Information processing device, information processing method, and program

Abstract

Previously, it was not possible to effectively utilize information based on the user's screen transition actions. [Solution] An information processing device 1 comprises a reception unit 12 that receives operation information, which is information based on the user's screen transition operations; a screen information acquisition unit 131 that acquires screen information, which is information about all or part of each of two or more screens corresponding to the operation information received by the reception unit 12; a vectorization unit 134 that acquires a vector using each of the two or more screen information acquired by the screen information acquisition unit 131; a prediction unit 135 that acquires behavioral information about the user's actions using the vector acquired by the vectorization unit 134 and learning information used when outputting behavioral information about the user's actions using the vector; and an output unit 14 that outputs the behavioral information acquired by the prediction unit 135. This information processing device 1 makes effective use of information based on the user's screen transition operations.

Description

[Technical Field]

[0001] The present invention relates to an information processing device, etc., that acquires and outputs information about user behavior using a vector based on user operations. [Background technology]

[0002] Conventionally, there have been technologies that allow for content expansion in accordance with user trends when viewing content (see Patent Document 1). [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2024-73889 [Overview of the project] [Problems that the invention aims to solve]

[0004] However, conventional technologies could not effectively utilize information based on the user's screen transition operations. [Means for solving the problem]

[0005] The first information processing device of this invention comprises: a receiving unit that receives operation information which is information based on the user's screen transition operations; a screen information acquisition unit that acquires screen information for two or more screens corresponding to the operation information received by the receiving unit; a vectorization unit that acquires a vector using the two or more screen information acquired by the screen information acquisition unit; a prediction unit that acquires behavior information relating to the user's actions using the vector acquired by the vectorization unit and learning information used when outputting behavior information relating to the user's actions using the vector; and an output unit that outputs the behavior information acquired by the prediction unit.

[0006] This configuration allows for the effective use of information based on the user's screen transition operations.

[0007] Furthermore, the information processing device of the second invention, compared to the first invention, has a reception unit that receives operation information which is associated with a user identifier, and further comprises a user attribute value acquisition unit that acquires one or more user attribute values ​​associated with the user identifier associated with the operation information from a user management unit which stores one or more user attribute values ​​associated with one or more user identifiers, and a vectorization unit which acquires a vector using two or more screen information acquired by the screen information acquisition unit and one or more user attribute values ​​acquired by the user attribute value acquisition unit.

[0008] This configuration allows for the effective use of information based on the user's screen transition operations and user attribute values.

[0009] Furthermore, the information processing device of this third invention, compared to the first or second invention, further comprises an external information acquisition unit that acquires external information that is not present on two or more screens using at least a portion of the screen information acquired by the screen information acquisition unit, and the vectorization unit is an information processing device that acquires a vector using the two or more screen information acquired by the screen information acquisition unit and the external information acquired by the external information acquisition unit.

[0010] This configuration allows for the effective use of information based on the user's screen transition operations and external information that can be obtained using that information.

[0011] Furthermore, the information processing device of this fourth invention is an information processing device that, in addition to any one of the first to third inventions, has a vectorization unit that acquires a vector using the screen information for each of the two or more screen information acquired by the screen information acquisition unit, and acquires one vector using the two or more vectors, and a prediction unit that acquires behavior information using the one vector acquired by the vectorization unit and the learning information.

[0012] This configuration allows for the effective use of information based on the user's screen transition operations.

[0013] Furthermore, the information processing device of this fifth invention is an information processing device in which, with respect to any one of the first to fourth inventions, the prediction unit provides the vectors acquired by the vectorization unit to the generating AI and acquires behavioral information from the generating AI.

[0014] With this configuration, useful behavioral information can be obtained by feeding vectors acquired from information based on the user's screen transition operations to the generating AI.

[0015] Furthermore, the information processing device of the sixth invention is an information processing device in which, with respect to any one of the first to fourth inventions, the prediction unit acquires behavioral information by machine learning prediction processing using the vector acquired by the vectorization unit and a learning model acquired by machine learning processing using two or more training data having vectors and behavioral information.

[0016] With this configuration, useful behavioral information can be obtained through machine learning prediction processing using vectors acquired from information based on the user's screen transition operations and a learning model.

[0017] Furthermore, the information processing device of the seventh invention is an information processing device in which, with respect to any one of the first to fourth inventions, the prediction unit searches for a vector that satisfies a similarity condition to the vector using a correspondence table having two or more training data having vectors and behavioral information, using the vector acquired by the vectorization unit, and uses the behavioral information paired with the vector to acquire the behavioral information to be output.

[0018] With this configuration, useful behavioral information can be obtained by searching a correspondence table using vectors acquired from information based on the user's screen transition operations.

[0019] Furthermore, the information processing device of the eighth invention is an information processing device in which, with respect to any one of the first to fourth inventions, the prediction unit substitutes each element of the vector acquired by the vectorization unit into an arithmetic formula, executes the arithmetic formula, obtains the execution result, and obtains the action information corresponding to the execution result from a correspondence table having correspondence information that shows the correspondence between the execution result and the action information.

[0020] With such a configuration, useful action information can be obtained by executing an arithmetic expression into which elements of a vector obtained from information based on a user's screen transition operation are substituted.

[0021] Further, the information processing apparatus of the ninth invention is an information processing apparatus in which, with respect to any one of the first to eighth inventions, the action information is prediction information regarding the next operation of the user's screen transition operation.

[0022] With such a configuration, information based on the user's screen transition operation can be effectively utilized, and prediction information regarding the next operation of the user's screen transition operation can be obtained.

[0023] Further, the information processing apparatus of the tenth invention is an information processing apparatus in which, with respect to any one of the first to eighth inventions, the action information is user information of another user who performs an operation similar to that of the user.

[0024] With such a configuration, information based on the user's screen transition operation can be effectively utilized, and user information of another user who performs an operation similar to that of the user can be obtained.

[0025] Further, the information processing apparatus of the eleventh invention is an information processing apparatus in which, with respect to any one of the first to eighth inventions, the prediction unit further includes a post-processing unit that obtains the action information of two or more users and clusters the two or more users using the action information of the two or more users, and the output unit outputs a clustering result that is the result of the clustering performed by the post-processing unit.

[0026] With such a configuration, information based on the screen transition operations of two or more users can be effectively utilized, and the two or more users can be clustered.

[0027] Further, the information processing apparatus of the twelfth invention is an information processing apparatus in which, with respect to any one of the first to eighth inventions, the action information is explanatory information that explains the user's action based on the operation information.

