same as that of the individuality extraction unit, and it is possible to reuse what has been learned once. Subsequently, the service characteristic extraction unit 12d uses the behavioral sequence generation model to learn a service embedding layer from the time-series data of each service to bridge the gap between behavior during service use and general behavioral tendencies. After learning, the service characteristic extraction unit 12d extracts the service embedding layer and stores it in the service characteristic storage unit 13c as a service characteristic vector.

[0030] Here, we will explain the process of extracting service characteristic vectors using Figure 5. Figure 5 is a diagram illustrating the process of extracting service characteristic vectors. As illustrated in Figure 5, the service characteristic extraction unit 12d inputs user behavior data of users utilizing a specific service (in the example of Figure 5, "Service X") as training data into the behavioral sequence generation model (see (2-1) in Figure 5), and learns the user embedding layer by learning the "deviation" from common behavioral tendencies as service characteristics (see (2-2) in Figure 5). Then, the service characteristic extraction unit 12d takes time-series data of multiple users utilizing Service X as input and extracts the service characteristic vector of Service X using the behavioral sequence generation model that includes the learned user embedding layer (see (2-3) in Figure 5).

[0031] The output unit 12e outputs the personal characteristics vector of the recommended user and the service characteristics vector related to the service as information for recommending to the recommended user who will use the service. In addition, the output unit 12e further outputs the personal characteristics vectors of users whose similarity to the recommended user is above a predetermined threshold.

[0032] For example, the output unit 12e outputs to the recommendation engine of service X the personal characteristics vector of the recommended user A, the service characteristics vector X related to the service, and the personal characteristics vector of user B whose similarity to the recommended user A is above a predetermined threshold, as information for recommending to the recommended user A who uses service X.

[0033] Here, we will explain an example of recommendation output using Figure 6. Figure 6 is a diagram illustrating an example of recommendation output. In the example in Figure 6, it is shown that user A purchased a T-shirt from service X, and three days later purchased socks and a hat. The output unit 12e outputs to the recommendation engine the personal characteristic vector extracted from the time-series data of user A's purchase history, the personal characteristic vectors of other users whose personal characteristic vectors are similar to user A's, and the service characteristic vector of service X. The recommendation engine can predict that the next item user A will want is shoes and make a recommendation.

[0034] Thus, the information processing device 10 can improve recommendation accuracy by utilizing a service characteristic vector from the time-series data of users using the service, a user's personal characteristic vector from service-independent time-series data, and the service characteristic vectors of other users similar to the user using the service.

[0035] Here, an example of processing of the information processing device 10 of the first embodiment will be explained using Figure 7. Figure 7 is a diagram illustrating an example of processing of the information processing device of the first embodiment. As illustrated in Figure 7, the information processing device 10 collects time-series data of user A using service X from server 20X that provides service X and stores it in storage unit 13.

[0036] The information processing device 10 then extracts a personal characteristic vector representing the characteristics of user A from the time-series data of user A that it has collected, and stores it in the storage unit 13. The information processing device 10 also compares the personal characteristic vector of user A with that of other users and estimates the degree of similarity for each.

[0037] The information processing device 10 then outputs the personality vector of user A, along with the personality vectors of users highly similar to user A, and the service characteristic vector of service X to the server 20X. As a result, the recommendation engine 22X of server 20X uses the personality vector of user A, the personality vectors of users highly similar to user A, and the service characteristic vector of service X as input data to predict that the next thing user A wants is shoes and makes a recommendation.

[0038] [Processing Procedure of Information Processing Device 10] Next, an example of the processing procedure of information processing performed by the information processing device 10 will be described using Figure 8. Figure 8 is a flowchart showing the processing flow of the information processing device according to the first embodiment.

[0039] As illustrated in Figure 8, the collection unit 12a of the information processing device 10 collects time-series data of each user using each service (step S101). For example, the collection unit 12a collects time-series data of each user from the server 20 of each service at predetermined time intervals and stores the collected time-series data in the time-series data storage unit 13a.

[0040] Next, the personal characteristic extraction unit 12b extracts a personal characteristic vector for each user from the time-series data collected by the collection unit 12a (step S102). Subsequently, the similarity estimation unit 12c compares the personal characteristic vectors of each user extracted by the personal characteristic extraction unit 12b and estimates the similarity (step S103). In addition, the service characteristic extraction unit 12d extracts a service characteristic vector for each service from the time-series data collected by the collection unit 12a, indicating the characteristics of users who use that service (step S104).

[0041] Subsequently, the similarity estimation unit 12c and the service characteristic extraction unit 12d store the estimated similarity and the extracted service characteristic vector in the storage unit 13 (step S105).

[0042] [Effects of Embodiment] As described above, the information processing apparatus 10 of the present embodiment collects time-series data of each user using each service, extracts, for each user from the collected time-series data, a personal characteristic vector indicating the characteristic of the user, and extracts, for each service from the collected time-series data, a service characteristic vector indicating the characteristic of users using the service. Thereafter, the information processing apparatus 10 outputs the personal characteristic vector of the recommendation target user and the service characteristic vector related to the service, as information for making a recommendation to the recommendation target user who uses the service.

