Information processing device, information processing method, and program

Abstract

[Problem] To implement intuitively understandable analysis of differences between the behavior of an analysis subject and the behavior of another task performer. [Solution] To provide an information processing device comprising an analysis unit that, on the basis of behavior record data indicating records of a behavior related to a prescribed task, analyzes tendencies in preferences related to the behavior. The analysis unit generates, on the basis of the behavior record data, a similarity map in which similarities in behavior between each of multiple task performers, which include an analysis subject, is expressed on a two-dimensional plane, and expresses the strength of preferences related to a selected type of behavior as a heat map in the similarity map.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] It is very important to properly analyze the behavior of a task performer who performs a certain task. For this reason, in recent years, many mechanisms have been proposed to automate or assist such analysis. For example, Patent Document 1 proposes a mechanism that analyzes the investment behavior of a mutual fund and provides a rating based on the results of the analysis. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2009-245368 Summary of the Invention [Problem to be solved by the invention]

[0004] Furthermore, if the above-described analysis is performed with high accuracy, it may be possible to use the results of the analysis to reproduce tasks that have already been performed or tasks that can be performed by a third party. [Means for solving the problem]

[0005] According to one aspect of the present disclosure, there is provided an information processing device including an analysis unit that generates a similarity map that represents the similarity between actions by each of a plurality of task performers on a two-dimensional plane based on behavioral performance data that indicates the performance of actions related to a predetermined task, and the analysis unit outputs behavior reproduction data having the same data structure as the behavioral performance data based on coordinate data that indicates the position of the plot in the similarity map.

[0006] According to another aspect of the present disclosure, there is provided an information processing method including: a processor generating, based on behavioral performance data indicating the performance of behavior related to a predetermined task, a similarity map that represents on a two-dimensional plane the similarity between the behaviors of each of a plurality of task performers; and outputting, based on coordinate data indicating the position of the plot in the similarity map, behavioral reproduction data having the same data structure as the behavioral performance data.

[0007] According to another aspect of the present disclosure, there is provided a program for causing a computer to function as an information processing device, comprising an analysis unit that generates, based on behavioral performance data indicating the performance of behavior related to a predetermined task, a similarity map that represents on a two-dimensional plane the similarities between the behaviors of each of a plurality of task performers, and the analysis unit outputs behavior reproduction data having the same data structure as the behavioral performance data, based on coordinate data indicating the position of the plot in the similarity map. [Brief explanation of the drawings]

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Embodiments for Carrying Out the Invention

[0009] Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In the present specification and drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant description is omitted.

[0010] Note that the description will be made in the following order. 1. First Embodiment 1.1. Overview 1.2. Example of System Configuration 1.3. Details of Analysis 1.4. Example of User Interface 1.5. Flow of Processing 2. Second Embodiment 2.1. Generation of Action Reproduction Data 2.2. Expression of Action Reproduction Data Using Existing Action Performance Data 3. Example of Hardware Configuration 4. Summary

[0011] <1. Embodiment> <<1.1. Overview>> First, the first embodiment of the present disclosure will be described. As described above, it is very important to appropriately analyze the actions of task executors who execute predetermined tasks, regardless of the industry type.

[0012] However, when the task executed by the task executor is highly specialized and the analyst does not have the same specialized knowledge about the task as the task executor, it may be difficult to perform an appropriate analysis.

[0013] Here, as an example, it is assumed that in a certain fund, an analysis of the actions of an investment fund (hereinafter simply referred to as a fund) is to be performed.

[0014] An analyst belonging to the fund, for example, shall perform an analysis of the actions of the analysis target, taking as the analysis targets the funds with which contracts have already been concluded and new funds that are candidates for future contracts.

[0015] However, here, when the analyst belonging to the fund does not have the same specialized knowledge as the analysis target, the analyst cannot grasp how the actions of the analysis target differ from those of other funds, and it is difficult to appropriately evaluate the analysis target.

[0016] The technical idea according to the first embodiment of the present disclosure was conceived by paying attention to the above points, and realizes an analysis that can intuitively grasp the difference between the actions of the analysis target and the actions of other task performers.

[0017] For this purpose, the analysis apparatus 20 according to the first embodiment of the present disclosure includes an analysis unit 210 that analyzes the preference tendency related to the action based on the action performance data indicating the performance of the action related to a predetermined task.

[0018] Further, the analysis unit 210 according to the present embodiment generates a similarity map that represents the similarity of the above actions by each of a plurality of task performers including the analysis target on a two-dimensional plane based on the action performance data, and represents the intensity of the preference related to the selected action type in the form of a heat map in the similarity map, which is one of the features.

[0019] In the following, the case where the predetermined task is asset management and the action related to the predetermined task is a trading transaction of financial products will be mainly described as an example. In this case, the task performer including the analysis target may be a fund.

[0020] Here, an example of the similarity map generated by the analysis unit 210 according to the present embodiment will be described. FIG. 1 is a diagram showing an example of the similarity map according to the present embodiment.

[0021] For example, the analysis unit 210 according to the present embodiment can generate a similarity map M1 as shown in FIG. 1 based on the weight (active weight) information of financial products that changes due to trading behavior.

[0022] In the similarity map M1 illustrated in FIG. 1, the characteristics of the trading behavior by each of Funds A to G are plotted using different symbols for each fund.

[0023] For example, in the case of the example shown in FIG. 1, the characteristics of the trading behavior by Fund A are plotted using a round symbol, and the characteristics of the trading behavior by Fund B are plotted using a pentagonal symbol.

[0024] Also, in the similarity map according to the present embodiment, the similarity of the characteristics of each trading behavior is expressed as the distance on the map.

[0025] For example, in the example shown in FIG. 1, both Fund A and Fund B may be funds that conduct trading behavior emphasizing a value strategy. In this case, the plots showing the characteristics of the trading behavior of Fund A and Fund B can be arranged at a close distance in the similarity map M1.

[0026] Similarly, the plots showing the characteristics of the trading behavior of Fund C and Fund D, which emphasize a large-cap growth strategy, can be arranged at a close distance in the similarity map M1.

[0027] Similarly, the plots showing the characteristics of the trading behavior of Fund E and Fund F, which emphasize a small-cap growth strategy, can be arranged at a close distance in the similarity map M1.

[0028] In this way, the similarity map generated by the analysis unit 210 according to this embodiment makes it possible to visually express the similarity in the characteristics of the trading behavior of funds using distance.

[0029] This allows analysts who refer to the similarity map to intuitively grasp funds with similar trading behavior characteristics and, conversely, funds with dissimilar trading behavior characteristics, making it possible to use this information in fund evaluations.

[0030] Furthermore, the analysis unit 210 according to this embodiment may represent, in the similarity map, changes in the behavior of each of a plurality of task performers over time.

[0031] For example, each plot in the similarity map M1 shown in Figure 1 indicates the characteristics of the trading behavior of the corresponding fund in a certain month. As an example, the characteristics of Fund A in a certain month are depicted by a black circle or a white circle in the similarity map M1.

[0032] In the example shown in FIG. 1, the white circles may represent the characteristics of Fund A's trading behavior in the most recent month, and the black circles may represent the characteristics of Fund A's trading behavior in the past months.

[0033] In this way, the analysis unit 210 according to this embodiment may express how the characteristics of a fund's trading behavior change over time by connecting plots relating to the same fund with lines in chronological order and distinguishing plots relating to the most recent month from plots for previous months.

[0034] Using the above expression, it is possible to analyze how the characteristics of a fund's trading behavior change over time.