[0028] This configuration allows for the effective use of information based on the user's screen transition operations, and enables the acquisition of user information for other users who perform similar operations to the user in question. [Effects of the Invention]

[0029] According to the information processing device of the present invention, information based on the user's screen transition operations can be effectively utilized. [Brief explanation of the drawing]

[0030] [Figure 1] Conceptual diagram of information processing system A in Embodiment 1 [Figure 2] Block diagram of the same information processing system A [Figure 3] Block diagram of the information processing device 1 [Figure 4] A flowchart illustrating an example of the operation of the information processing device 1. [Figure 5] A flowchart illustrating an example of the process for obtaining information from the same screen. [Figure 6] A flowchart illustrating an example of the vector acquisition process. [Figure 7] A flowchart illustrating the first example of this prediction process. [Figure 8] A flowchart illustrating a second example of the same prediction process. [Figure 9] A flowchart illustrating a third example of the prediction process. [Figure 10] A flowchart illustrating the fourth example of the same prediction process. [Figure 11] Block diagram of the computer system [Modes for carrying out the invention]

[0031] The embodiments of the information processing device, etc., will be described below with reference to the drawings. In the embodiments, components that are denoted by the same reference numerals perform the same operation, and therefore, further explanation may be omitted.

[0032] (Embodiment 1) In this embodiment, an information processing device is described that acquires all or part of the information of two or more screens based on the user's screen transition operations, constructs a vector from that information, and uses that vector and learned information to acquire and output behavioral information related to the user's actions.

[0033] This embodiment describes an information processing device that acquires external information other than all or part of the information on the screen, constructs a vector using that external information, and uses that vector and learned information to acquire and output behavioral information related to the user's actions.

[0034] In this embodiment, using learning information means, for example, using a generative AI, using a learning model obtained by machine learning, using a correspondence table, or using an arithmetic formula.

[0035] Furthermore, in this embodiment, the behavioral information includes, for example, prediction information for the next operation after a screen transition operation performed by the user, user information of users similar to the user who performed the screen transition operation, descriptive information explaining the user's actions based on the operation information, and the results of clustering users based on the operation information.

[0036] In this specification, information X being associated with information Y means that information Y can be obtained from information X, or information X can be obtained from information Y, and the method of association is irrelevant. Information X and information Y may be linked, may reside in the same buffer, may information X be contained in information Y, or information Y may be contained in information X, and so on.

[0037] Furthermore, in this specification, selecting or determining information Z means obtaining information Z, obtaining a pointer to information Z, obtaining the ID of information Z, setting a flag on information Z, etc., and it is sufficient to be able to access information Z.

[0038] Figure 1 is a conceptual diagram of information processing system A in this embodiment. Information processing system A comprises an information processing device 1, one or more terminal devices 2, and a generation AI device 3.

[0039] Information processing device 1 is usually a server, but it can also be a terminal. If information processing device 1 is a server, it can be, for example, a cloud server or an ASP server, but the type is not specified. If information processing device 1 is a terminal, it can be, for example, a personal computer, a smartphone, or a tablet device, but the type is not specified. Furthermore, if information processing device 1 is a terminal, it can be assumed that terminal device 2 is not necessary for information system A, or that information processing device 1 also functions as terminal device 2.

[0040] Terminal device 2 is a device used by the user. The user is, for example, someone who performs screen transitions. Terminal device 2 can be, for example, a personal computer, smartphone, or tablet device, but the type is not limited.

[0041] The generation AI device 3 is a device that has the functionality of a generation AI. In this context, the generation AI device 3 typically has the functionality of a text generation AI. The generation AI may be, for example, ChatGPT, Google Bard, or Gemini, but is not limited to these. Note that Google is a registered trademark. The generation AI device 3 may be, for example, a cloud server or an ASP server, but is not limited to these types. The generation AI device 3 will be referred to as the generation AI as appropriate. Furthermore, the information processing device 1 may also have the functionality of a generation AI. In such a case, the generation AI device 3 is not necessary in information processing system A.

[0042] Figure 2 is a block diagram of information processing system A in this embodiment. Figure 3 is a block diagram of information processing device 1.

[0043] The information processing device 1 comprises a storage unit 11, a reception unit 12, a processing unit 13, and an output unit 14. The storage unit 11 comprises a user management unit 111. The processing unit 13 comprises a learning unit 130, a screen information acquisition unit 131, an external information acquisition unit 132, a user attribute value acquisition unit 133, a vectorization unit 134, a prediction unit 135, and a post-processing unit 136.

[0044] The terminal device 2 includes a terminal storage unit 21, a terminal receiving unit 22, a terminal processing unit 23, a terminal transmission unit 24, a terminal receiving unit 25, and a terminal output unit 26.

[0045] The storage unit 11, which constitutes the information processing device 1, stores various types of information. These types of information include, for example, one or more training data sets, learning information, or user information. The learning information includes, for example, a learning model, a correspondence table, or an arithmetic formula.

[0046] The user management unit 111 stores one or more user information. User information refers to information about a user. User information includes a user identifier and one or more user attribute values. User attribute values ​​can be static or dynamic. Static attribute values ​​are usually unchanging attribute values, but can also be thought of as attribute values ​​that do not change moment by moment. Examples of static attribute values ​​include name, age, address, telephone number, credit card number, email address, user terminal identifier, user identifier, password, etc. The user terminal identifier is information that identifies the terminal device 2, and can be, for example, a cookie ID, session identifier, IP address, MAC address, etc. The user identifier is information that identifies the user, and can be, for example, an ID. The user identifier may also be a telephone number, credit card number, email address, etc. Dynamic attribute values ​​are attribute values ​​that can change dynamically due to user operations, etc. Examples of dynamic attribute values ​​include real-time dynamic attribute values ​​and history information-based dynamic attribute values. Real-time dynamic attribute values ​​are attribute values ​​that change moment by moment in real time. Real-time dynamic attribute values ​​include, for example, the time spent on the web page the user is currently viewing, and the number of web pages viewed during the current visit. Historical information-based dynamic attribute values ​​are dynamic attribute values ​​obtained using the history of operations during past visits. Historical information-based dynamic attribute values ​​include, for example, the number of purchases, the purchase amount, the total purchase amount, the average time spent, the average number of page views, the number of visits, the score, and the rank. The score is, for example, information indicating the importance of the user from the administrator's perspective. The score is calculated using one or more user attribute values ​​of the user and two or more pieces of information from one or more operation information received from the user's terminal device 2. The rank is the user's rank. The rank is, for example, information obtained using one or more user attribute values ​​(for example, the purchase amount). Note that the number of page views is the number of page views. Furthermore, the method of obtaining user attribute values ​​is not specified. The technology for obtaining user attribute values ​​is publicly known technology.