[0043] Accordingly, the information processing apparatus 10 can improve the recommendation accuracy even for new users or services with a small amount of data. In addition, the information processing apparatus 10 extracts the service characteristic vector from the time-series data of the users using the service, and extracts the personal characteristic vector of the user from the service-independent time-series data, and uses these vectors, so that, for example, even in a cross-domain scenario where available user-item information varies greatly across domains, the recommendation accuracy can be improved.

[0044] Further, the information processing apparatus 10 compares the personal characteristic vectors of the respective users extracted by the personal characteristic extraction unit, estimates the similarity, and further outputs the personal characteristic vector of a user whose similarity to the recommendation target user is equal to or greater than a predetermined threshold. Accordingly, the information processing apparatus 10 can further improve the recommendation accuracy by also utilizing the personal characteristic vectors of other similar users.

[0045] [Second Embodiment] In the first embodiment described above, the case where the information processing apparatus 10 collects time-series data has been described, but there are also services that cannot output data to the outside from the viewpoint of privacy and the like. In such a case, the storage unit and each functional unit may be deployed on the server side that provides the service.

[0046] For example, as illustrated in Figure 9, in the recommendation system 1A according to the second embodiment, the server 20X that provides service X may include a storage unit 23X, an individuality extraction unit 24X, a similarity estimation unit 25X, and a service characteristic estimation unit 26X. In addition, the server 20X may appropriately acquire necessary information from the storage unit 13 of the information processing apparatus 10. Figure 9 is a diagram showing a configuration example of the recommendation system according to the second embodiment.

[0047] By doing this, in the recommendation system 1A according to the second embodiment, the server side extracts the personal characteristic vector and the service characteristic vector, and uses the data of the information processing apparatus 10, so that privacy can be protected, and the recommendation performance can be improved with less data compared with the prior art.

[0048] [Third Embodiment] Even if the items of a service are homogeneous, it does not mean that the users or services are similar. In addition, in order to improve the recommendation accuracy, it is also important to grasp services with high service similarity. Therefore, the service itself may be recommended using the similarity of the service characteristic vectors. Accordingly, in the third embodiment, a case where the service itself is recommended using the similarity of service characteristic vectors will be described.

[0049] Here, the information processing apparatus 10A according to the third embodiment will be described with reference to Figure 10. Figure 10 is a block diagram illustrating the configuration of the information processing apparatus according to the third embodiment. As illustrated in Figure 10, the information processing apparatus 10A differs from the information processing apparatus 10 in that it further includes a service similarity estimation unit 12f.

[0050] The service similarity estimation unit 12f compares the service characteristic vectors extracted by the service characteristic extraction unit 12d and estimates their similarity. For example, the service similarity estimation unit 12f reads service characteristic vectors from the service characteristic storage unit 13c, compares two of these service characteristic vectors to estimate their similarity, and stores the estimated similarity in the similarity storage unit 13d. The service similarity estimation unit 12f then repeatedly performs the process of estimating similarity for all service combinations. Note that the method for estimating similarity is not limited and any method may be used.

[0051] The output unit 12e further outputs service characteristic vectors whose similarity to the services used by the recommended user is above a predetermined threshold. Alternatively, for example, the output unit 12e may recommend to the user services with high service similarity, based on the service characteristic vectors whose similarity to the services used by the recommended user is above a predetermined threshold.

[0052] Here, an example of processing by the information processing device 10A of the third embodiment will be explained using Figure 11. Figure 11 is a diagram illustrating an example of processing by the information processing device of the third embodiment. As illustrated in Figure 11, the information processing device 10A of the recommendation system 1B according to the third embodiment compares the service characteristic vectors extracted by the service characteristic extraction unit 12d and estimates the similarity. The information processing device 10A then outputs the personality vector of user A, the personality vectors of users with a high similarity to user A, the service characteristic vector of service X, and the service characteristic vectors of other services with a high similarity to service X. Alternatively, for example, the information processing device 10A may recommend service Y, whose service characteristic vector has a similarity to the service used by the user being recommended that is above a predetermined threshold, to the user as a service with high similarity.

[0053] In this way, the recommendation system 1A according to the third embodiment can improve recommendation accuracy by outputting service characteristic vectors of other services that are highly similar to the service being used by the user. Furthermore, the recommendation system 1A according to the third embodiment can recommend services themselves by utilizing the similarity of the service characteristic vectors.

[0054] [Other Embodiments] In the embodiments described above, recommendations are made using two vectors: a service characteristic vector and an individual characteristic vector. However, the weighting system may be used to change which of the user's preferences and the service's tendencies are given more weight. For example, as illustrated in Figure 12, the weighting ratio may be automatically changed based on feedback from the user and the service provider. Figure 12 illustrates an example of processing by an information processing device of another embodiment.

[0055] [Program] In one embodiment, the information processing device can be implemented by installing an information processing program that performs the above processing as packaged software or online software on a desired computer. For example, by having the information processing device execute the above information processing program, the information processing device can be made to function as an information processing device. The information processing device referred to here includes desktop or notebook personal computers. In addition, the information processing device also includes smartphones, tablet terminals, etc.