[0035] For example, in the case of an example shown in FIG. 1, between the sixth plot and the seventh plot counted from the most recent past of Fund C, the distance in the similarity map M1 is widely spaced. In this case, the analyst can infer that there was a significant change in the trading behavior of Fund C between the month corresponding to the sixth plot and the month corresponding to the seventh plot.

[0036] Also, according to the time-series representation as shown in FIG. 1, it is also possible to analyze how the characteristics of the trading behavior between certain funds change over time.

[0037] For example, in the case of an example shown in FIG. 1, each plot related to Fund A and Fund B that emphasizes the value strategy is generally arranged at a relatively close distance in the time series. In this case, the analyst can infer that Fund A and Fund B adopted similar trading strategies in response to past market changes and the like.

[0038] On the other hand, in the case of an example shown in FIG. 1, each plot related to Fund E and Fund F that emphasizes the small-cap growth strategy has a widening distance over time. In this case, the analyst can infer that Fund E and Fund F adopted dissimilar trading strategies in response to past market changes and the like.

[0039] As described above, an example of the similarity map according to the present embodiment has been described. According to the similarity map as described above, it is possible to visually and intuitively perform an analysis regarding the similarity of trading behavior between funds, the change over time of the trading behavior of a certain fund, and the like.

[0040] In addition, the analysis unit 210 according to the present embodiment may represent the intensity of preference related to the selected type of action in the form of a heat map in the similarity map as described above.

[0041] Here, the type of behavior can be appropriately set depending on the target task. For example, if the task is asset management, the types of behavior according to this embodiment include holding financial products of a predetermined brand, holding financial products corresponding to a predetermined industry, holding financial products corresponding to a predetermined factor, etc.

[0042] In other words, when analyzing trading behavior by funds, the analysis unit 210 of this embodiment may represent the strength of preference of financial products corresponding to specified stocks, specified industries, and specified factors by each fund in the same category in the form of a heat map in the similarity map.

[0043] 2 is a diagram showing an example of an affinity map that represents the strength of preferences in the form of a heat map according to this embodiment. FIG. 2 illustrates an affinity map M2 (hereinafter also referred to as a mining map M2) that represents the strength of preferences for holding financial products corresponding to the mining industry in the form of a heat map.

[0044] In the example shown in Figure 2, the strength of preference for holding financial products corresponding to the mining industry is represented by dots and diagonal lines.

[0045] Specifically, when represented by dots on the mining map M2, it indicates a strong preference for holding financial products corresponding to mining, and the higher the density of the dots, the stronger the degree of preference.

[0046] Conversely, when the mining map M2 is represented by diagonal lines, it indicates a low preference for holding financial products corresponding to mining, and the higher the density of the dots, the lower the degree of preference.

[0047] In addition, in the mining map M2 illustrated in FIG. 2, the plots relating to each fund are written according to the same rules as those in the legend shown in FIG.

[0048] Based on these, referring to the mining map M2, it can be intuitively grasped that while Funds A to F do not have a strong preference regarding the holding of financial products corresponding to mining, Fund G has a very strong preference for holding financial products related to mining.

[0049] Thus, according to the analysis unit 210 according to this embodiment, after expressing the similarity in the trading behaviors among funds using distances on a two-dimensional plane, it is possible to express the intensity of preference for the holding of a predetermined financial product in the form of a heat map.

[0050] Also, according to this, it becomes possible to visually and intuitively grasp the differences in the trading behaviors among funds by focusing on the preferences regarding the holding of financial products.

[0051] The above describes the outline of the analysis by the analysis unit 210 according to this embodiment. Note that in the above, the case where the predetermined task is asset management and the actions related to the predetermined task are buying and selling transactions of financial products was described as an example.

[0052] However, the predetermined task and the actions related to the predetermined task according to this embodiment are not limited to such examples.

[0053] For example, the predetermined task according to this embodiment may be the expansion of product sales. Also, actions related to the predetermined task include marketing.

[0054] In this case, the analysis unit 210 according to this embodiment can, for example, perform analysis regarding the similarity and preference of existing channels and newly adopted channels based on the action performance data regarding the channels used for product sales and promotion. The results of such analysis are expected to be utilized for new adoption by marketers, contract cancellation, optimization of marketing costs, etc.

[0055] Also, for example, the predetermined task according to this embodiment may be the acquisition of contracts. Also, actions related to the predetermined task include various business activities.

[0056] In this case, the analysis unit 210 according to this embodiment can analyze the similarities and preferences of various sales activities (e.g., visits, phone calls, emails, presentations, etc.) based on, for example, behavioral performance data related to sales activities. The results of such analysis are expected to be used to evaluate the similarities between existing employees and job candidates, or between transfer candidates and personnel at the transfer destination. In this way, the analysis method according to this embodiment can also be applied to personnel evaluation in companies, etc.

[0057] An example of a system configuration for implementing the above-described analysis will be described in detail below.

[0058] <<1.2. System configuration example>> The system according to this embodiment includes a learning device 10 that performs learning using a machine learning algorithm, and an analysis device 20 that performs analysis using an encoder and a decoder generated by learning by the learning device 10.

[0059] (Learning Device 10) First, a description will be given of an example of the functional configuration of the learning device 10 according to this embodiment. Fig. 3 is a block diagram showing an example of the functional configuration of the learning device 10 according to this embodiment.

[0060] As shown in FIG. 3, the learning device 10 according to this embodiment may include a learning unit 110 and a storage unit 120.

[0061] (Learning Section 110) The learning unit 110 according to this embodiment performs learning using a machine learning algorithm.

[0062] For example, the learning unit 110 according to this embodiment performs learning related to a variational auto-encoder (VAE). A variational auto-encoder is a neural network-based generative model that performs learning using an auto-encoding variational Bayes algorithm.

[0063] Specifically, the learning unit 110 according to this embodiment inputs behavioral performance data into a neural network (encoder) and performs learning to generate a similarity map.

[0064] Furthermore, the learning unit 110 according to this embodiment inputs the latent variables at any point on the similarity map to a neural network (decoder) and performs learning to reproduce the behavioral performance data.

[0065] The functions of the learning unit 110 according to this embodiment are realized by a processor such as a GPU.

[0066] (Storage unit 120) The storage unit 120 according to this embodiment stores various types of information related to the learning performed by the learning unit 110. For example, the storage unit 120 stores the structure of a network used for learning by the learning unit 110, various parameters related to the network, learning data, and the like.

[0067] The above describes an example of the functional configuration of the learning device 10 according to this embodiment. Note that the functional configuration described above using Fig. 3 is merely an example, and the functional configuration of the learning device 10 according to this embodiment is not limited to this example.

[0068] For example, the study device 10 according to this embodiment may further include an operation unit that accepts operations by the user, a display unit that displays various types of information, and the like.

[0069] The functional configuration of the learning device 10 according to this embodiment can be flexibly modified according to specifications and operations.

[0070] (Analyzer 20) Next, a description will be given of an example of the functional configuration of the analysis device 20 according to this embodiment. The analysis device 20 according to this embodiment is an example of an information processing device that analyzes preference trends related to behavior based on behavioral performance data.

[0071] 4 is a block diagram showing an example of the functional configuration of the analysis device 20 according to this embodiment. As shown in FIG. 4, the analysis device 20 according to this embodiment may include an analysis unit 210, a storage unit 220, a display unit 230, and an operation unit 240.

[0072] (Analysis Department 210) The analysis unit 210 according to this embodiment analyzes preference trends related to a behavior based on behavior performance data indicating the performance of the behavior related to a predetermined task.

[0073] In this case, one of the features of the analysis unit 210 according to this embodiment is that it generates a similarity map based on the behavioral performance data, which represents the similarity between the behaviors of each of the multiple task performers, including the subject of analysis, on a two-dimensional plane, and represents the strength of preference for the selected type of behavior in the form of a heat map on the similarity map.