[0047] The reception unit 12 receives various types of information and instructions. For example, the reception unit 12 receives various types of information and instructions from the terminal device 2. These types of information and instructions include, for example, operation information and instructions to output action information.

[0048] Operation information refers to information obtained through the user's operation of the device. The device is, for example, terminal device 2, but is not limited to that. Preferably, the operation information here is information based on screen transition operations. However, operation information may also include information based on operations other than screen transition operations. Operation information includes, for example, screen information of the screen before the transition and screen information of the screen after the transition. A screen transition operation is an operation in which a screen changes. A screen transition usually involves switching from the current screen to the next screen, but it may also include the display of a pop-up or a separate window on the current screen, or a change in the content of the screen. Screen information refers to information about a screen. Screen information usually refers to all or part of the information of the screen. Screen information includes, for example, one or more types of information from HTML for constructing the screen, XML for constructing the screen, or a screen image, or one or more strings of one or more that are displayed on the screen. Screen information may also include video and audio. Screen information may also be screen summary information. Screen summary information is a sentence that describes what kind of screen it is, and one or more keywords that indicate what kind of screen it is. The screen summary information consists of the HTML, XML, or image used to construct the screen, or one or more strings displayed on the screen, and the information contained in the response obtained by the AI ​​generating a prompt to retrieve the screen summary information (for example, "Using the following image, summarize the content of the screen in 100 characters or less!").

[0049] Operation information includes, for example, information about an operation to jump to another screen, information indicating that a button on the screen was pressed, information indicating that an anchor on the screen was selected, and information entered into a field on the screen. Operation information may include, for example, a screen identifier for the screen that was operated on. Operation information may include, for example, an identifier for the selected object. Operation information may include, for example, a click ID. Operation information may include, for example, "rightButtonON" (indicating that the right mouse button was pressed), "click objectA" (indicating that the object identified as objectA was clicked), "drug object A" (indicating that object A was dragged), "<purchase product ID>123 <quantity>3" (indicating that 3 units of the product identified as 123 were purchased), "jump http: / / www.xxx.co.jp" (indicating that the webpage identified by the URL "http: / / www.xxx.co.jp" was accessed), etc. Operation information here may be, for example, information about primitive operations, but it may also be information that a person can see and judge the meaning and significance of the operation. In other words, the data structure and granularity of the operation information are not specified. Operation information typically identifies the actions performed by the user, but may also include information about the processing performed by the information processing device 1 as a result of the user's actions. Information that a human can see and judge the meaning and significance of an operation is, for example, information about one or more primitive operations (e.g., events, function names, or method names generated by the operation) given to the generating AI, and a prompt given to the generating AI to obtain information indicating the meaning and significance of the operation (e.g., "Explain the meaning of the operation using the following operation information!"), and the information contained in the response obtained from the generating AI.

[0050] A screen is anything displayed on the display medium of a device. A screen can be anything, such as a window, a web page, an application screen, a pop-up, or a single frame on a screen. A screen can be the entire screen displayed on the display medium, or just a part of the screen displayed on the display medium. A part of the screen displayed on the display medium includes windows within the screen, and web pages or application screens displayed in a specific area of ​​the screen.

[0051] An action information output instruction is an instruction to output action information. Action information output instructions typically include operation information.

[0052] Here, "reception" typically refers to the reception of information transmitted via wired or wireless communication lines, but it may also be a concept that includes the reception of information input from input devices such as keyboards, mice, and touch panels, as well as the reception of information read from recording media such as optical discs, magnetic discs, and semiconductor memory.

[0053] The processing unit 13 performs various processes. These processes include, for example, those performed by the learning unit 130, the screen information acquisition unit 131, the external information acquisition unit 132, the user attribute value acquisition unit 133, the vectorization unit 134, or the prediction unit 135.

[0054] The learning unit 130 acquires a learning model by machine learning processing using two or more training data sets from the storage unit 11. The training data is information that has vectors and behavioral information. The vectors are explanatory variables, and the behavioral information is the target variable. Preferably, the vectors are information acquired by the vectorization unit 134. The behavioral information is, for example, information entered manually, but the method of acquisition is not limited.

[0055] A learning model is information constructed through the learning process of machine learning, and is used in the prediction process of machine learning. A learning model can also be called a learner, classifier, or classification model. The machine learning algorithm can be deep learning, random forest, decision tree, SVR, etc. Furthermore, various machine learning functions and various existing libraries can be used in machine learning, such as the TensorFlow® library, the R language's random forest module, fastText, TinySVM, etc.

[0056] The screen information acquisition unit 131 acquires screen information, which is information about all or part of each of the two or more screens corresponding to the operation information received by the reception unit 12.

[0057] The screen information acquisition unit 131 acquires screen information, which is information about all or part of the screen, for each of the two or more screens corresponding to the operation information received by the reception unit 12.

[0058] The external information acquisition unit 132 uses at least some of the screen information acquired by the screen information acquisition unit 131 to acquire external information that is not present on two or more screens.

[0059] The external information acquisition unit 132, for example, acquires terms contained in the screen information acquired by the screen information acquisition unit 131, searches the database using those terms as keys, and acquires external information that corresponds to those terms. For example, if the destination screen is an e-commerce site screen, the external information acquisition unit 132, for example, acquires a product code from the screen information of that screen, searches the product database using that product code as a key, and acquires one or more pieces of external information that are not displayed on the screen, such as the color of the product or the price of the product.

[0060] The user attribute value acquisition unit 133 acquires one or more user attribute values ​​from the user management unit 111 that correspond to the user identifier associated with the operation information received by the reception unit 12.

[0061] The vectorization unit 134 obtains a vector using two or more screen information acquired by the screen information acquisition unit 131. Note that each of the two or more elements constituting the vector can be a number or a string. The data type of the elements is not restricted. The method of obtaining the vector using screen information is not restricted.

[0062] The vectorization unit 134 preferably constructs a vector using two or more types of information. Two or more types of information include, for example, two or more types of information from among the screen image, text strings on the screen, screen description information, operation information, user attribute values, and external information contained in the screen information.

