Matching system and system for improving matching accuracy

Abstract

To improve the accuracy of matching between demand and supply using AI.SOLUTION: A matching system includes: a holding unit (2) which holds a history of demand data for AI learning and a history of supply data for AI learning, or holds a history of AI models for matching between demand and supply in addition to the above histories, in a blockchain; an evaluation unit (3) which evaluates, based on the held histories, authenticity of each of the demand data and the supply data, or each of the AI models in addition to the demand and supply data; an assignment unit (4) which assigns a predetermined kind of incentive or penalty according to the evaluated authenticity, to each data or to each of the AI models in addition to the data; and a learning unit (21) which creates an AI model based on the demand data and the supply data to perform the matching.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a technical field of a matching system for matching demand and supply by AI, applicable to various matchings such as job hunting matching between a taxi company on the demand side that recruits taxi drivers, for example, and candidates or applicants on the supply side who apply for it, matching between an automobile company or vehicle type and a car user, matching between a hospital or doctor and a patient, and matchmaking matching, and a matching accuracy improvement system for improving the accuracy of matching in such a matching system.

Background Art

[0002] As such a matching system, a matching system using a so-called traditional AI learning system of supervised learning method, unsupervised learning method, or reinforcement learning method has been developed and already put into practical use (see Patent Document 1). In the AI learning in such a system, the quality of the demand data, supply data, and the authenticity of the AI model are basically entrusted to the demanders, suppliers, etc. who are the providers of those data and models.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, according to the aforementioned background art, although it is said to facilitate matching, for example, between the supply side such as a taxi company and the demand side such as potential users of taxis, there is a possibility that demand data or supply data of uncertain authenticity is being used, such as data that is intentionally or negligently different from the facts being provided on the demand side or the supply side. In addition, it is difficult to evaluate the authenticity itself. Furthermore, there is a technical problem that an AI model of uncertain authenticity, such as when the provider or vendor is unclear, may be created as a learning result, and in addition, it is difficult to evaluate the authenticity itself.

[0005] More specifically, generally, on the demand side such as the company side that conducts recruitment, there is a motivation to present itself as well as possible, and on the supply side such as the applicant side, there is a motivation to present itself as well as possible. For this reason, the authenticity of the data used as the basis for AI learning for matching is likely to be suspect by its nature. On the other hand, from the perspective of the provider who offers the model, the motivation to invest funds in data collection and model creation to enhance authenticity is not strong, nor is the motivation to disclose the provider of data and models that may also be confidential information. As a result, it is not very likely that learning results of high authenticity will be provided. In addition, for data and models where the provider or source is unclear, it is difficult to detect forgery, whether initial or subsequent, that is, there is also a technical problem that it is easy to be forged.

[0006] Generally speaking, if the demand data, supply data, and AI model have such low authenticity, there is a technical problem that the accuracy of the final matching will ultimately be low.

[0007] An object of the present invention is to provide a matching system that can improve the matching accuracy by improving the authenticity of demand data and supply data, or in addition to these, improving the authenticity of the AI model, and a matching accuracy improvement system that can improve the matching accuracy in the matching system.

Means for Solving the Problem

[0008] One aspect of the matching system according to the present invention is to solve the above problems. A holding unit that holds, in a blockchain, the history of demand data for AI learning related to demand and the history of supply data for AI learning related to supply, or in addition to the history, the history of an AI model for matching the demand and the supply; an evaluation unit that evaluates the authenticity of each of the demand data, each of the supply data, or each of the AI models in addition to each of the data, based on the held history; an awarding unit that awards a predetermined type of incentive or penalty according to the evaluated degree of authenticity to the provider of each of the data or each of the AI models in addition to each of the data; and a learning unit that performs the matching by creating the AI model based on the demand data and the supply data.

[0009] One aspect of the matching accuracy improvement system according to the present invention is a matching accuracy improvement system for improving the accuracy of matching in a matching system that performs matching by creating an AI model based on demand data for AI learning related to demand and supply data for AI learning related to supply. The system includes a holding unit that holds, in a blockchain, the history of the demand data and the history of the supply data, or in addition to the history, the history of the AI model; an evaluation unit that evaluates the authenticity of each of the demand data, each of the supply data, or each of the AI models in addition to each of the data, based on the held history; and an awarding unit that awards a predetermined type of incentive or penalty according to the evaluated degree of authenticity to the provider of each of the data or each of the AI models in addition to each of the data.