[0074] For example, the analysis unit 210 according to this embodiment may generate a similarity map that represents the similarity between at least the behavior of the subject of analysis and the behavior of a specified comparison subject on a two-dimensional plane, and may represent the strength of preference of the subject of analysis and the strength of preference of the comparison subject for the selected type of behavior in the form of a heat map in the similarity map.

[0075] According to the above-described analysis, the analyst can visually and intuitively grasp the strength of the preference of the subject of analysis and the strength of the preference of the comparison subject for the selected type of behavior.

[0076] Details of the analysis by the analysis unit 210 according to this embodiment will be explained separately. The functions of the analysis unit 210 according to this embodiment are realized by a processor such as a GPU.

[0077] (Storage unit 220) The storage unit 220 according to this embodiment stores various types of information used by the analysis device 20. The storage unit 220 stores, for example, behavioral record data, the structures and parameters of the encoder and decoder used by the analysis unit 210, analysis results, and the like.

[0078] (Display section 230) The display unit 230 according to the present embodiment displays various types of visual information, and therefore includes a display.

[0079] For example, the display unit 230 according to this embodiment displays the results of the analysis by the analysis unit 210 under the control of the analysis unit 210. The results of the analysis include an affinity map and the like.

[0080] (Operation unit 240) The operation unit 240 according to this embodiment accepts operations by the user, and therefore includes various input devices such as a keyboard and a mouse.

[0081] The functional configuration of the analysis device 20 according to this embodiment has been described above. Note that the functional configuration described above using Fig. 4 is merely an example, and the functional configuration of the analysis device 20 according to this embodiment is not limited to this example.

[0082] For example, the analysis unit 210 and storage unit 220 according to this embodiment, and the display unit 230 and operation unit 240 may be provided in separate devices. For example, the analysis unit 210 and storage unit 220 may be provided in an information processing device located on the cloud, and the display unit 230 and operation unit 240 may be provided in a locally located information processing device.

[0083] The functional configuration of the analysis device 20 according to this embodiment can be flexibly modified according to the specifications and operation.

[0084] <<1.3. Analysis Details>> Next, a detailed description will be given of the analysis by the analysis unit 210 according to this embodiment. Fig. 5 is a diagram for explaining the analysis by the analysis unit 210 according to this embodiment.

[0085] First, as shown in the figure, the analysis unit 210 according to this embodiment generates a similarity map M01 by inputting the behavior performance data RD into an encoder 212 generated by learning related to the variational autoencoder by the learning device 10.

[0086] The content of the behavior performance data RD according to this embodiment can be designed according to the task to be analyzed. For example, when the predetermined task is asset management, the behavior performance data RD may include weight information related to financial products for each fund.

[0087] FIG. 6 and FIG. 7 are diagrams for explaining an example of the behavior performance data RD according to this embodiment.

[0088] For example, in the case of the example shown in FIG. 6, the behavior performance data RD1 includes weight information or active weight information related to the stocks held by a certain fund in a certain month.

[0089] When adopting weight information, the behavior performance data RD1 may include the absolute value of the holding amount for each stock.

[0090] On the other hand, when adopting active weight information, the behavior performance data RD1 may include information related to the difference between the holding amount of the stocks held by the fund and the benchmark.

[0091] The behavior performance data RD1 may include the above-mentioned weight information or active weight information for each fund for the period of analysis.

[0092] On the other hand, in the case of the example shown in FIG. 7, the behavior performance data RD2 includes information related to the ratio for each industry related to the stocks held by a certain fund in a certain month.

[0093] In this way, the behavioral performance data RD according to this embodiment does not necessarily have to include information at the granularity of the stock. For example, even when the behavioral performance data RD2 shown in Fig. 7 is used, it is possible to analyze preferences for a specific industry as shown in Fig. 2.

[0094] The information included in the behavior record data RD according to this embodiment may be selected according to the type of behavior by analyzing preferences.

[0095] The information contained in the behavioral performance data RD according to this embodiment includes, in addition to the examples shown in Figures 6 and 7, for example, the number of stocks held, the number of stocks newly added, the total number of stocks sold, the trading turnover rate, and returns.

[0096] The description will continue with reference to Figure 5. After generating the affinity map M01, the analysis unit 210 according to this embodiment inputs the latent variables at any point on the affinity map M01 to the decoder 214 generated by learning related to the variational autoencoder by the learning device 10, and obtains output data OD.

[0097] The output data OD output by the decoder 214 according to this embodiment may be data that reproduces the behavioral performance data input to the encoder 212.

[0098] For example, when the behavioral performance data RD1 shown in FIG. 6 is input to the encoder 212, the output data OD includes weight information for each brand for each coordinate on the affinity map M01.

[0099] Furthermore, when the behavioral performance data RD2 as shown in FIG. 7 is input to the encoder 212, the output data OD includes information regarding the ratio of each industry for each coordinate on the similarity map M01.

[0100] That is, the decoder 214 according to this embodiment can obtain output data OD that is interpolated with unknown data that is not included in the behavioral record data RD.

[0101] In addition, based on the output data OD as described above, the analysis unit 210 according to the present embodiment generates a similarity map M02 that represents the intensity of preference related to the type of behavior in the form of a heatmap.

[0102] According to this, as in the similarity map M2 shown in FIG. 2, it is possible to realize a heatmap representation across the entire map.

[0103] Note that the analysis unit 210 according to the present embodiment may automatically select the type of behavior with a large difference in the intensity of preference between the analysis target person and the comparison target person, and generate a similarity map M02 that represents the intensity of preference related to the selected type of behavior in the form of a heatmap.

[0104] The projector 218 included in the analysis unit 210 according to the present embodiment is configured to automatically select the type of behavior with a large difference in the intensity of preference between the analysis target person and the comparison target person, and generate a heatmap representation related to the selected type of behavior.

[0105] The projector 218 according to the present embodiment may, for example, select the type of behavior with a difference in the intensity of preference between the analysis target person and the comparison target person based on the contribution degree of each element included in the behavior performance data RD for generating the similarity map M01.

[0106] ]>The above contribution degree is calculated, for example, by SHAP (SHapley Additive exPlanations) 226 included in the analysis unit 210.

[0107] SHAP is a model that calculates, in a learned model, a value indicating how each element in the input data affects the predicted value of the model, that is, the contribution degree of each data to the predicted value.

[0108] For example, the SHAP 216 according to the present embodiment calculates the contribution degree (SHAP value) of each element included in the behavior performance data RD for generating the similarity map M01 by the encoder 212.

[0109] Figures 8 and 9 are diagrams showing an example of visually expressing the contribution degree calculated by SHAP216 according to the present embodiment.

[0110] For example, in an example shown in FIG. 8, in the action performance data related to Fund A in March 2020, each element included in the action performance data (each stock in an example shown in FIG. 8) is shown in a ranking order according to the contribution degree to the X-axis or Y-axis of the similarity map.

[0111] For example, in an example shown in FIG. 8, it can be understood that the contribution degree of the stock corresponding to Company A is the highest both in the X-axis and the Y-axis.

[0112] On the other hand, in an example shown in FIG. 9, in the action performance data related to Fund G in March 2020, each element included in the action performance data (each stock in an example shown in FIG. 9) is shown in a ranking order according to the contribution degree to the X-axis or Y-axis of the similarity map.

[0113] For example, in an example shown in FIG. 9, it can be understood that for the contribution degree to the X-axis, the stock corresponding to Company J is the highest, and for the contribution degree to the Y-axis, the stock corresponding to Company I is the highest.

[0114] The projector 218 according to the present embodiment may automatically select the type of action to be expressed in the form of a heat map based on the contribution degree calculated as described above.