[0063] The vectorization unit 134, for example, obtains a vector using the screen information for each of the two or more screen information acquired by the screen information acquisition unit 131, and then obtains a single vector using those two or more vectors.

[0064] The vectorization unit 134 obtains, for example, two or more strings from screen information. Next, the vectorization unit 134 constructs a vector using, for example, the two or more strings. For constructing a vector using two or more strings, known techniques such as "Bag of Words," "Word2Vec," "TF-IDF," "FastText," and "BERT" can be used. Note that the strings obtained from screen information are usually the strings displayed on the screen.

[0065] The vectorization unit 134 obtains a vector using, for example, source information, which is the information from which the vector is obtained. The vectorization unit 134 provides the source information to the generating AI and also provides the generating AI with a prompt to obtain a vector (for example, "Please output a vector from the following information (source information)."), and obtains a vector from the generating AI.

[0066] The source information is typically the information obtained by the screen information acquisition unit 131 from two or more individual screen information sets, or from two or more individual screen information sets. The source information is, for example, one or more types of information from among the screen image, text within the screen, screen description information, operation information, user attribute values, and external information contained in the screen information.

[0067] The vectorization unit 134 obtains a vector using, for example, two or more screen information acquired by the screen information acquisition unit 131 and external information acquired by the external information acquisition unit 132.

[0068] The vectorization unit 134, for example, obtains two or more image features from the screen image contained in the screen information, and obtains a vector whose elements are each of these two or more image features.

[0069] The vectorization unit 134 obtains, for example, a vector whose elements are the descriptive information of each of the two or more transitioned screens.

[0070] The vectorization unit 134 obtains, for example, a vector whose elements are two or more strings obtained by character recognition processing from the screen images of two or more transitioned screens.

[0071] The vectorization unit 134 may construct a single vector that includes a vector obtained from the generating AI and information obtained without using the generating AI.

[0072] The prediction unit 135 obtains behavioral information using one or more vectors acquired by the vectorization unit 134 and the learning information.

[0073] Behavioral information refers to information about a user's actions. Behavioral information is obtained based on the operation information received. Examples of behavioral information include prediction information for the next operation after a screen transition operation performed by the user, user information of users similar to the user who performed the screen transition operation, descriptive information that explains the user's actions based on the operation information, and the results of clustering users based on the operation information. However, the content of the behavioral information is not restricted. Behavioral information can be, for example, a string, an ID, a vector, etc., and its data type is not restricted.

[0074] One vector obtained by the vectorization unit 134 is, for example, a vector obtained by combining vectors obtained by the vectorization unit 134 for each screen. Two or more vectors obtained by the vectorization unit 134 are, for example, vectors obtained by the vectorization unit 134 for each screen.

[0075] Learning information refers to the information used to output behavioral information about user actions using vectors. Examples of learning information include the LLM of a generative AI, a learning model, a correspondence table, or an arithmetic formula.

[0076] The following describes an example of an algorithm used by the prediction unit 135 to acquire behavioral information. (1) When using a generative AI

[0077] The prediction unit 135 provides the one or more vectors acquired by the vectorization unit 134 to the generating AI and obtains action information from the generating AI.

[0078] For example, the prediction unit 135 provides the generating AI with one or more training data sets having vectors and behavioral information, and in a situation where the generating AI has learned from one or more training data sets, it is preferable for the vectorization unit 134 to provide the generating AI with one or more vectors acquired by the vectorization unit 134 and acquire behavioral information from the generating AI.

[0079] The prediction unit 135 provides the generating AI with, for example, one of the following prompt examples 1, 2, or 3, obtains a response from the generating AI, and acquires the behavioral information contained in that response. In the prompt examples, strings enclosed in "<" and ">" are variables. That is, the variable <vector> is assigned one or more vectors obtained by the vectorization unit 134. (1-1) Example prompt 1

[0080] Prompt template example 1 is: "The following are vectors obtained based on screen transition operations. Please output a summary of the screen transition operations indicated by the following vectors. [Vector]<Vector>"

[0081] Then, the prediction unit 135 obtains, for example, prompt template example 1 from the storage unit 11, and the vectorization unit 134 obtains one or more vectors (for example, (x1, x2, ..., x) from the variable <vector> of prompt template example 1. n Replace with )) and prompt example 1: "The following is a vector obtained based on screen transition operations. Output a summary of the screen transition operations indicated by the following vector. [Vector](x1,x2,···,y m) is obtained. Next, the prediction unit 135 provides the prompt example 1 to the generation AI and obtains an answer from the generation AI. Next, the prediction unit 135 obtains action information from the answer. Here, the action information is a summary of the screen transition operation. The action information is, for example, "The user moved from the web page 'Personal Computer Page' of the EC site to the purchase page for purchasing 'PC-AAA'."

[0082] Note that, (x1, x2, ···, y m ) is, for example, a single vector obtained by merging vectors for each screen information of two or more transitioned screens. (1-2) Prompt Example 2

[0083] The prompt template example 2 is "The following is a vector obtained based on the screen transition operation. Please output the operation predicted to be performed by the user after the screen transition operation indicated by the following vector. [Vector]<Vector>".

[0084] Then, the prediction unit 135, for example, obtains the prompt template example 2 from the storage unit 11, replaces the variable <Vector> of the prompt template example 2 with one or more vectors (for example, (x1, x2, ···, x n )(y1, y2, ···, y m )) obtained by the vectorization unit 134, and obtains the prompt example 2. Next, the prediction unit 135 provides the prompt example 2 to the generation AI and obtains an answer from the generation AI. Next, the prediction unit 135 obtains action information from the answer. Here, the action information is the operation predicted to be performed by the user after the screen transition operation. The action information is, for example, "Put the product 'PC-AAA' into the shopping cart."

[0085] Note that, among (x1, x2, ···, x n )(y1, y2, ···, y m ), (x1, x2, ···, x n ) is a vector obtained from the screen information of the screen before the transition, and (y1, y2, ···, y m) is a vector obtained from the screen information of the screen after the transition. (1-3) Example prompt 3

[0086] Prompt template example 3 is: "The following are vectors obtained based on the user's actions today. Please output the user information of users similar to the user performing the actions indicated by the following vectors from the user management table. The user management table is stored in the user management unit 111 of the information processing device 1. [Vector]<Vector>"

[0087] Then, the prediction unit 135 obtains, for example, prompt template example 3 from the storage unit 11, and the vectorization unit 134 obtains one or more vectors (for example, (x1, x2, ..., z) from the variable <vector> of prompt template example 3. n Replace it with )) and obtain prompt example 3. Next, the prediction unit 135 gives prompt example 3 to the generating AI and obtains a response from the generating AI. Next, the prediction unit 135 obtains behavioral information from the response. In this case, the behavioral information is user information of a user similar to the user in question. The behavioral information is, for example, "<User Information><Age>25 <Gender>Female <Hobbies>Travel...".

