JP7001559B2 - Publishing device, publishing method and publishing program - Google Patents

Description

The present invention relates to a public device, a public method and a public program.

Conventionally, there is known a technique of generating information related to the input information based on the analysis result of the input information and outputting the generated information as a response. As an example of such a technique, words, sentences, and contexts contained in the input text are converted into a multidimensional vector and analyzed, and based on the analysis result, text similar to the input text or input. A natural language processing technique that infers the text following the text and outputs the analogy result is known.

Japanese Unexamined Patent Publication No. 2015-22433 JP-A-2017-59082

“DEEPCODER: LEARNING TO WRITE PROGRAMS”, Matej Balog, Alexander L. Gaunt, Marc Brockschmidt, Sebastian Nowozin, Daniel Tarlow, Published as a conference paper at ICLR 2017 "Introduction to Neural Turing Machine", Naoto Moriyama, <Internet> https://www.slideshare.net/naotomoriyama/ss-69549087 (Searched May 1, 2018) “Dynamics on expanding spaces: modeling the emergence of novelties”, Vittorio Loreto, Vito D.P. Servedio, Steven H. Strogatz, Francesca Tria <Internet>

In recent years, the life cycle of services provided to users has become shorter. However, conventionally, no proposal has been made for a method for automating the process from drafting to provision of a service to a user.

The present application has been made in view of the above, and an object thereof is to automate the process from drafting to provision of a service.

The publishing device according to the present application is a word set included in a sentence indicating a predetermined process, and uses an extraction unit for extracting a word set having a predetermined structure and a word set extracted by the extraction unit. It is characterized by having a generation unit that generates a program that realizes a predetermined process, and a public unit that publishes the program generated by the generation unit.

According to one aspect of the embodiment, it is possible to automate the process from drafting to provision of a service.

FIG. 1 is a diagram showing an example of a process executed by a public device according to an embodiment. FIG. 2 is a diagram showing a configuration example of a public device according to an embodiment. FIG. 3 is a diagram showing an example of information registered in the literature database according to the embodiment. FIG. 4 is a diagram showing an example of information registered in the technical graph database according to the embodiment. FIG. 5 is a diagram showing an example of information registered in the service information database according to the embodiment. FIG. 6 is a flowchart illustrating an example of a flow of processing executed by the public device according to the embodiment. FIG. 7 is a diagram showing an example of a hardware configuration.

Hereinafter, a public device, a public method, and a mode for implementing a public program (hereinafter, referred to as “embodiment”) according to the present application will be described in detail with reference to the drawings. It should be noted that this embodiment does not limit the publishing device, publishing method, and publishing program according to the present application. Further, in each of the following embodiments, the same parts are designated by the same reference numerals, and duplicate explanations are omitted.

[Embodiment]
[1-1. An example of a public device]
First, an example of the process executed by the public device will be described with reference to FIG. In the following description, as an example of the process executed by the publishing device, an application (that is, software) for providing a predetermined service is automatically generated, and the generated application is published, and the application is used. An example of a process of accepting an accompanying user registration and providing an application and a list of users who use the application (that is, a user list) for sale will be described, but the embodiment is limited to this. is not it.

FIG. 1 is a diagram showing an example of a process executed by a public device according to an embodiment. In FIG. 1, the public device 10 is an information processing device that executes a determination process described below, and is realized by, for example, a server device, a cloud system, or the like.

More specifically, the public device 10 communicates with an arbitrary device such as a management device 100, a business device 200, and a user terminal 300 (see, for example, FIG. 2) via a predetermined network N such as the Internet. Is possible.

The management device 100 is an information processing device that manages various contents, and is realized by, for example, a server device, a cloud system, or the like. For example, the management device 100 manages contents such as papers belonging to various technical fields and publications related to patent applications. The management device 100 may manage arbitrary information including texts, such as recipes posted by users and posts to SNS (Social Networking Service), as contents.

The business operator device 200 is a terminal device used by various business operators, and is realized by, for example, a server device, a cloud system, a PC (Personal Computer), or the like. For example, when the business operator device 200 accepts a set of an application and a user list as service information indicating a service that can be sold, it presents the contents of the received service information to the business operator and purchases such service information. Has a function of notifying the public device 10.