[0115] FIG. 10 is a diagram for explaining the automatic selection of the type of action based on the contribution degree according to the present embodiment. For example, the projector 218 according to the present embodiment may calculate the difference in the contribution degree (SHAP value) for each stock between the analysis target person and the comparison target person as shown in FIG. 10.

[0116] The projector 218 may, for example, calculate the difference in the contribution degree for each stock between the analysis target person and the comparison target person respectively on the X-axis and the Y-axis, and further calculate the average of the difference on the X-axis and the difference on the Y-axis.

[0117] In this case, the projector 218 can determine that the brand with a higher average of the difference on the X-axis and the difference on the Y-axis is a brand with a larger difference in the intensity of preference.

[0118] From this, the projector 218 may select, from the top, brands with a high average of the difference on the X-axis and the difference on the Y-axis, and generate a similarity map M02 that represents, in a heatmap-like manner, the intensity of preference related to the brand, the industry related to the brand, the factors related to the brand, and the like.

[0119] Alternatively, the projector 218 may perform clustering of brands based on the difference in contribution as shown in FIG. 10, identify a group with a large weighted average value of the difference average, and generate a similarity map M02 that represents, in a heatmap-like manner, the intensity of preference related to the group.

[0120] In addition, the projector 218 according to the present embodiment may accumulate user feedback on the similarity map represented in a heatmap-like manner, and automatically select the type of action to be heatmapped based on the feedback.

[0121] FIG. 11 is a diagram for explaining the automatic selection of the type of action based on user feedback according to the present embodiment.

[0122] For example, the user checks a similarity map in which the intensity of preference for each brand is represented in a heatmap-like manner, and as shown in the upper part of FIG. 11, provides feedback on an evaluation value that represents, on a five-point scale, whether the preference for the corresponding brand is characteristic of the corresponding fund.

[0123] Thereafter, as shown in the lower part of FIG. 11, the projector 218 performs matrix decomposition based on the value given as user feedback, and predicts the evaluation value of the brand for which no feedback has been given. In FIG. 11, the evaluation value predicted by the projector 218 is emphasized and shown by dots.

[0124] Next, based on the evaluation values feedback by the user and the predicted evaluation values, the projector 218 automatically selects stocks, and presents to the user a similarity map that expresses the strength of preference for the selected stocks in the form of a heat map.

[0125] By repeatedly executing the presentation of the similarity map as described above, the feedback of the evaluation values by the user, the prediction of the evaluation values based on the feedback, etc., the projector 218 can improve the selection accuracy of stocks having a preference tendency characteristic of the fund.

[0126] <<1.4. Example of User Interface>> The details of the analysis by the analysis unit 210 according to the present embodiment have been described above. Next, an example of a user interface for presenting the results of the analysis by the analysis unit 210 according to the present embodiment will be described.

[0127] FIGS. 12 to 16 are diagrams showing an example of a user interface UI1 for presenting the results of the analysis by the analysis unit 210 according to the present embodiment. The analysis unit 210 according to the present embodiment may control the user interface UI1 by the display unit 230.

[0128] In each similarity map illustrated in FIGS. 12 to 16, the plots related to each fund are described according to the same rules as the legend shown in FIG. 1.

[0129] FIG. 12 shows an example of the "Similarity Map" tab of the user interface UI1.

[0130] For example, a similarity map M1 generated by the analysis unit 210 may be displayed on the "Similarity Map" tab.

[0131] Also, the user may select a fund for which information is to be displayed on the similarity map M1 using, for example, a field as shown in the lower left of FIG. 12.

[0132] Furthermore, the user may be able to use fields such as those shown in the lower left of Figure 12 to arbitrarily select a fund to be analyzed and a fund to be compared with the fund to be analyzed.

[0133] The above-described setting of the analysis subject and comparison subject is particularly effective when it is desired to compare the similarity of behavior between certain specific funds.

[0134] For example, suppose a user is searching for a candidate fund to enter into a new contract with when the current contract with the current fund expires. In this case, the user may set the current fund (e.g., Fund A) as the comparison target and the candidate fund (e.g., Fund G) as the analysis target.

[0135] The above settings make it possible to perform a detailed analysis that focuses on the similarity between the behavior of the fund under contract and the behavior of the candidate fund.

[0136] In this case, the analysis unit 210 may also display, on the user interface UI1, information indicating the distance of each fund on the similarity map M1 based on the fund set as the comparison target.

[0137] In the example shown in Figure 12, the analysis unit 210 displays the distance of each fund on the similarity map M1 based on fund A, which is set as the comparison target, and highlights the distance related to fund G, which is set as the analysis target.

[0138] According to the display control described above, it becomes possible to relatively grasp how far apart the analysis target and the comparison target are in the similarity map M1 by comparing with other funds.

[0139] In addition, FIGS. 13 to 15 show an example of the "Difference" tab of the user interface UI1. In this tab, the analysis unit 210 according to the present embodiment may present a plurality of users with types of actions having a large difference in the intensity of preference between the analysis target person and the comparison target person.

[0140] Also, in this case, the analysis unit 210 according to the present embodiment may display a similarity map in which the intensity of preference related to the type of action selected by the user is represented in the form of a heat map.

[0141] For example, in the case of the example shown in FIGS. 13 to 15, the analysis unit 210 presents "Industry: Mining", "Industry: Transportation Equipment", and "Underpriced" as types of actions having a large difference in the intensity of preference between the fund G set as the analysis target person and the fund A set as the comparison target person.

[0142] The analysis unit 210 according to the present embodiment can automatically select the types of actions as described above based on the contribution degree calculated using the above-described SHAP216 and the like.

[0143] Here, for example, as shown in FIG. 13, when the user selects "Industry: Transportation Equipment", the analysis unit 210 may display on the user interface UI1 a similarity map M3 (hereinafter also referred to as the transportation equipment map M3) in which the intensity of preference regarding the holding of financial products corresponding to "Industry: Transportation Equipment" is represented in the form of a heat map.

[0144] According to the transportation equipment map M3 shown in FIG. 13, while the fund A set as the comparison target person has a strong preference regarding the holding of financial products corresponding to transportation equipment, the user can visually and intuitively grasp that the fund G set as the analysis target person has a weak degree of preference for holding financial products for transportation equipment.

[0145] Also, according to this, the user can infer that if the fund A is replaced with the fund G, the holding amount of financial products corresponding to transportation equipment will be greatly reduced.

[0146] On the one hand, when the user selects "Industry: Mining", the analysis unit 210 may display a mining map M2 as shown in FIG. 2 on the user interface UI1.

[0147] Also, in this case, the analysis unit 210 may present individual brands that have particular differences in preference in "Industry: Mining" like "Enterprise L" shown in FIG. 14.

[0148] Here, when the user selects "Enterprise L", the analysis unit 210 may display on the user interface UI1 a similarity map M4 (hereinafter, also referred to as the enterprise L map M4) that represents the intensity of preference regarding the holding of financial products corresponding to "Enterprise L" in the form of a heat map.

[0149] According to the mining map M2 shown in FIG. 2 and the enterprise L map M4 shown in FIG. 14, it is possible for the user to visually and intuitively grasp that the fund A set as the comparison target does not have a strong preference for holding financial products corresponding to mining, while the fund G set as the analysis target has a strong preference for holding financial products for mining.

[0150] Also, according to this, the user can infer that if the fund A is replaced with the fund G, the holding amount of financial products corresponding to mining will increase significantly, and can consider whether they can tolerate the risk.