[0088] Note that (x1,x2,···,z n This is a vector obtained based on screen transition operations over a long period of time. (2) When using machine learning prediction processing

[0089] The prediction unit 135 uses the vectors acquired by the vectorization unit 134 and the learning model to obtain behavioral information through machine learning prediction processing.

[0090] The learning model is a model acquired by the learning unit 130 through machine learning training using two or more training data sets that contain vectors and behavioral information. (3) When using a correspondence table

[0091] The prediction unit 135 uses the vectors acquired by the vectorization unit 134 to search for one or more vectors that satisfy the similarity condition for each vector in a correspondence table having two or more training data sets having vectors and behavior information, and uses the behavior information paired with each of the one or more vectors to acquire the behavior information to be output.

[0092] The prediction unit 135, for example, determines the vector with the highest similarity to the vector obtained by the vectorization unit 134, and obtains the corresponding behavioral information. In this case, the similarity condition is that the similarity is maximized.

[0093] The prediction unit 135, for example, determines one or more vectors whose similarity to the vectors obtained by the vectorization unit 134 satisfies the similarity condition, obtains action information paired with each of these one or more vectors, and obtains a representative value of the one or more action information. The representative value is, for example, action information obtained by majority vote of the one or more action information. The similarity condition here is that the similarity is equal to or greater than a threshold, or the similarity is greater than a threshold. (4) When using an arithmetic formula

[0094] The prediction unit 135 substitutes each element of the vector obtained by the vectorization unit 134 into a calculation formula, executes the calculation formula, obtains the execution result, and obtains the action information corresponding to the execution result from a correspondence table that has correspondence information showing the correspondence between the execution result and the action information. The execution result is a numerical value.

[0095] The post-processing unit 136 performs processing using the behavior information of one or more users and obtains the behavior information to be output.

[0096] The post-processing unit 136 clusters two or more users using the behavioral information of each user. Clustering determines the user's class. Clustering involves, for example, obtaining a class identifier by associating it with a user identifier. The classes can be, for example, five levels from "1" to "5" or three levels such as "Gold," "Silver," and "Bronze," but are not limited to these. Note that the process of clustering information (in this case, behavioral information) is a publicly known technique.

[0097] The post-processing unit 136 presents information to the user, for example, using the behavioral information of one or more users. The post-processing unit 136 retrieves information from a correspondence table that corresponds to the conditions that the user's behavioral information satisfies, and transmits this information to the user's terminal device 2. In this case, the correspondence table has two or more correspondence pieces of information that have conditions and information that the user's behavioral information satisfies.

[0098] The output unit 14 outputs various types of information. The output unit 14 normally transmits various types of information to the terminal device 2. The output unit 14 outputs behavioral information acquired by the prediction unit 135. The output unit 14 outputs information acquired by the post-processing unit 136.

[0099] Here, output usually refers to transmission to terminal device 2, but it may also be a concept that includes display on a screen, projection using a projector, printing with a printer, sound output, transmission to devices other than terminal device 2, storage on a recording medium, and delivery of processing results to other processing devices or other programs.

[0100] The terminal storage unit 21, which constitutes the terminal device 2, stores various types of information. These types of information include, for example, operation information and user identifiers.

[0101] The terminal reception unit 22 receives various types of information and instructions. These types of information and instructions include, for example, operation information and operations.

[0102] The means of inputting various information and instructions can be anything, such as a touch panel, keyboard, mouse, or menu screen.

[0103] The terminal processing unit 23 performs various processes. These processes include, for example, converting received information and instructions into information and instructions for transmission. Other processes include, for example, converting received information into information for output.

[0104] The terminal transmission unit 24 transmits various information and instructions to the information processing device 1 or other devices (not shown). These various information and instructions include, for example, operation information.

[0105] The terminal receiving unit 25 receives various types of information. These types of information include, for example, web pages.

[0106] The terminal output unit 26 outputs various types of information. These types of information include, for example, various screens.

[0107] The storage unit 11, the user management unit 111, and the terminal storage unit 21 are preferably made of non-volatile recording media, but can also be made of volatile recording media.

[0108] The process by which information is stored in the storage unit 11, etc. is not relevant. For example, information may be stored in the storage unit 11, etc. via a recording medium, information transmitted via a communication line, etc. may be stored in the storage unit 11, etc., or information input via an input device may be stored in the storage unit 11, etc.

[0109] The reception unit 12 is preferably implemented by wireless or wired communication means, but may also be implemented by means of receiving broadcasts, device drivers for input means such as touch panels and keyboards, or control software for menu screens.

[0110] The processing unit 13, screen information acquisition unit 131, external information acquisition unit 132, user attribute value acquisition unit 133, vectorization unit 134, prediction unit 135, post-processing unit 136, and terminal processing unit 23 can typically be implemented using a processor, memory, etc. The processing procedures of the processing unit 13, etc., are typically implemented in software, and this software is recorded on a recording medium such as ROM. However, it may also be implemented in hardware (dedicated circuitry). The processor can be a CPU, MPU, GPU, etc., and the type is not limited.

[0111] The output unit 14 is usually implemented by wireless or wired communication means, but it may also be implemented by driver software for an output device such as a display or speaker, or by driver software for an output device and the output device itself.

[0112] The terminal reception unit 22 can be implemented using device drivers for input means such as touch panels and keyboards, or control software for menu screens, etc.

[0113] The terminal transmission unit 24 is usually implemented by wireless or wired communication means, but it may also be implemented by broadcasting means.

[0114] The terminal receiving unit 25 is usually implemented by wireless or wired communication means, but it may also be implemented by means of receiving broadcasts.

[0115] The terminal output unit 26 may or may not be considered to include output devices such as a display or speakers. The terminal output unit 26 can be implemented using driver software for an output device, or driver software for an output device and an output device.

[0116] Next, an example of the operation of the information processing device 1 will be explained using the flowchart in Figure 4.