[1-2. Overview of public processing and provision processing]
Here, the prior art will be described. The life cycle of services provided to users is shortening. However, conventionally, no proposal has been made for a method for automating the process from drafting to provision of a service to a user.

Therefore, the publishing device 10 executes the following publishing process. First, the publishing device 10 extracts a word set having a predetermined structure, which is a word set included in a sentence indicating a predetermined process. Subsequently, the publishing device 10 uses the extracted word set to generate a program that realizes a predetermined process. Then, the publishing device 10 publishes the generated program.

Further, the public device 10 executes the following provision process. First, the public device 10 acquires information on a user who uses the published program. Then, the public appliance 10 provides the service information in which the information about the program and the information of the user who uses the program are associated with each other as the information about the service that can be sold to the business operator or the like.

That is, the publishing device 10 generates a program from the word set and publishes the generated program. Further, the public device 10 receives the user's information registration regarding the use of the published program. Then, the public device 10 discloses the service information, which is a set of the program and the user list of the user who uses the program, as a sales target. That is, the public device 10 automatically generates a program that realizes a new business, and sells the automatically generated program and the information of the user who uses the program as a set. As a result of such processing, the public device 10 can automate a series of flows of service generation and service sale, in other words, a series of company generation and company sale.

[1-3. About program generation]
Here, the publishing device 10 generates a program from the word set. For example, the publishing device 10 extracts, as a word set, a word set including a word that becomes a noun, a word or a word set that becomes an adverbial phrase, and a word that becomes a verb related to the word that becomes a noun. Then, the publishing device 10 generates a program so that when the information indicated by the noun phrase included in the word set is input, the information indicated by the adverbial phrase included in the word set is output.

To give a specific example, for example, a sentence that may explain the characteristics of an idea or invention, that is, the characteristics of an idea, is a noun indicating the object of processing, a state expression word indicating the state of the object indicated by the noun, or an adverbial phrase. , And a verb that indicates the content of the process. In addition, in such a sentence, if the part of speech that can exist between words such as particles is enclosed in curly braces, "(noun) {to} (state expression word, adverbial phrase) {by, with} (verb). ) {Can, can} ”is expected to have a structure.

Therefore, the publishing device 10 extracts a sentence having the above-mentioned structure from the content belonging to a predetermined field, and conceives three words included in the extracted sentence, that is, a noun, a state expression word or an adverbial phrase, and a verb. Is extracted as a word set indicating. The word set extracted in this way can reproduce a sentence showing an idea such as an invention or an idea only by adding a predetermined case particle. That is, it is possible to show the concept of an idea such as an invention or an idea.

Here, when the word sets "A", "B", and "C" are extracted from the sentence "A is C by B" and the information indicated by "A" is input, "C" is used. When the program is generated so as to output the indicated information, it is considered that the publishing device 10 can generate the program that realizes the concept indicated by the text. Therefore, the publishing device 10 automatically generates a program so as to output the information indicated by the verb "C" when the information indicated by the noun "A" extracted from the word set is input.

That is, a program applicable to "B" can be generated by collecting pairs of teacher data of the input "A" and the output "C" and changing the internal parameters by the program automatic generation device utilizing the neural Turing machine. Alternatively, some programs may be automatically generated by changing some arbitrary internal parameters, and those applicable to the word "B" may be extracted from the automatically generated programs.

For example, the public device 10 has the sentence "grouping users by attribute information", the "user" corresponding to the above "A", the "attribute information" corresponding to the above "B", and the above "C". Extract the "grouping" corresponding to. Then, the public device 10 collects various information about the user as the information indicated by the word "user". It also collects information about "grouping". For example, the public device 10 inputs gender information indicated by the word "user", a pair of male and female grouping information, and the like into a neural Turing machine device that automatically generates a program as teacher data, and "attribute information" as an internal parameter. By making adjustments so that the weight of the thing related to "" becomes large, the program intended by these three words can be automatically generated.