[0151] On the other hand, as shown in FIG. 15, when the user selects "Inexpensive", the analysis unit 210 may display on the user interface UI1 a similarity map M5 (hereinafter, also referred to as the inexpensive map M5) that represents the intensity of preference regarding the holding of financial products corresponding to the factor "Inexpensive" in the form of a heat map.

[0152] In this way, the analysis unit 210 according to the present embodiment can also represent the intensity of preference regarding the holding of financial products corresponding to a predetermined factor in the form of a heat map on a similarity map.

[0153] According to the discount map M5 shown in FIG. 15, while the fund A set as the comparison target has a strong preference for holding financial products corresponding to the factor "discount", the user can visually and intuitively grasp that the fund G set as the analysis target has a weak degree of preference for holding financial products for the factor "discount".

[0154] Also according to this, the user can infer that if the fund A is replaced with the fund G, the holding amount of the financial product corresponding to the factor "discount" will be greatly reduced.

[0155] As described above, an example of information presentation in the "Difference" tab according to this embodiment has been shown. According to the information presentation as exemplified in FIGS. 13 to 15, by simply selecting the items automatically listed by the analysis unit 210, the user can visually and intuitively grasp the information related to the types of actions with a large difference in preference between the analysis target and the comparison target.

[0156] On the other hand, in addition to the types of actions listed by the analysis unit 210, the user may be able to display on the user interface a similarity map that represents the degree of preference for the types of actions selected by the user in the form of a heat map.

[0157] FIG. 16 shows an example of the "Details" tab of the user interface UI1. The analysis unit 210 according to this embodiment may, for example, present a graph that visually represents the contribution degree for each brand as shown in FIGS. 8 and 9 in this tab.

[0158] In this case, the analysis unit 210 according to this embodiment may represent the intensity of preference for the type of action corresponding to the element selected by the user in the similarity map in the form of a heat map based on the presented contribution degree.

[0159] For example, in the case of an example shown in FIG. 16, based on the user selecting "Company L" in a graph visually representing the contribution degree for each stock, the analysis unit 210 causes the Company L map M4 to be displayed on the user interface UI1.

[0160] In addition to the Company L map M4, the analysis unit 210 may further display a similarity map visually representing the intensity of preference for the industry type or factors corresponding to "Company L" in the form of a heat map.

[0161] According to the above control, it becomes possible for the user to visually and intuitively grasp the differences in trading behaviors between funds by focusing on the preferences related to the holding of any financial product.

[0162] <<1.5. Flow of Processing>> Next, an example of the processing flow by the analysis apparatus 20 according to the present embodiment will be described in detail. FIG. 17 is a flowchart showing an example of the processing flow by the analysis apparatus 20 according to the present embodiment.

[0163] In the case of an example shown in FIG. 17, first, the analysis unit 210 inputs the behavior performance data into the encoder 212 and generates a similarity map (S102).

[0164] [[ID=2,2]]Next, the analysis unit 210 inputs the latent variable at an arbitrary point on the similarity map generated in step S102 into the decoder 214 and acquires output data (S104).

[0165] Also, the analysis unit 210 uses SHAP216 to calculate the contribution degree of each element included in the behavior performance data to the generation of the similarity map in step S102 (S106).

[0166] Next, the analysis unit 210 selects the type of behavior with a large difference in the intensity of preference between the analysis target person and the comparison target person based on the contribution degree and the like calculated in step S106 (S108).

[0167] Next, the analysis unit 210 presents the type of behavior selected in step S108 to the user via the user interface UI1 (S110).

[0168] Next, the analysis unit 210 displays, on the user interface UI1, an affinity map that represents, in the form of a heat map, the strength of preferences related to the types of behavior selected by the user on the user interface UI1 (S112).

[0169] The processing flow by the analysis device 20 according to this embodiment has been described in detail above using an example. Note that the processing flow described above using Fig. 17 is merely an example, and the processing flow by the analysis device 20 according to this embodiment is not limited to this example.

[0170] The processing flow by the analysis device 20 according to this embodiment can be flexibly modified according to the specifications and operation.

[0171] 2. Second embodiment <<2.1. Generation of behavioral reproduction data>> Next, a second embodiment of the present disclosure will be described. Note that the following description will focus on the differences from the first embodiment, and detailed description of configurations and effects common to the first embodiment will be omitted.

[0172] In the first embodiment, the analysis device 20 mainly analyzes the difference (similarity) between the behavior of the analysis subject and the behavior of other task performers using a similarity map.

[0173] However, the use of the affinity map is not limited to the above-described analysis. For example, analysis device 20 can use the affinity map to reproduce tasks that have already been performed or tasks that can be performed by a third party.

[0174] In what follows, similar to the first embodiment, when the predetermined task is asset management and the actions related to the predetermined task are trading transactions of financial products, this will be mainly described as an example. Also, assume that the task executor is a fund.

[0175] For example, in asset management, risk diversification is a major issue. Since the market value of each asset fluctuates according to the market environment, it can be said that an asset with large price movements and expected returns has a high possibility of incurring losses on the other hand. For this reason, it is important to diversify risks by combining assets with different movements depending on the market situation.

[0176] As an example, when entering into a contract with a new fund in a fund or the like, risk diversification can be achieved by adopting a fund with different investment assets and method characteristics from the already contracted funds.

[0177] However, with conventional tools, when an analyst does not have specialized knowledge, it has been difficult to intuitively grasp the differences in asset management between funds, and thus it has also been difficult to find a new fund that conducts asset management different from the already contracted funds.

[0178] On the other hand, according to the similarity map generated by the analysis device 20 according to an embodiment of the present disclosure, even when an analyst does not have specialized knowledge, it becomes possible to intuitively grasp the differences in asset management between funds.

[0179] Furthermore, the analysis device 20 according to the second embodiment of the present disclosure can output action reproduction data that reproduces tasks that can be executed by a third party having characteristics different from those of the already contracted funds by using the generated similarity map.

[0180] FIGS. 18 and 19 are diagrams for explaining the generation of action reproduction data according to the second embodiment of the present disclosure.

[0181] 18 shows the similarity map M02 generated by the analysis unit 210. As in the first embodiment, each plot in the similarity map M02 represents the asset management characteristics of a given fund over a given period. That is, the closer the distance between plots, the more similar the asset management characteristics are, and the farther the distance between plots, the more different the asset management characteristics are.

[0182] Here, attention is focused on areas with low density of plots, particularly blank areas BS1 to BS4 where no plots exist, in the similarity map M02 shown in FIG.

[0183] Since there are no plots based on asset management by Funds A to G in blank areas BS1 to BS4, it can be assumed that these areas are areas that are not fully covered by asset management by existing funds and that they likely contain portfolios using new management methods.

[0184] That is, if asset management corresponding to the plots located in the blank areas BS1 to BS4 can be reproduced, it becomes possible to obtain data on asset management having characteristics different from those of existing funds.

[0185] The technical concept of the second embodiment of the present disclosure was conceived with the above points in mind, and makes it possible to reproduce with high accuracy the tasks corresponding to the coordinates specified in the similarity map.

[0186] For this reason, one of the features of the analysis unit 210 of the analysis device 20 according to this embodiment is that it outputs behavior reproduction data having the same data structure as the behavior performance data based on coordinate data representing the position of the plot in the similarity map.

[0187] Furthermore, as described above, when representing the time-series changes in the behavior of each of multiple task performers in the similarity map, the analysis unit 210 may output behavior reproduction data that represents the time-series changes based on the coordinate data.

[0188] For example, in the example shown in Figure 19, in the blank area BS1 of the similarity map M02, plots (hexagons) that assume the characteristics of asset management by a fictitious fund Z are drawn in the same time series as the behavioral performance data used to generate the similarity map M02.