[0056] Figure 13 shows an example of a computer that executes an information processing program. Computer 1000 has, for example, memory 1010 and CPU 1020. Computer 1000 also has a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0057] Memory 1010 includes ROM (Read Only Memory) 1011 and RAM (Random Access Memory) 1012. ROM 1011 stores, for example, a boot program such as BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1090. The disk drive interface 1040 is connected to the disk drive 1100. For example, a removable storage medium such as a magnetic disk or optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to, for example, a mouse 1110 and a keyboard 1120. The video adapter 1060 is connected to, for example, a display 1130.

[0058] The hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. That is, the programs that define each process of the information processing device are implemented as program modules 1093 in which executable code for the computer is written. The program modules 1093 are stored, for example, in the hard disk drive 1090. For example, a program module 1093 for performing processes similar to the functional configuration of the information processing device is stored in the hard disk drive 1090. Note that the hard disk drive 1090 may be replaced by an SSD.

[0059] Furthermore, the configuration data used in the processing of the above-described embodiment is stored as program data 1094 in, for example, memory 1010 or hard disk drive 1090. The CPU 1020 then reads the program module 1093 and program data 1094 stored in memory 1010 or hard disk drive 1090 into RAM 1012 as needed and executes the processing of the above-described embodiment.

[0060] Furthermore, the program module 1093 and program data 1094 are not limited to being stored in the hard disk drive 1090; for example, they may be stored in a removable storage medium and read by the CPU 1020 via a disk drive 1100 or the like. Alternatively, the program module 1093 and program data 1094 may be stored in another computer connected via a network (LAN (Local Area Network), WAN (Wide Area Network), etc.). The program module 1093 and program data 1094 may then be read by the CPU 1020 from the other computer via a network interface 1070.

[0061] Furthermore, the functions of the elements disclosed herein may be implemented using circuits or processing circuitry that include general-purpose processors, special-purpose processors, integrated circuits, ASICs (Application Specific Integrated Circuits), FPGAs (Field Programmable Gate Arrays), conventional circuits, and / or combinations thereof that are programmed using one or more programs stored in one or more memories, or otherwise configured to perform the disclosed functions. A processor is considered processing circuitry or circuitry because it includes transistors and other circuits. A processor may be a programmed processor that executes programs stored in memory. In this disclosure, a circuit, unit, or means is hardware that performs the enumerated functions, or hardware programmed to perform the enumerated functions. Hardware may be any hardware disclosed herein that is programmed or configured to perform the enumerated functions.

[0062] Furthermore, there is memory for storing computer programs that include computer instructions. These computer instructions provide logic and routines that enable hardware (e.g., processing circuitry or circuitry) to perform the methods disclosed herein. This computer program can be implemented in commonly known forms, such as computer-readable storage media, computer program products, memory devices, recording media such as CD-ROMs and DVDs, and / or memory of FPGAs and ASICs.

[0063] Although embodiments applying the invention made by the present inventors have been described above, the present invention is not limited by the descriptions and drawings that constitute part of the disclosure of the present invention in this embodiment. That is, all other embodiments, examples, and operational techniques made by those skilled in the art based on this embodiment are included in the scope of the present invention.

[0064] 10 Information processing device 11 Communication processing unit 12 Control unit 12a Collection unit 12b Personal characteristics extraction unit 12c Similarity estimation unit 12d Service characteristics extraction unit 12e Output unit 12f Service similarity estimation unit 13 Storage unit 13a Time series data storage unit 13b Personal characteristics storage unit 13c Service characteristics storage unit 13d Similarity storage unit 20X, 20Y Server

Claims

1. An information processing device comprising: a collection unit that collects time-series data of each user using each service; a personal characteristic extraction unit that extracts a personal characteristic vector representing the characteristics of each user from the time-series data collected by the collection unit; a service characteristic extraction unit that extracts a service characteristic vector representing the characteristics of users using each service from the time-series data collected by the collection unit; and an output unit that outputs the personal characteristic vector of the recommended user and the service characteristic vector related to the service as information for recommending to a recommended user who uses the service.

2. The information processing apparatus according to claim 1, further comprising a similarity estimation unit that compares the personal characteristic vectors of each user extracted by the personal characteristic extraction unit and estimates the degree of similarity, wherein the output unit further outputs the personal characteristic vectors of users whose similarity to the recommended user is above a predetermined threshold.

3. The information processing apparatus according to claim 1 or 2, further comprising a service similarity estimation unit that compares service characteristic vectors extracted by the service characteristic extraction unit and estimates the degree of similarity, wherein the output unit further outputs service characteristic vectors whose similarity to the service used by the recommended user is greater than or equal to a predetermined threshold.

4. An information processing method performed by an information processing device, wherein the computer performs the following processes: collects time-series data of each user using each service; extracts a personal characteristic vector representing the characteristics of each user from the collected time-series data; extracts a service characteristic vector representing the characteristics of each user using each service from the collected time-series data; and outputs the personal characteristic vector of the recommended user and the service characteristic vector related to the service as information for recommending to the recommended user who uses the service.