[0117] (Step S401) The reception unit 12 determines whether or not it has received an action information output instruction. If it has received an action information output instruction, it proceeds to step S402; otherwise, it returns to step S401.

[0118] (Step S402) The screen information acquisition unit 131 acquires operation information from the action information output instruction received in step S401.

[0119] (Step S403) The screen information acquisition unit 131 acquires screen information using the operation information acquired in step S402. An example of such screen information acquisition process will be explained using the flowchart in Figure 5.

[0120] (Step S404) The vectorization unit 134, etc., acquires a vector using the screen information acquired in step S403. An example of such vector acquisition process will be explained using the flowchart in Figure 6.

[0121] (Step S405) The prediction unit 135 performs prediction processing using the vector obtained in step S404 and obtains behavioral information. An example of such prediction processing will be explained using the flowcharts in Figures 7, 8, 9, and 10.

[0122] (Step S406) The output unit 14 outputs the action information acquired in step S405. Return to step S401.

[0123] In the flowchart shown in Figure 4, processing is terminated by power-off or processing termination interrupts.

[0124] Next, an example of the screen information acquisition process in step S403 will be explained using the flowchart in Figure 5.

[0125] (Step S501) The screen information acquisition unit 131 assigns 1 to counter i.

[0126] (Step S502) The screen information acquisition unit 131 determines whether or not the i-th operation information exists. If the i-th operation information exists, the process proceeds to step S503; otherwise, it returns to the higher-level process.

[0127] (Step S503) The screen information acquisition unit 131 acquires the i-th operation information that has been received.

[0128] (Step S504) The screen information acquisition unit 131 determines whether the i-th operation information acquired in step S503 is operation information that specifies an operation that will change the display screen. If it is operation information that specifies an operation that will change the display screen, the unit proceeds to step S505; otherwise, the unit proceeds to step S506. Note that the operations that will change the display screen are predetermined, such as operations that include "jump" or operations that include "show".

[0129] (Step S505) The screen information acquisition unit 131 acquires the screen information of the changed display screen and temporarily stores the screen information in a buffer (not shown).

[0130] In this context, screen information includes, for example, one or more strings displayed on the screen, a screen image, and one or more strings obtained by character recognition processing on the screen image. The screen information acquisition unit 131 also temporarily stores the acquired screen information in a buffer (not shown) by associating it with the i-th operation information.

[0131] Furthermore, the screen information here refers to, for example, a screen image, and descriptive information of the screen obtained by providing the generating AI with one or more types of information from one or more strings obtained by character recognition processing on the screen image.

[0132] (Step S506) The screen information acquisition unit 131 increments counter i by 1. The process returns to step S502.

[0133] Next, an example of the vector acquisition process in step S404 will be explained using the flowchart in Figure 6.

[0134] (Step S601) The vectorization unit 134 assigns 1 to counter i.

[0135] (Step S602) The vectorization unit 134 determines whether or not the i-th screen information exists. If the i-th screen information exists, the unit proceeds to step S603; otherwise, the unit proceeds to step S612.

[0136] (Step S603) The vectorization unit 134 obtains one or more source information from the i-th screen information. Source information is information used to create a vector. Source information includes, for example, a screen image, a string, and feature quantities of the screen image. The string is usually a term or sentence displayed on the screen that the user has transitioned to.

[0137] (Step S604) The external information acquisition unit 132 assigns counter j1.

[0138] (Step S605) The external information acquisition unit 132 determines whether the j-th piece of original information exists among the original information acquired in step S603. If the j-th piece of original information exists, the unit proceeds to step S606; otherwise, the unit proceeds to step S610.

[0139] (Step S606) The external information acquisition unit 132 acquires the class of the j-th piece of original information. The class can also be called a category or type. If the original information is a term, the class is, for example, whether or not it is a product name.

[0140] (Step S607) The external information acquisition unit 132 determines whether the class acquired in step S606 satisfies the external information acquisition conditions. If the external information acquisition conditions are met, the unit proceeds to step S608; otherwise, it proceeds to step S609.

[0141] (Step S608) The external information acquisition unit 132 searches the database where the external information is stored using the j-th original information as a key, acquires one or more external information that is paired with the j-th original information, and associates them with the j-th original information.

[0142] (Step S609) The external information acquisition unit 132 increments counter j by 1. Return to step S605.

[0143] (Step S610) The vectorization unit 134 constructs a vector using one or more source information obtained in step S603, or one or more source information obtained in step S603 and one or more external information obtained in step S608.

[0144] If the source information and external information are strings, the vectorization unit 134 obtains a vector from one or more strings. As described above, there are various ways to obtain a vector from strings.

[0145] (Step S611) The vectorization unit 134 increments counter i by 1. Return to step S602.

[0146] (Step S612) The vectorization unit 134 constructs a single vector using two or more vectors obtained in step S610. It then returns to the higher-level processing. The single vector constructed using two or more vectors is, for example, a combined vector formed by combining vectors based on two or more screen information. The combined vector is (vector based on the first screen information, vector based on the second screen information, ..., vector based on the nth screen information).

[0147] In the flowchart of Figure 6, the vectorization unit 134 may obtain two or more vectors without performing step S612.

[0148] Furthermore, in the flowchart of Figure 6, the vectorization unit 134 does not acquire a vector for each piece of screen information, but may acquire a single vector using two or more pieces of source information acquired from two or more pieces of screen information.

[0149] Furthermore, in the flowchart of Figure 6, it is preferable that the user attribute value acquisition unit 133 acquires one or more user attribute values ​​from the user management unit 111 that are paired with the user identifier included in the action output instruction, and that the vectorization unit 134 constructs a vector using these one or more user attribute values.

[0150] Next, we will explain the first example of the prediction process in step S405 using the flowchart in Figure 7. The first example is when the prediction unit 135 uses a generated AI.

[0151] (Step S701) The prediction unit 135 acquires one or more training data from the storage unit 11, passes the one or more training data to the generating AI, and allows the generating AI to learn.

[0152] (Step S702) The prediction unit 135 obtains a prompt template from the storage unit 11.

[0153] (Step S703) The prediction unit 135 obtains the vector obtained in step S404.

[0154] (Step S704) The prediction unit 135 places the vector obtained in step S703 into the variable <vector> in the prompt template to construct the prompt.

[0155] (Step S705) The prediction unit 135 provides the prompt configured in step S704 to the generation AI.

[0156] (Step S706) The prediction unit 135 determines whether or not it has obtained an answer from the generating AI. If an answer has been obtained, it proceeds to step S707; otherwise, it returns to step S706.