[0189] Each plot related to Fund Z is sufficiently far away from each plot based on the performance data of existing Funds A to G that was used to generate Similarity Map M02, and it can be said that it represents asset management with different characteristics from existing Funds A to G.

[0190] The analysis unit 210 according to this embodiment uses coordinate data (x z ,y z ) to the trained decoder 214 shown in FIG. 5, it is possible to output behavior reproduction data having the same data structure as the behavior record data.

[0191] For example, assume that the behavioral performance data used to generate the similarity map M02 includes weight information or active weight information regarding stocks 1 to n held by a certain fund during a certain period, as shown in Fig. 6. In this case, the behavioral reproduction data may also include weight information or active weight information regarding stocks 1 to n for each period.

[0192] Also, for example, assume that the behavioral performance data used to generate the similarity map M02 includes information on the ratios of stocks held by a certain fund for each of industry types 1 to n during a certain period, as shown in Fig. 7. In this case, the behavioral reproduction data may also include weight information or active weight information for each of industry types 1 to n during each period.

[0193] Note that the coordinate data used for generating the action reproduction data may be provided by an analyst using an external file or the like. Alternatively, the analysis unit 210 may convert the points and lines drawn by the analyst on the similarity map M02 into coordinate data and use them for generating the action reproduction data.

[0194] Furthermore, the analysis unit 210 may extract an area where the plot density is low (for example, a blank area) on the generated similarity map M02 and automatically generate coordinate data representing the positions of the plots within the area.

[0195] In this case, the analysis unit 210 may generate coordinate data so that plots are generated at positions as far as possible from existing plots. Also, when generating time-series action reproduction data, the coordinate data related to each plot may be generated so that a smooth line without abrupt changes is formed.

[0196] Thus, according to the analysis unit 210 according to the present embodiment, it is possible to simply and accurately reproduce data related to asset management that can be executed by a new fund having characteristics different from those of existing funds.

[0197] Also, by subjecting the obtained Action reproduction data to various analyses, it is also possible to clarify characteristics such as the weight distribution, factor exposure, and sector exposure, and to evaluate investment realizability such as liquidity and investment costs.

[0198] Note that the action performance data and the results of various analyses based on the action performance data can be widely used for searching for new funds, developing R&D management using in-house funds, and the like.

[0199] Here, the flow of the process related to the generation of the action reproduction data will be described in more detail. FIG. 20 is a flowchart showing an example of the flow of action reproduction data generation according to the present embodiment.

[0200] In the example shown in FIG. 20, the analysis unit 210 first inputs the behavioral performance data to the trained encoder 212 and generates a similarity map (S202).

[0201] The generation of the affinity map in step S202 may be the same as the processing in the first embodiment, and therefore a detailed description thereof will be omitted.

[0202] Next, the analysis unit 210 generates coordinate data (S204).

[0203] For example, the analysis unit 210 may generate coordinate data based on points and lines drawn on the affinity map, or may automatically generate coordinate data based on extracted blank areas.

[0204] On the other hand, if the coordinate data is provided using an external file or the like, the process in step S204 may be skipped.

[0205] Next, the analysis unit 210 inputs the coordinate data to the trained decoder 214 and outputs behavior reproduction data (S206).

[0206] As described above, the behavior reproduction data output in step S206 may have the same data structure as the behavior result data used to generate the similarity map in step S202.

[0207] The flow of generating behavior reproduction data according to this embodiment has been described above with an example.

[0208] Until now, most of the search for new investment methods has mainly consisted of examining portfolios similar to existing investment methods or analyzing portfolios brought in from outside.

[0209] For this reason, the analysis of those responsible for developing new investment methods tends to be biased towards waiting for information from outside, and since the portfolio obtained as information from outside is not necessarily the portfolio that the company is looking for, the development of new investment methods tends to be highly personal and dependent on accidental factors.

[0210] On the other hand, the above-described analysis method according to this embodiment enables the person in charge to proactively search for a portfolio and investment management company with desirable characteristics that suit their own situation. As a result, by using the analysis method according to this embodiment, it is expected that the dependence on individuals and chance factors in the development process of new investment methods will be reduced, and that the productivity and efficiency of the development process will be improved.

[0211] <<2.2. Representation of behavior reproduction data using existing behavioral performance data>> Next, a description will be given of how to express behavior reproduction data using existing behavior record data according to this embodiment.

[0212] The above describes a case where the analysis unit 210 generates behavior reproduction data by inputting coordinate data into the trained decoder 214. This makes it possible to easily obtain data that reproduces asset management that may be carried out by a fictitious new fund with characteristics different from existing funds, for example, and to utilize the data for adopting a new fund, etc.

[0213] On the other hand, in many cases, each investment entity (asset management company, etc.) sets its own investment limit for the investment portfolio, taking into account factors such as the liquidity of investment stocks and investment efficiency. Since asset owners such as funds cannot make additional investments beyond this investment limit, an upper limit is inevitably imposed on the investment.

[0214] On the other hand, if it is possible to replicate a specific fund with a small investment limit using multiple funds with a large investment limit, it will be possible to alleviate the constraints of the investment limit.

[0215] In view of the above, the analysis unit 210 according to this embodiment may express behavior reproduction data corresponding to arbitrary coordinate data using behavioral performance data of a plurality of task performers.

[0216] As an example, the analysis unit 210 according to this embodiment may represent behavior reproduction data corresponding to the behavior performance data of a selected existing task performer using behavior performance data of multiple other task performers different from the specified task performer.

[0217] Furthermore, the analysis unit 210 according to this embodiment may output information regarding the composite ratio of the behavioral performance data of a plurality of task performers used to express the behavior reproduction data.

[0218] This makes it possible to replicate asset management by any existing fund by combining asset management by multiple other existing funds, thereby making it possible to significantly relax restrictions on investment limits.

[0219] FIG. 21 is a diagram for explaining the representation of behavior reproduction data using existing behavior result data according to this embodiment.

[0220] FIG. 21 shows an example of a case where behavioral reproduction data that reproduces asset management by an existing fund C is expressed using behavioral performance data of existing funds B, E, and G.

[0221] When attempting to generate behavioral reproduction data that reproduces asset management by a given existing fund, behavioral performance data of three other existing funds is used.

[0222] In this case, the behavioral performance data of other existing funds that can be used can, in principle, be any option, but it is predicted that the accuracy of reproduction will be higher the longer the trajectory formed by the plot based on the behavioral performance data, the greater the distance between the centers of gravity of the trajectory, and the closer the angle of the line segment connecting the centers of gravity is to a right angle.

[0223] The analysis unit according to this embodiment may calculate a synthesis ratio based on the positional relationship between plots based on the action performance data of each of a plurality of task performers in the similarity map.

[0224] More specifically, the analysis unit 210 may generate a linear combination vector representing specified coordinate data using two position vectors obtained from three plots based on the action performance data of three different task performers, and calculate a synthesis ratio based on the linear combination vector.

[0225] For example, in the case of an example shown in FIG. 21, the analysis unit 210 calculates a synthesis ratio based on the positional relationship between plots based on the action performance data of Fund B, Fund E, and Fund G in the similarity map M02.

[0226] Here, the coordinates of each plot are defined as in the following mathematical formula (1). In this case, the position vector 1 (starting from the plot of Fund G) formed by the plot of Fund G and the plot of Fund B in a certain period can be expressed as in the following mathematical formula (2). Similarly, the position vector 2 (starting from the plot of Fund G) formed by the plot of Fund G and the plot of Fund E in a certain period can be expressed as in the following mathematical formula (3).

[0227]

Equation

[0228] The analysis unit 210 according to this embodiment generates a linear combination vector corresponding to the plot P of Fund C to be reproduced using the above position vector 1 and position vector 2. At this time, the linear combination vector can be calculated as in the following mathematical formula (4).