Effect of the Invention

[0010] According to one aspect of the matching system according to the present invention or one aspect of the matching accuracy improvement system according to the present invention, by giving incentives or penalties to the providers of demand data, supply data, and the AI model respectively, as the AI learning progresses, the authenticity of these data and models can be gradually enhanced, and the matching accuracy can be improved.

[0011] Such an effect of the present invention will be made clearer by the embodiments of the invention described below.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Embodiments for Carrying Out the Invention

[0013] First, with reference to FIG. 1, the overall configuration of the matching system according to the embodiment will be described. This embodiment is applicable to various matchings such as job hunting matching between a taxi company on the demand side recruiting taxi drivers and candidates or applicants on the supply side applying for it, matching between an automobile company or vehicle type and car users, matching between a hospital or doctor and a patient, matching between a caregiver and an elderly person, matchmaking, adoption matching, matching between a company and a client, and matching between companies in corporate succession. In any case, it is possible to appropriately improve the matching accuracy.

[0014] As shown in FIG. 1, the matching system includes a matching accuracy improvement system 1, a demand data Dm providing unit 11, a supply data Sl providing unit 12, and an AI model Mn providing unit 21.

[0015] The matching accuracy improvement system 1 is configured to receive the provision of demand data Dm and supply data Sl from a demand data Dm providing unit 11 and a supply data Sl providing unit 12 via a network, respectively. Further, the matching accuracy improvement system 1 is configured to receive the provision of the AI model Mn from an AI model Mn providing unit 21 via a network. The matching accuracy improvement system 1 that performs these centralized processing or distributed processing, a plurality m of demand data Dm providing units 11 and a plurality l of supply data Sl providing units 12 at each data provider, and a plurality n of AI model Mn providing units 21 at each model provider are accommodated in a network such as the Internet.

[0016] The demand data Dm (D1, D2,..., Dm) is data related to the demand side required for the AI learning for various matchings according to its use. On the other hand, the supply data Sl (S1, S2,..., Sl) is data related to the supply side required for the AI learning for various matchings according to its use.

[0017] For example, in the case of job hunting matching between a taxi company on the demand side recruiting taxi drivers and candidates or applicants on the supply side applying for it, the demand data Dm is data having various formats that can be handled by an AI model Mn providing unit (i.e., an AI learning unit) 21 related to recruitment conditions indicating salaries, treatments, etc. that the company can offer, company information indicating the company's performance, scale, etc. On the other hand, the supply data Sl is data having various formats that can be handled by an AI model Mn providing unit (i.e., an AI learning unit) 21 related to application conditions indicating salaries, treatments, etc. that the applicant hopes for, personal information indicating the applicant's personal history and abilities within the range permitted by the applicant.

[0018] The demand data Dm providing unit 11 and the supply data Sl providing unit 12 are various computer-mounted devices and various computer devices. The demand data Dm and the supply data Sl collected therein are provided to the matching accuracy improvement system 1 and the AI model Mn providing unit 21 via a network in their original form or in a data format obtained by performing a predetermined type of processing.

[0019] The AI model Mn providing unit 21 includes one or more various computer devices that perform centralized processing or distributed processing. It receives the provision of the demand data Dm and the supply data Sl from the demand data Dm providing unit 11 and the supply data Sl providing unit 12, and performs AI learning (in other words, machine learning or deep learning) such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, and generative AI. Furthermore, the AI model Mn obtained as a result of the learning is configured to be provided to the matching accuracy improvement system 1 sequentially or collectively via a network at regular or irregular intervals. The AI model Mn (M1, M2,..., Mn) generated here is various models learned by the AI learning according to its use or purpose.

[0020] The AI model Mn providing unit 21 (in other words, the AI learning unit) may be configured using a neural network that performs efficient AI learning through representation learning, transfer learning, feature selection, fine-tuning, ensemble learning, etc., or may be configured as a generative AI that learns the patterns and relationships of the demand data Dm and the supply data Sl and generates content data different from the demand data Dm and the supply data Sl.