[0157] (Step S707) The prediction unit 135 obtains behavioral information based on the response obtained in step S706. It returns to the higher-level processing.

[0158] Next, a second example of the prediction process in step S405 will be explained using the flowchart in Figure 8. The second example is when the prediction unit 135 uses machine learning prediction processing.

[0159] (Step S801) The prediction unit 135 obtains the vector obtained in step S404.

[0160] (Step S802) The prediction unit 135 retrieves the learning model from the storage unit 11.

[0161] (Step S803) The prediction unit 135 provides the vector and the learning model to the machine learning prediction module and executes the prediction module.

[0162] (Step S804) The prediction unit 135 obtains the behavior information which is the result of the prediction module's execution. It then returns to the higher-level processing.

[0163] Next, a third example of the prediction process in step S405 will be explained using the flowchart in Figure 9. The third example is when the prediction unit 135 uses a correspondence table.

[0164] (Step S901) The prediction unit 135 obtains the vector obtained in step S404.

[0165] (Step S902) The prediction unit 135 assigns 1 to counter i.

[0166] (Step S903) The prediction unit 135 determines whether the i-th correspondence information exists in the correspondence table of the storage unit 11. If the i-th correspondence information exists, the unit proceeds to step S904; otherwise, the unit proceeds to step S906.

[0167] (Step S904) The prediction unit 135 obtains the vector contained in the i-th corresponding information. The prediction unit 135 calculates the similarity between this vector and the vector obtained in step S901.

[0168] (Step S905) The prediction unit 135 increments counter i by 1. The process returns to step S903.

[0169] (Step S906) The prediction unit 135 determines one or more vectors whose similarity calculated in step S904 satisfies the similarity condition. The prediction unit 135 obtains the corresponding behavioral information from the correspondence table for each of the one or more vectors determined.

[0170] (Step S907) The prediction unit 135 determines whether the behavioral information obtained in step S906 is 2 or more. If it is 2 or more, it proceeds to step S908; otherwise, it returns to the higher-level process.

[0171] (Step S908) The prediction unit 135 obtains a representative value of two or more behavioral information obtained in step S906. It returns to the higher-level processing. This representative value is the behavioral information to be output.

[0172] Next, a fourth example of the prediction process in step S405 will be explained using the flowchart in Figure 10. The fourth example is when the prediction unit 135 uses a calculation formula.

[0173] (Step S1001) The prediction unit 135 obtains the vector obtained in step S404.

[0174] (Step S1002) The prediction unit 135 obtains the calculation formula from the storage unit 11.

[0175] (Step S1003) The prediction unit 135 substitutes two or more elements that are included in the vector obtained in step S404 and that become parameters of the calculation formula into the calculation formula.

[0176] (Step S1004) The prediction unit 135 executes the calculation formula and obtains the calculation result.

[0177] (Step S1005) The prediction unit 135 acquires action information corresponding to the calculation result obtained in step S1004 and temporarily stores the action information in a buffer (not shown). It then returns to the higher-level processing.

[0178] As described above, according to this embodiment, information based on the user's screen transition operations can be effectively utilized.

[0179] Furthermore, according to this embodiment, information based on the user's screen transition operations and user attribute values ​​can be effectively utilized.

[0180] Furthermore, according to this embodiment, information based on the user's screen transition operations and external information that can be obtained using such information can be effectively utilized.

[0181] Furthermore, according to this embodiment, useful behavioral information can be obtained by providing the generating AI with vectors acquired from information based on the user's screen transition operations.

[0182] Furthermore, according to this embodiment, useful behavioral information can be obtained through machine learning prediction processing using vectors acquired from information based on the user's screen transition operations and a learning model.

[0183] Furthermore, according to this embodiment, useful behavioral information can be obtained by searching a correspondence table using vectors acquired from information based on the user's screen transition operations.

[0184] Furthermore, according to this embodiment, useful behavioral information can be obtained by executing an arithmetic expression in which the elements of a vector obtained from information based on the user's screen transition operations are substituted.

[0185] Furthermore, according to this embodiment, information based on the user's screen transition operations can be effectively utilized to obtain predictive information regarding the user's next operation in the screen transition process.

[0186] Furthermore, according to this embodiment, information based on the user's screen transition operations can be effectively utilized to obtain user information of other users who perform operations similar to those of the user in question.

[0187] Furthermore, according to this embodiment, information based on the screen transition operations of two or more users can be effectively utilized, and these two or more users can be clustered.

[0188] The processing in this embodiment may be implemented in software. This software may be distributed via software download or the like. Alternatively, this software may be recorded on a recording medium such as a CD-ROM and distributed. This also applies to other embodiments described herein. The software that implements information processing 1 in this embodiment is the following program. In other words, this program enables the computer to... A reception unit that receives operation information, which is information based on the user's screen transition operations, A screen information acquisition unit acquires screen information which is information about all or part of each of the two or more screens corresponding to the operation information received by the reception unit, A vectorization unit that acquires a vector using two or more screen information acquired by the screen information acquisition unit, A prediction unit obtains behavioral information about the user's actions using the vectors obtained by the vectorization unit and the learning information used when outputting behavioral information about the user's actions using the vectors. This is a program to function as an output unit that outputs the behavioral information acquired by the prediction unit.

[0189] Figure 11 is a block diagram of a computer system 300 that executes the program described herein to realize the various embodiments of the information processing device 1 described above.

[0190] In Figure 11, the computer system 300 includes a computer 301 with a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.

[0191] In Figure 11, the computer 301 includes, in addition to the CD-ROM drive 3012, an MPU 3013, a bus 3014 connected to the CD-ROM drive 3012, a ROM 3015 for storing programs such as boot-up programs, a RAM 3016 connected to the MPU 3013 for temporarily storing instructions for application programs and providing temporary storage space, and a hard disk 3017 for storing application programs, system programs, and data. Although not shown here, the computer 301 may further include a network card for providing connectivity to a LAN.

[0192] The program that causes the computer system 300 to execute the functions of the information processing device 1, etc., as described above, may be stored on the CD-ROM 3101, inserted into the CD-ROM drive 3012, and then transferred to the hard disk 3017. Alternatively, the program may be transmitted to the computer 301 via a network (not shown) and stored on the hard disk 3017. The program is loaded into the RAM 3016 during execution. The program may also be loaded directly from the CD-ROM 3101 or the network.