[0229]

Equation

[0230] Next, as shown in the following mathematical formula (5), the analysis unit 210 according to the present embodiment calculates the synthesis ratio of the behavioral performance data corresponding to each plot of Fund B, Fund E, and Fund G in reverse using the coefficients (α1, α2) of the linear combination. At this time, each plot of Fund B, Fund E, and Fund G can be represented by the following mathematical formulas (6) to (8).

[0231]

Number

[0232] As described above, according to the analysis unit 210 according to the present embodiment, it is possible to represent the behavior reproduction data corresponding to arbitrary coordinate data using the behavioral performance data of a plurality of task performers.

[0233] In the above, the case where the asset management by a certain existing fund is reproduced by the behavioral performance data of a plurality of other existing funds has been exemplified. However, the analysis unit 210 according to the present embodiment can represent a task corresponding to an arbitrary coordinate by the existing behavioral performance data regardless of whether it is existing or not.

[0234] For example, as described above, the analysis unit 210 may represent the plots automatically generated in the blank area using the existing plots. According to this, it is also possible to realize asset management having characteristics different from those of existing funds by synthesizing existing funds.

[0235] Next, with reference to FIG. 22, the flow of output of the behavior reproduction data using the existing behavioral performance data according to the present embodiment will be described in detail.

[0236] In the case of an example shown in FIG. 22, the analysis unit 210 first inputs the behavioral performance data into the learned encoder 212 to generate a similarity map (S302).

[0237] Since the generation of the similarity map in step S302 may be the same as the processing in the first embodiment, a detailed description thereof will be omitted.

[0238] Next, the coordinate data of the plot corresponding to the behavior reproduction data to be reproduced is designated (S304).

[0239] For example, when generating behavior reproduction data related to an existing fund, any existing fund may be specified using the user interface. Note that the coordinate data specified in step S304 is not limited to the plots related to the existing fund as described above.

[0240] Next, the analysis unit 210 calculates a synthesis ratio relating to the behavioral performance data of the multiple task performers based on the coordinate data designated in step S304 (S306).

[0241] Next, the analysis unit 210 outputs information relating to the synthesis ratio calculated in step S306 and the behavior reproduction data generated based on the synthesis ratio (S308).

[0242] <3. Hardware configuration example> Next, an example of a hardware configuration common to the learning device 10 and the analysis device 20 according to an embodiment of the present disclosure will be described. Fig. 23 is a block diagram showing an example of the hardware configuration of an information processing device 90 according to an embodiment of the present disclosure. The information processing device 90 may be a device having a hardware configuration equivalent to that of each of the above devices.

[0243] 23, the information processing device 90 includes, for example, a processor 871, a ROM 872, a RAM 873, a host bus 874, a bridge 875, an external bus 876, an interface 877, an input device 878, an output device 879, a storage 880, a drive 881, a connection port 882, and a communication device 883. Note that the hardware configuration shown here is an example, and some of the components may be omitted. Furthermore, the information processing device 90 may further include components other than those shown here.

[0244] (Processor 871) The processor 871 functions as, for example, an arithmetic processing unit or a control unit, and controls the overall operation or a part of the operation of each component based on various programs recorded in the ROM 872, the RAM 873, the storage 880, or the removable storage medium 901.

[0245] (ROM872, RAM873) The ROM 872 is a means for storing programs read by the processor 871, data used for calculations, and the like. In the RAM 873, for example, programs read by the processor 871 and various parameters that change as appropriate when executing the programs are stored temporarily or permanently.

[0246] (Host bus 874, bridge 875, external bus 876, interface 877) The processor 871, the ROM 872, and the RAM 873 are interconnected via, for example, a host bus 874 capable of high-speed data transmission. On the other hand, the host bus 874 is connected to an external bus 876 with a relatively low data transmission speed via, for example, a bridge 875. Further, the external bus 876 is connected to various components via an interface 877.

[0247] (Input device 878) For the input device 878, for example, a mouse, a keyboard, a touch panel, buttons, switches, levers, and the like are used. Further, as the input device 878, a remote controller (hereinafter, remote control) capable of transmitting a control signal using infrared rays or other radio waves may be used. In addition, the input device 878 includes a voice input device such as a microphone.

[0248] (Output device 879) The output device 879 is a device capable of visually or auditorily notifying the user of the acquired information, such as a display device such as a CRT (Cathode Ray Tube), LCD, or organic EL, an audio output device such as a speaker or headphones, a printer, a mobile phone, or a facsimile. Further, the output device 879 according to the present disclosure includes various vibration devices capable of outputting tactile stimuli.

[0249] (Storage 880) The storage 880 is a device for storing various data. As the storage 880, for example, a magnetic storage device such as a hard disk drive (HDD), a semiconductor storage device, an optical storage device, or a magneto-optical storage device is used.

[0250] (Drive 881) The drive 881 is a device that reads information recorded on a removable storage medium 901 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, or writes information to the removable storage medium 901.

[0251] (Removable storage medium 901) The removable storage medium 901 is, for example, a DVD medium, a Blu-ray (registered trademark) medium, an HD DVD medium, various semiconductor storage media, etc. Of course, the removable storage medium 901 may be, for example, an IC card equipped with a non-contact type IC chip, or an electronic device.

[0252] (Connection port 882) The connection port 882 is a port for connecting an external connection device 902 such as a USB (Universal Serial Bus) port, an IEEE1394 port, a SCSI (Small Computer System Interface), an RS-232C port, or an optical audio terminal.

[0253] (External connection device 902) The external connected device 902 is, for example, a printer, a portable music player, a digital camera, a digital video camera, or an IC recorder, etc.

[0254] (Communication device 883) The communication device 883 is a communication device for connecting to a network, and is, for example, a communication card for wired or wireless LAN, Bluetooth (registered trademark), or WUSB (Wireless USB), a router for optical communication, a router for ADSL (Asymmetric Digital Subscriber Line), or a modem for various communications, etc.

[0255] <4. Summary> As described above, the analysis device 20 according to the second embodiment of the present disclosure includes an analysis unit 210 that generates a similarity map representing the similarity between actions by each of a plurality of task performers on a two-dimensional plane based on action performance data indicating the performance of actions related to a predetermined task.

[0256] Also, the analysis unit 210 according to the second embodiment of the present disclosure is characterized in that it outputs action reproduction data having the same data structure as the action performance data based on the coordinate data representing the position of the plot in the similarity map.

[0257] According to the above configuration, it is possible to accurately reproduce the executed tasks and the tasks that can be executed by a third party.

[0258] As described above, the preferred embodiments of the present disclosure have been described in detail with reference to the accompanying drawings, but the technical scope of the present disclosure is not limited to such examples. It is obvious that those having ordinary knowledge in the technical field of the present disclosure can conceive of various modification examples or correction examples within the scope of the technical idea described in the claims, and these are also naturally understood to belong to the technical scope of the present disclosure.

[0259] Furthermore, the steps of the processes described in this specification do not necessarily have to be processed in chronological order according to the order shown in the flowcharts or sequence diagrams. For example, the steps of the processes of each device may be processed in an order different from the order shown, or may be processed in parallel.

[0260] Furthermore, the series of processes performed by each device described in this specification may be realized using software, hardware, or a combination of software and hardware. The programs constituting the software may be provided, for example, inside or outside each device and stored in advance in a non-transitory computer-readable medium. Each program is then loaded into RAM when executed by a computer and executed by various processors. Examples of the storage medium include a magnetic disk, an optical disk, a magneto-optical disk, and a flash memory. Furthermore, the computer programs may be distributed, for example, via a network, without using a storage medium.