[0021] The matching accuracy improvement system 1 includes a memory, a processor, etc., and is configured as a centralized system that performs centralized processing or a distributed system that performs distributed processing, and includes a holding unit 2, an evaluation unit 3, and an imparting unit 4.

[0022] The holding unit 2 includes the memories of a plurality of computers housed in the network, and is configured using existing or future-upgraded blockchain technology. The holding unit 2 sequentially or at appropriate timings holds the history of the demand data Dm, the history of the supply data Sl, and the history of the AI model Mn in the blockchain.

[0023] The evaluation unit 3 is configured to evaluate the authenticity of each of the demand data Dm, each of the supply data Sl, and each of the AI models Mn based on the history held by the holding unit 2. The evaluation unit 3 may be configured to perform the evaluation by AI learning that performs authenticity evaluation using the input data as a history. Such an evaluation unit 3 performs AI learning such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, generative AI, etc., and sequentially or collectively provides the learned AI models to the imparting unit 4 via the network. The evaluation unit 3 may be configured using a neural neural network that performs efficient AI learning through representation learning, transfer learning, feature selection, fine-tuning, ensemble learning, etc., or may be configured as a generative AI that learns the patterns and relationships of the history and generates content data different from the history. The evaluation unit 3 may be configured to perform a Gold evaluation when the authenticity is the highest, a Silver evaluation when the authenticity is very high, and a Bronze evaluation when the authenticity is moderately high, for example.

[0024] The awarding unit 4 adds or associates, in the form of NFTs, ranking data DRm (DR1, DR2,..., DRm) indicating the ranking of demand data Dm, ranking data SRl (SR1, SR2,..., SRl) indicating the ranking of supply data Sl, and ranking data MRn (MR1, MR2,..., MRn) indicating the ranking of AI model Mn, to the corresponding demand data Dm, supply data Sl, and AI model Mn respectively, according to the degree of authenticity evaluated by the evaluation unit 3. For example, NFTs such as Gold NFT indicating the highest authenticity, Silver NFT indicating very high authenticity, and Bronze NFT indicating relatively high authenticity are added as ranking data. That is, the awarding unit 4 is configured to award incentives or penalties according to the degree of authenticity to the providers of each demand data Dm, each supply data Sl, and each AI model Mn in the form of awarding ranking data. Further, instead of or in addition to providing the ranking data DRm, SRl, and MRn that give such "endorsements" or status to the corresponding providers, the awarding unit 4 may be configured to award convertible points or gift points to the corresponding providers.

[0025] On the other hand, the linking data DRm, SRl, and MRn thus awarded by the awarding unit 4 are stored in the holding unit 2 in the blockchain together with the history.

[0026] Next, with reference to the sequence chart of FIG. 2, an example of the process of awarding incentives in the matching system according to this embodiment will be described.

[0027] In FIG. 2, first, the demand data Dm providing unit 11 collects the demand data Dm (step S1). The demand data Dm providing unit 11 sequentially, regularly, irregularly, or at a predetermined timing provides the collected demand data Dm to the AI model Mn providing unit 21 and the matching accuracy improvement system 1 (step S2). In parallel with or before and after these, the supply data Sl providing unit 12 collects the supply data Sl (step S3). The supply data Sl providing unit 12 sequentially, regularly, irregularly, or at a predetermined timing provides the collected supply data Sl to the AI model Mn providing unit 21 and the matching accuracy improvement system 1 (step S4). The data provision here is efficiently performed in an environment where the data source (i.e., the demand data Dm providing unit 11 and the supply data Sl providing unit 12) and the data destination (the AI model Mn providing unit 21) are accommodated in the same network.