It should be noted that such information may be acquired, for example, based on the similarity between the item name of the database in which various data are stored and the noun extracted as a word set. For example, when the similarity between the distributed expression of the item registered in the database and the distributed expression of the extracted noun exceeds a predetermined threshold value, the public device 10 collects the information of the item as the information indicated by the noun. Just do it. Further, for example, the publishing device 10 identifies and identifies an item corresponding to a distributed expression similar to the sum of the distributed expression of the extracted noun and the distributed expression of the extracted adverbial phrase from each item of the database. All you have to do is get the item information.

Further, the public device 10 collects information corresponding to "grouping". For example, the publishing device 10 collects information corresponding to the words "user" and "grouping". To give a more specific example, the public appliance 10 collects various information indicating the group to which the user belongs from the database.

Then, the public appliance 10 automatically generates a program from the collected information. For example, the publishing device 10 generates a program using a neural Turing machine so that when the information indicated by the extracted noun word is input, the information indicated by the verb word is output. For example, as shown in Non-Patent Document 1, a method has been proposed in which a program for generating output data from input data is automatically generated by using a neural Turing machine which is a Turing machine using a neural network.

Such a neural Turing machine M1 is, for example, as shown in FIG. 1, a memory (Memory), a read head (ReadHeads) of a Turing machine, a write head (WriteHeads) of a Turing machine, and an input (External Input). ), It has a controller (Control) that operates a read head and a write head and generates an output (External Output) according to an input. Further, in such a neural tuling machine M1, the controller is realized by a recurrent neural network such as RNN (Recurrent Neural Network) or LSTM (Long short-term memory).

When such a neural Turing machine M1 is trained to generate a certain output data from a certain input data, a program for outputting a certain output data from a certain input data is stored in the memory of the neural Turing machine M1. It will be stored. Therefore, the publishing device 10 realizes automatic generation of a program by learning the neural Turing machine M1 so as to output the information indicated by the verb from the information indicated by the noun of the word set extracted from the content.

[1-4. About narrowing down fields using graphs]
Here, as described above, when a program is automatically generated using a word set as learning data, it is considered that the novelty of the generated program depends on the word set that is the learning data. Therefore, it is desirable that the extracted word set is a word set indicating a concept belonging to a novel technical field, that is, a field in which innovation is likely to occur.

In fields where such innovation is likely to occur, that is, in fields with abundant development potential, the ease with which new technologies are likely to occur from existing technologies is new from existing technologies belonging to that field. Although it can be judged by identifying how the technology occurs, it is thought that a lot of information is needed to accurately and continuously grasp the development of such technology. In addition to the cases required for user dialogue, it can also be used for applications such as technological development as long as it can identify fields in which innovation is likely to occur.

Therefore, the publishing device 10 estimates the growth based on the rate of increase of the content belonging to the technical field, and selects the technical field in which the estimated growth potential satisfies a predetermined condition. For example, the publishing device 10 selects a technical field in which the growth potential exceeds a predetermined threshold. Then, the publishing device 10 extracts a word set included in the sentence transmitted from the content belonging to the selected technical field.

For example, the publishing device 10 uses a technical element represented by a word set belonging to a predetermined technical field as a node, and is described as graph information (hereinafter, referred to as "technical graph") showing a relationship between each element as a link between the nodes. .) Is generated. Such graph information is information indicating what kind of technical element the technical element belonging to a certain field originated from. For example, when the first technical element is a combination of the second technical element and the third technical element, or when the technology is developed from the second technical element and the third technical element, it corresponds to the first technical element. The node is connected to the node corresponding to the second technical element and the node corresponding to the third technical element via a link.

Here, the public device 10 may use any information as a technical element as long as it is information indicating a technology belonging to a certain field. For example, the publishing device 10 may have a patented invention or a patent application belonging to a certain field as a technical element, and may incorporate the relationship between each technical element into a technical graph. Further, the publishing device 10 may generate a technical graph by regarding one paper as one technical element and connecting the papers having a citation relationship with each other by a link. Further, the publishing device 10 may regard a plurality of papers having a commonality as one technical element.