[0261] Furthermore, the effects described herein are merely descriptive or exemplary and are not limiting. In other words, the technology according to the present disclosure may achieve other effects that are apparent to those skilled in the art from the description of this specification, in addition to or in place of the above-described effects.

[0262] The following configurations also fall within the technical scope of the present disclosure. (1) an analysis unit that generates a similarity map that represents similarities between actions by each of a plurality of task performers on a two-dimensional plane based on action performance data that indicates the performance of the actions related to a predetermined task; Equipped with the analysis unit outputs behavior reproduction data having the same data structure as the behavior performance data based on coordinate data representing the position of a plot in the similarity map. Information processing device. (2) The analysis unit represents changes in the time series of the actions by each of the plurality of task performers in the similarity map, and outputs the action reproduction data representing the changes in the time series based on the plurality of the coordinate data. The information processing apparatus according to (1) above. (3) The analysis unit extracts a region with a low plot density on the generated similarity map, and outputs the action reproduction data based on the coordinate data representing the positions of the plots within the region. The information processing apparatus according to (1) or (2) above. (4) The analysis unit represents the action reproduction data corresponding to the coordinate data using the action performance data of the plurality of task performers. The information processing apparatus according to (1) or (2) above. (5) The analysis unit represents the action reproduction data corresponding to the action performance data of a selected predetermined task performer using the action performance data of a plurality of other task performers different from the predetermined task performer. The information processing apparatus according to (4) above. (6) The analysis unit outputs information regarding the synthesis ratio of the action performance data of the plurality of task performers used to represent the action reproduction data. The information processing apparatus according to (4) or (5) above. (7) The analysis unit calculates the synthesis ratio based on the positional relationship between the plots based on the action performance data of each of the plurality of task performers in the similarity map. The information processing apparatus according to (6) above. (8) The analysis unit generates a linear combination vector representing the specified coordinate data using two position vectors obtained from three plots based on the action performance data of three different task performers, and calculates the synthesis ratio based on the linear combination vector. The information processing apparatus according to (7) above. (9) The analysis unit generates the similarity map by inputting the action performance data into a learned encoder. The information processing apparatus according to (1) above. (10) The analysis unit inputs the coordinate data into a learned decoder and outputs the action reproduction data. The information processing apparatus according to (9) above. (11) The encoder and the decoder are generated by learning related to a variational autoencoder. The information processing apparatus according to (10) above. (12) The predetermined task includes asset management. The information processing apparatus according to (1) above. (13) The action includes buying and selling transactions of financial products. The information processing apparatus according to (12) above. (14) The type of the action includes holding the financial product of a predetermined brand. The information processing apparatus according to (13) above. (15) The type of the action includes holding the financial product corresponding to a predetermined industry. The information processing apparatus according to (13) above. (16) The type of the action includes holding the financial product corresponding to a predetermined factor. The information processing apparatus according to (13) above. (17) The action performance data and the action reproduction data include weight information related to financial products. The information processing apparatus according to (13) above. (18) A display unit that displays the similarity map. The information processing apparatus according to (1) above further includes. The information processing apparatus according to (1) above. (19) The processor generates a similarity map that represents the similarity between the actions by each of a plurality of task performers on a two-dimensional plane based on action performance data indicating the performance of actions related to a predetermined task; Output action reproduction data having the same data structure as the action performance data based on coordinate data representing the position of a plot in the similarity map; including An information processing method. (20) A computer An analysis unit that generates a similarity map that represents the similarity between the actions by each of a plurality of task performers on a two-dimensional plane based on action performance data indicating the performance of actions related to a predetermined task; equipped with Based on the coordinate data representing the position of the plot in the similarity map, the analysis unit outputs action reproduction data having the same data structure as the action performance data. An information processing apparatus A program for causing it to function.

Explanation of Signs

[0263] 10 Learning apparatus 110 Learning unit 120 Storage unit 20 Analysis apparatus 210 Analysis unit 212 Encoder 214 Decoder 216 SHAP 218 Projector 220 Storage unit 230 Display unit 240 Operation unit

Claims

1. An analysis unit that analyzes a preference tendency related to the buying and selling transactions of the financial product based on the behavioral performance data indicating the performance of the buying and selling transactions of the financial product. Comprising The analysis unit generates a similarity map that represents, on a two-dimensional plane, the similarity between the buying and selling transactions of the financial product by at least the analysis target person and the buying and selling transactions of the financial product by a designated comparison target person based on the behavioral performance data, and represents, in the form of a heat map in the similarity map, the strength of the preference of the analysis target person regarding the holding of a selected predetermined financial product and the strength of the preference of the comparison target person. An information processing apparatus.

2. The analysis unit represents, in the similarity map, the change in the time series of the buying and selling transactions of the financial product. The information processing apparatus according to claim 1.

3. The analysis unit selects the predetermined financial product with a large difference in the strength of preference between the analysis target person and the comparison target person, and represents, in the form of a heat map in the similarity map, the strength of the preference regarding the holding of the selected predetermined financial product. The information processing apparatus according to claim 1.

4. The analysis unit presents a plurality of the predetermined financial products with a large difference in the strength of preference between the analysis target person and the comparison target person to the user, and represents, in the form of a heat map in the similarity map, the strength of the preference regarding the holding of the predetermined financial product selected by the user. The information processing apparatus according to claim 1.

5. The analysis unit selects the predetermined financial product with a large difference in the strength of preference between the analysis target person and the comparison target person based on the contribution degree of each element included in the behavioral performance data to the generation of the similarity map. The information processing apparatus according to claim 3.

6. The analysis unit presents the contribution degree to the user. The information processing apparatus according to claim 5.

7. The analysis unit represents, in the form of a heat map in the similarity map, the strength of the preference regarding the holding of the predetermined financial product corresponding to the element included in the behavioral performance data selected by the user based on the presented contribution degree. The information processing apparatus according to claim 6.

8. The analysis unit generates the similarity map by inputting the behavioral performance data into a trained encoder. The information processing apparatus according to claim 1.

9. The analysis unit expresses, in a heatmap form on the similarity map, the strength of preference regarding the holding of the predetermined financial product based on output data obtained by inputting a latent variable at an arbitrary point on the similarity map to a learned decoder. The information processing apparatus according to claim 8.

10. The encoder and the decoder are generated by learning related to a variational autoencoder. The information processing apparatus according to claim 9.

11. The action performance data includes weight information related to financial products. The information processing apparatus according to claim 1.

12. A display unit that displays the similarity map. further comprising The information processing apparatus according to claim 1.

13. A processor analyzes a tendency of preference regarding the buying and selling transactions of financial products based on action performance data indicating the results of the buying and selling transactions of financial products. including The analyzing includes generating a similarity map that represents, on a two-dimensional plane, the similarity between at least the buying and selling transactions of the financial product by the analysis target person and the buying and selling transactions of the financial product by a designated comparison target person based on the action performance data, and expressing, in a heatmap form on the similarity map, the strength of preference of the analysis target person regarding the holding of the selected predetermined financial product and the strength of preference of the comparison target person. Information processing method.

14. A computer An analysis unit that analyzes a tendency of preference regarding the buying and selling transactions of financial products based on action performance data indicating the results of the buying and selling transactions of financial products. comprising The analysis unit generates a similarity map that represents, on a two-dimensional plane, the similarity between at least the buying and selling transactions of the financial product by the analysis target person and the buying and selling transactions of the financial product by a designated comparison target person based on the action performance data, and expresses, in a heatmap form on the similarity map, the strength of preference of the analysis target person regarding the holding of the selected predetermined financial product and the strength of preference of the comparison target person. Information processing apparatus A program for causing the computer to function as such.

Images