[0028] In response to this, the AI model Mn providing unit 21 functions as an AI learning unit and executes a preset type of AI learning, such as AI learning for employment matching in a taxi company, using the demand data Dm and the supply data Sl as inputs (step S5). In parallel with or before and after this, the matching accuracy improvement system 1, by its holding unit 2, holds the demand data Dm and the supply data Sl in the blockchain as part of the data history (step S6). Even if the history of the demand data Dm and the supply data Sl is held in the database possessed by the holding unit 2 without using the blockchain here, the effect of improving the authenticity by granting incentives or penalties according to the present embodiment can be correspondingly obtained. However, by using the blockchain technology to hold here, the reliability of the ranking related to the demand data Dm and the supply data Sl can be enhanced and the possibility of forgery can be eliminated, and the value of the ranking can be increased. Finally, the effect of improving the authenticity in the present embodiment becomes more remarkable.

[0029] Subsequently, the AI model Mn providing unit 21 provides the AI model Mn obtained by AI learning in step S5 to the matching accuracy improvement system 1 sequentially, periodically, irregularly, or at a predetermined timing (step S7). The data provision here is efficiently performed in an environment where the provider (i.e., the AI model Mn providing unit 21) and the recipient (the matching accuracy improvement system 1) are accommodated in the same network.

[0030] In response to this, the matching accuracy improvement system 1 has its holding unit 2 hold the AI model Mn as part of the AI model history in the blockchain (step S8). Even if the history of the AI model Mn is held in the database of the holding unit 2 without using the blockchain here, the effect of improving authenticity by granting incentives or penalties according to the present embodiment can be obtained accordingly. However, by using the blockchain technology to hold it here, the reliability of the ranking related to the AI model can be enhanced and the possibility of forgery can be eliminated, and the value of the ranking can be increased. Finally, the effect of improving authenticity in the present embodiment becomes even more remarkable.

[0031] Subsequently, the matching accuracy improvement system 1 has its evaluation unit 3 evaluate the authenticity of each demand data Dm, each supply data Sl, and each AI model Mn based on the history held in the blockchain (step S9). The evaluation unit 3 may perform the evaluation using various AI learnings as described above.

[0032] Subsequently, the matching accuracy improvement system 1, by the granting unit 4, sequentially, periodically, irregularly, or at a predetermined timing, grants the ranking data DRm related to the demand data Dm and the ranking data SRl related to the supply data Sl obtained by AI learning or the like in step S9 to the demand data Dm providing unit 11, the supply data Sl providing unit 12, and the holding unit 2 in the NFT format (step S10). Here, the holding unit 2 holds the granted ranking data DRm and ranking data SRl on the blockchain. The granting unit 4 converts the ranking data DRm and ranking data SRl into the NFT format and then grants them to the demand data Dm providing unit 11 and the supply data Sl providing unit 12. At this time, preferably as described above, in addition to or instead of the granting of the ranking data DRm and ranking data SRl, points that increase as the ranking gets higher are granted to the source that provides them.

[0033] In parallel with or before and after this, the matching accuracy improvement system 1, by the granting unit 4, sequentially, periodically, irregularly, or at a predetermined timing, grants the ranking data MRn related to the AI model Mn obtained by AI learning or the like in step S9 to the AI model Mn providing unit 21 and the holding unit 2 (step S10). Here, the holding unit 2 holds the granted ranking data MRn on the blockchain. The granting unit 4 converts the ranking data MRm into the NFT format and then grants it to the AI model Mn providing unit 11. At this time, preferably as described above, in addition to or instead of the granting of the ranking data MRm, points that increase as the ranking gets higher are granted to the source that provides it.

[0034] Subsequently, the AI model Mn providing unit 21 refers to the ranking data MRm assigned in step S10 and reflects it in subsequent AI learning (step S11). That is, the higher the ranking indicated by the ranking data, the stronger the objectively obtained "endorsement" for the AI model provided by itself. More preferably, it is considered that a higher score can be obtained. As a result, the motivation to provide a higher-rated AI model and not provide a lower-rated AI model occurs in each AI model Mn provider 21 according to the situation.

[0035] In parallel with or subsequent to this, the demand data Dm providing unit 11 refers to the ranking data DRm assigned in step S10 and reflects it in the processes of data collection to data provision thereafter (step S12). Similarly, the supply data Sl providing unit 12 refers to the ranking data SRl assigned in step S10 and reflects it in the processes of data collection to data provision thereafter (step S13). That is, the higher the ranking indicated by the ranking data, the stronger the objectively obtained "endorsement" for the data provided by itself. More preferably, it is considered that a higher score can be obtained. As a result, the motivation to provide higher-rated data and not provide lower-rated data occurs in each demand data Dm providing unit 11 and each supply data Sl providing unit 12 according to the situation.