In such a technical graph, the increasing tendency of the number of nodes indicates the increasing tendency of the technical elements belonging to the field, that is, the increasing tendency of the new technical elements. The increasing trend of such new technological elements can be regarded as an indicator of the likelihood of innovation in the field. Further, in such a technology graph, the increasing tendency of the number of links is considered to indicate how likely it is that another technology is generated from one technology in the field. In other words, the increasing trend in the number of links is information that can be an indicator of the potential for technological development, how likely it is that a new technology will emerge based on that technology when a new technology emerges in that field. be.

Therefore, the publishing device 10 generates a technical graph corresponding to a certain field, and specifies a development mode of the generated technical graph. More specifically, the public device 10 estimates an increasing tendency of the number of nodes and an increasing tendency of the number of links. Then, based on the estimated increasing tendency, the public device 10 has technological development in the field, such as whether or not the field corresponding to the technical graph is a field in which growth is likely to occur, and whether or not the field is a field in which the field grows. Estimate the tendency of.

For example, the public device 10 may estimate the tendency of new technology belonging to the corresponding field to occur based on the increasing tendency of the number of nodes. For example, the public device 10 is a field in which the field to be processed is likely to grow when the number of nodes increasing within a predetermined time exceeds a predetermined threshold in the technical graph corresponding to the field to be processed. You may presume that.

Further, for example, the public device 10 may estimate the tendency of new technology belonging to the corresponding field to occur based on the tendency of the number of links to increase. For example, the publishing device 10 estimates that the target field is a field in which the target field is likely to grow when the number of links increasing within a predetermined time exceeds a predetermined threshold in the technical graph corresponding to the field to be processed. You may.

Here, when the increasing tendency of the number of nodes and the increasing tendency of the number of links are similar (for example, when the number of nodes increases by one and the number of links increases by one), the technology in the field is based on the existing technology. It is estimated that it will develop in stages. Further, when the increasing tendency of the number of links exceeds the increasing tendency of the number of nodes by a predetermined threshold value (for example, when the number of nodes increases by one, the number of links tends to increase by five or more), one is one. It is presumed that this is a field where multiple new technologies are likely to occur from technology.

Therefore, the public device 10 may estimate the tendency of new technology to belong to the corresponding field based on the tendency of the number of nodes to increase and the tendency of the number of links to increase. For example, the public device 10 may estimate the tendency of new technology to belong to the corresponding field based on the tendency of increasing the number of nodes and the tendency of increasing the number of links and the difference. For example, in the public device 10, when the number of links tends to increase as compared with the tendency of the number of nodes increasing, it may be presumed that the technology belonging to the target field is easy to develop.

In this way, the public device 10 estimates the ease of technological development in the field based on the increasing tendency of the number of nodes and the increasing tendency of the number of links in the technical graph. As a result, the public device 10 can present to the user a field in which innovation is likely to occur.

[1-5. About the development of technology graphs]
Here, the public device 10 estimates a parameter indicating the development of the technical graph in the field based on the change in the number of technical elements in the past and the change in the relationship between the technical elements, and based on the estimated parameter. By changing the number of nodes and links included in the technology graph, the ease of future technology development in the target field may be predicted.

For example, the development of technology in a certain field can be regarded as a change in the number of nodes and links in a technology graph, and such a change in the number of nodes and links is virtually reproduced by a power distribution. (See, for example, Non-Patent Document 1). Therefore, even if the public device 10 estimates the tendency of change between the number of nodes and the number of links, assuming that a new node connected to the node by a link is generated from the existing node according to the power distribution. good.

The public device 10 may adopt a graph having an arbitrary configuration as an initial model of the technical graph. For example, the publishing device 10 uses the Barabashi-Albert model to generate the initial state of the technical graph, and it is assumed that a new node connected to the node by a link is generated from the node included in the generated graph information. , The tendency of change between the number of nodes and the number of links may be estimated.

For example, the publishing device 10 generates a complete graph with m nodes as an initial model of a technical graph. Subsequently, the public device 10 adds one new node and sets a link connecting the added node and another node with a predetermined probability. By repeating such a process until the number of nodes reaches a predetermined number, the publishing device 10 may generate an initial model. In addition to the above-mentioned techniques, the publishing device 10 may generate an initial model of a technical graph by adopting various methods for generating an initial model of a complex graph, such as a Watts-Strogats model. ..