[0193] The program sends the information processing device 1 of the above embodiment to the computer 301. The operating system (OS) or third-party programs that perform such functions do not necessarily have to be included. The program only needs to contain the instruction portion that calls the appropriate functions (modules) in a controlled manner and obtains the desired result. How the computer system 300 operates is well known, so a detailed explanation is omitted.

[0194] In the above program, steps such as sending information and receiving information do not include hardware-based processing, such as processing performed by a modem or interface card in the transmission step (processing that can only be performed by hardware).

[0195] Furthermore, the computer running the above program may be a single computer or multiple computers. In other words, it may perform centralized processing or distributed processing.

[0196] Furthermore, it goes without saying that in each of the above embodiments, two or more communication means present in a single device may be physically implemented in a single medium.

[0197] Furthermore, in each of the above embodiments, each process may be implemented by centralized processing by a single device, or by distributed processing by multiple devices.

[0198] It goes without saying that the present invention is not limited to the embodiments described above, and various modifications are possible, all of which are also included within the scope of the present invention. [Industrial applicability]

[0199] As described above, the information processing device 1 according to the present invention has the effect of being able to effectively utilize information based on the user's screen transition operations, and is useful as a server or the like for receiving operation information. [Explanation of symbols]

[0200] 1. Information Processing Device 2 Terminal devices 3 Generation AI device 11 Storage Unit 12 Reception Department 13 Processing Unit 14 Output section 21 Terminal storage section 22 Terminal Reception Section 23 Terminal Processing Unit 24 Terminal transmission unit 25 Receiving part of the terminal 26 Terminal output section 111 User Management Department 130 Learning Department 131 Screen information acquisition section 132 External Information Acquisition Department 133 User attribute value acquisition unit 134 Vectorization section 135 Prediction Section 136 Post-processing

Claims

1. A reception unit that receives operation information, which is information based on the user's screen transition operations, A screen information acquisition unit acquires screen information for two or more screens corresponding to the operation information received by the reception unit, A vectorization unit that acquires a vector using two or more pieces of screen information acquired by the screen information acquisition unit, A prediction unit obtains behavioral information about the user's actions using the vectors obtained by the vectorization unit and the learning information used when outputting behavioral information about the user's actions using the vectors. An information processing device comprising: an output unit that outputs the behavioral information acquired by the prediction unit.

2. The operation information received by the aforementioned reception unit is associated with a user identifier. The system further comprises a user attribute value acquisition unit that acquires one or more user attribute values ​​corresponding to a user identifier corresponding to the operation information from a user management unit that stores one or more user attribute values ​​corresponding to one or more user identifiers, The vectorization unit, The information processing apparatus according to claim 1, which acquires a vector using two or more screen information acquired by the screen information acquisition unit and one or more user attribute values ​​acquired by the user attribute value acquisition unit.

3. The system further comprises an external information acquisition unit that uses at least a portion of the screen information acquired by the screen information acquisition unit to acquire external information that is not present in the two or more screens, The vectorization unit, The information processing apparatus according to claim 1, which acquires a vector using two or more screen information acquired by the screen information acquisition unit and the external information acquired by the external information acquisition unit.

4. The vectorization unit, For each of the two or more screen information acquired by the screen information acquisition unit, a vector is acquired using the screen information, and a single vector is acquired using the two or more vectors. The prediction unit, The information processing apparatus according to claim 1, wherein the vectorization unit acquires the one vector and the learning information to acquire the behavior information.

5. The prediction unit, The information processing apparatus according to any one of claims 1 to 4, wherein the vectorization unit provides the vector acquired to a generating AI and acquires the behavior information from the generating AI.

6. The prediction unit, The information processing apparatus according to any one of claims 1 to 4, wherein the vectors acquired by the vectorization unit and a learning model acquired by machine learning processing using two or more training data having vectors and behavioral information are used to acquire the behavioral information by machine learning prediction processing.

7. The prediction unit, An information processing apparatus according to any one of claims 1 to 4, which uses a vector acquired by the vectorization unit to search for a vector that satisfies a similarity condition to a correspondence table having two or more training data having vectors and behavior information, and uses the behavior information paired with the vector to acquire the behavior information to be output.

8. The prediction unit, The information processing apparatus according to any one of claims 1 to 4, wherein each element of the vector acquired by the vectorization unit is substituted into an arithmetic expression, the arithmetic expression is executed, the execution result is obtained, and the action information corresponding to the execution result is obtained from a correspondence table having correspondence information showing the correspondence between the execution result and the action information.

9. The information processing apparatus according to claim 1, wherein the behavioral information is predictive information regarding the next operation of the user's screen transition operation.

10. The information processing apparatus according to claim 1, wherein the behavioral information is user information of other users who perform operations similar to the user.

11. The prediction unit, Obtain behavioral information for two or more users, The system further comprises a post-processing unit that clusters the two or more users using the behavioral information of each of the two or more users. The output unit is, The information processing apparatus according to claim 1, which outputs a clustering result, which is the result of clustering performed by the post-processing unit.

12. The information processing apparatus according to claim 1, wherein the behavioral information is descriptive information that explains the user's actions based on the operation information.

13. An information processing method implemented by a reception unit, a screen information acquisition unit, a vectorization unit, a prediction unit, and an output unit, The reception unit receives operation information, which is information based on the user's screen transition operations, in a reception step, The screen information acquisition step involves the screen information acquisition unit acquiring screen information which is information of all or part of each of the two or more screens corresponding to the operation information received by the reception unit, The vectorization unit performs a vectorization step of obtaining a vector using two or more screen information acquired by the screen information acquisition unit, The prediction unit performs a prediction step in which it obtains behavioral information relating to the user's actions using the vectors obtained by the vectorization unit and the learning information used when outputting behavioral information relating to the user's actions using the vectors. An information processing method comprising: an output step in which the output unit outputs the behavior information acquired by the prediction unit.

14. Computers, A reception unit that receives operation information, which is information based on the user's screen transition operations, A screen information acquisition unit acquires screen information which is information about all or part of each of the two or more screens corresponding to the operation information received by the reception unit, A vectorization unit that acquires a vector using two or more pieces of screen information acquired by the screen information acquisition unit, A prediction unit obtains behavioral information about the user's actions using the vectors obtained by the vectorization unit and the learning information used when outputting behavioral information about the user's actions using the vectors. A program to function as an output unit that outputs the behavioral information acquired by the prediction unit.

Images