[0036] As described in detail above, the history of demand data and supply data, or in addition thereto, the history of the AI model is held in the blockchain, and the authenticity of each of the demand data and supply data, or in addition thereto, each of the AI models is evaluated. Depending on the degree of authenticity evaluated here, a predetermined type of incentive or penalty is given to the provider of each of the demand data, supply data, or in addition thereto, each of the AI models. By these means, it becomes possible to generate a motivation to enhance the authenticity with respect to the demand data, supply data, or in addition thereto, the AI model for the providers such as demanders and suppliers. At this time, since it is held in the blockchain, it is substantially impossible to prevent forgery from the beginning or retrospectively. As a result, as the AI learning related to matching progresses, the authenticity is gradually improved, and ultimately, it leads to an improvement in the accuracy of matching.

[0037] Supplementary Note Regarding the embodiments described above, the following supplementary notes are further disclosed.

[0038] [Supplementary Note 1] The matching system according to Supplementary Note 1 of the present invention includes a holding unit that holds the history of demand data for AI learning related to demand and the history of supply data for AI learning related to supply, or in addition to the history, the history of the AI model for matching the demand and the supply in the blockchain, an evaluation unit that evaluates the authenticity of each of the demand data and supply data, or in addition to the data, each of the AI models based on the held history, an awarding unit that awards a predetermined type of incentive or penalty according to the evaluated degree of authenticity to the provider of each of the data or in addition to the data, each of the AI models, and a learning unit that performs the matching by creating the AI model based on the demand data and the supply data.

[0039] According to the matching system described in Supplementary Note 1, since the demand data and supply data, whose authenticity is uncertain as they are respectively entrusted to the demanders and suppliers who are the data providers, or the history of the AI models, whose authenticity is uncertain such as the provider or vendor of the AI model being unclear, are stored in a blockchain by the storage unit. Here, it is made practically impossible to falsify data or models either initially or subsequently. By the evaluation unit, for each of the demand data and each of the supply data, the degree of authenticity or classification by degree (in other words, "ranking of data") is specified, and for each of the AI models, the degree of authenticity or classification by degree (in other words, "ranking of models") is specified as an evaluation. Further, by the awarding unit, as the evaluation result by the evaluation unit of the authenticity of each of the demand data and each of the supply data, or each of the AI models, when the authenticity is high, an incentive corresponding to the degree of the highness is given to the provider, or when the authenticity is low, a penalty corresponding to the degree of the lowness is given to the provider. The motivation of the provider to improve the authenticity is enhanced not only to obtain the incentive or avoid the penalty, but also to obtain "endorsement" or "social status" or "business status" based on a highly reliable evaluation result. As a result, as the AI learning progresses, the authenticity of the data and models is gradually improved, and finally the accuracy of the matching is improved.

[0040] [Supplementary Note 2] The matching system described in Supplementary Note 2 according to the present invention is the matching system described in Supplementary Note 1, characterized in that the awarding unit awards ranking data indicating the ranking corresponding to the degree as at least part of the incentive or penalty in the form of an NFT.

[0041] According to the matching system described in Supplementary Note 2 of the present invention, the granting unit adds or associates ranking data indicating a ranking according to the degree of authenticity evaluated by the evaluation unit, to each of the demand data and the supply data, or to each of these data in addition to AI models, using NFTs. For example, "endorsed" items such as gold-rank demand data or supply data, or AI models with low or negative points, are granted to the provider. At this time, it is substantially impossible to falsify the ranking data in NFT format corresponding to the evaluation result. Since such high reliability is based on the demand data, supply data, or AI models, and further on the evaluation results, the motivation of the provider to improve authenticity is further enhanced.

[0042] [Supplementary Note 3] The matching system described in Supplementary Note 3 of the present invention is the matching system according to Supplementary Note 1 or 2, characterized in that the granting unit grants convertible points or gift points according to the degree as at least part of the incentive or penalty.

[0043] According to the matching system described in Supplementary Note 2 of the present invention, the granting unit grants, for example, convertible points or gift points according to the evaluated degree of authenticity or the classification according to the degree, to the provider. For example, the higher the authenticity, the higher the amount of convertible points or gift points granted to the provider. Therefore, the provider is more directly motivated to increase authenticity.

[0044] [Supplementary Note 4] The matching system described in Supplementary Note 4 of the present invention is the matching system according to any one of Supplementary Notes 1 to 3, characterized in that the evaluation unit evaluates the authenticity by AI.

[0045] According to the matching system described in Supplementary Note 4 of the present invention, in the evaluation unit, the authenticity of the data and models as described above is evaluated by AI. Therefore, as the AI learning in the evaluation unit progresses, the number of samples or the scale of learning related to the data and models for evaluating authenticity increases, and more appropriate authenticity evaluation becomes possible, leading to an improvement in matching accuracy.

[0046] [Supplementary Note 5] The matching accuracy improvement system described in Supplementary Note 5 of the present invention is a matching accuracy improvement system for performing matching by creating an AI model based on demand data for AI learning related to demand and supply data for AI learning related to supply in a matching system. The matching accuracy improvement system includes: a holding unit that holds the history of the demand data and the history of the supply data, or in addition to the history, the history of the AI model, on a blockchain; an evaluation unit that evaluates the authenticity of each of the demand data, each of the supply data, or each of the AI models in addition to each of the data, based on the held history; and an awarding unit that awards a predetermined type of incentive or penalty according to the degree of the evaluated authenticity to the source of each of the data, or each of the AI models in addition to each of the data.

[0047] According to the matching accuracy improvement system described in Supplementary Note 5 of the present invention, if the matching accuracy improvement system is adopted in a system that matches the demand side and the supply side by AI, the same effects as those of the matching system described in Supplementary Note 1 above are achieved.

[0048] The present invention can be appropriately modified within the scope not contrary to the gist or idea of the invention that can be read from the claims and the entire specification, and the matching system and the matching accuracy improvement system with such modifications are also included in the technical idea of the present invention.

Explanation of Signs

[0049] Matching accuracy improvement system... 1 Holding unit... 2 Evaluation unit... 3 Granting unit... 4 Demand data Dm providing unit... 11 Supply data Sl providing unit... 12 AI model Mn providing unit... 21

Claims

1. A holding unit that holds, on a blockchain, the history of demand data for AI learning related to demand and the history of supply data for AI learning related to supply, or in addition to the history, the history of an AI model for matching the demand and the supply; An evaluation unit that evaluates the authenticity of each of the demand data, each of the supply data, or each of the AI models in addition to each of the data, based on the held history; An awarding unit that awards a predetermined type of incentive or penalty according to the degree of the evaluated authenticity to the provider of each of the data or each of the AI models in addition to each of the data; A learning unit that performs the matching by creating the AI model based on the demand data and the supply data; A matching system comprising the above.

2. The matching system according to claim 1, wherein the awarding unit awards, as at least part of the incentive or penalty, ranking data indicating a ranking according to the degree, in the form of an NFT.

3. The matching system according to claim 1 or 2, wherein the awarding unit awards, as at least part of the incentive or penalty, convertible points or gift points according to the degree.

4. The matching system according to any one of claims 1 to 3, wherein the evaluation unit evaluates the authenticity by AI.

5. A matching accuracy improvement system for improving the accuracy of matching in a matching system that performs matching by creating an AI model based on demand data for AI learning related to demand and supply data for AI learning related to supply, comprising: A holding unit that holds, on a blockchain, the history of the demand data and the history of the supply data, or in addition to the history, the history of the AI model; An evaluation unit that evaluates the authenticity of each of the demand data, each of the supply data, or each of the AI models in addition to each of the data, based on the held history; An awarding unit that awards a predetermined type of incentive or penalty according to the degree of the evaluated authenticity to the provider of each of the data or each of the AI models in addition to each of the data; The matching accuracy improvement system is characterized by comprising the above.

Images