[1-6. About Polya's jar algorithm]
Here, as shown in Non-Patent Document 3, a method of mathematically expressing the development of the technical field by the Polya pot algorithm can be considered. Therefore, the publishing device 10 may develop the technical graph by using the polya jar algorithm, and estimate the future development tendency in the target field based on the increasing tendency of the number of nodes and links in the technical graph. More specifically, the publishing device 10 uses a technical element indicated by a sentence belonging to a certain technical field as a node, and in a graph structure showing the relationship between each technical element as a link between the nodes, a probability based on a polya jar. Estimate the tendency of change between the number of nodes and the number of links, assuming that a new node connected to the node by a link is generated from the existing node, and the rate of increase between the number of nodes and the number of links. May select a technical field in which is greater than a predetermined threshold.

Hereinafter, an example of the Polya pot algorithm will be described. For example, in the Polya jar algorithm, one ball is randomly taken out from a jar containing balls of different colors, and it is determined whether or not the taken out ball is the color of the first taken out ball. Random. If the removed ball has the same color as the previously removed ball, the removed ball is returned to the jar and a predetermined number of balls of the same color as the ball are added to the jar. .. On the other hand, if the removed ball is the first color taken out, the removed ball is returned to the jar, a predetermined number of balls of the same color as the ball are added to the jar, and the colors are different. A predetermined number of balls, each with a new color, are placed in the jar.

Such a polya jar algorithm can be regarded as a stochastic model of the development of technological elements in a certain field. For example, if a ball placed in a jar is used as a technical element, the probability that the color of the ball taken out of the jar is the color of the ball taken out for the first time corresponds to the probability that a new technical element will occur. be able to. Also, if the ejected ball is the first extracted color, the number of new colored balls that will be newly placed in the jar will be the new number of new colored balls that will be generated from that technology when new technological elements occur. It can be associated with a parameter indicating the number of technical elements. Also, the number of balls that are put into the jar together with the balls that are taken out and that are the same color as the balls that are taken out of the jar can be associated with the relationship between the technical elements, that is, the parameter that causes the link. can.

Therefore, the public device 10 develops a technical graph by using the Polya's pot algorithm, and estimates the future development tendency in the target field. For example, when any element is selected from a set of a plurality of types of elements, the publishing device 10 is different from each other after adding an element of a type different from the selected element together with the selected element to the set. Based on the probability that an element will be selected, you may estimate the trend of change between the number of nodes and the number of links, assuming that an existing node will generate a new node connected to that node by a link. ..

[1-7. About data expansion]
Here, when the program is automatically generated by learning the neural Turing machine, it is considered that a lot of learning data is required. Therefore, the public device 10 may automatically expand the data. For example, the publishing device 10 generates a new word set from the extracted word set by combining the extracted word set and a plurality of words indicating a predetermined operation, and uses the generated word set to create a program. May be generated.

For example, the publishing device 10 may have event data in which a plurality of nouns are registered, and may newly generate a word set in which the nouns included in the word set are changed to the nouns stored in advance. Further, the publishing device 10 vectorizes the nouns included in the word set by using a technique such as word2vec, and identifies other nouns similar to the nouns included in the word set based on the comparison result of the vectors. Then, the publishing device 10 may generate a word set obtained by converting a noun included in the word set into another specified noun. Further, the publishing device 10 identifies a word corresponding to the vector of the noun and the vector in which the Euclidean distance on the vector space is equal to or more than a predetermined threshold value but the cosine distance is within the predetermined range, and the word including the specified word. By generating a set, a word set having a different field but a similar concept may be generated.

Further, for example, the publishing device 10 vectorizes each word included in the word set, and generates a vector showing the concept indicated by the word set based on the vector of each word. Then, the publishing device 10 may specify a vector similar to the generated vector and extract the word set corresponding to the specified vector as learning data.

[1-8. About an example of processing]
Hereinafter, an example of the process executed by the publishing device 10 will be described with reference to FIG. 1. First, the public device 10 collects content belonging to each technical field from the management device 100 (step S1). For example, the publication device 10 collects publications and papers belonging to each technical field. In such a case, the public device 10 has a technical element belonging to a certain field as a node and a link between the technical elements, so that the technical element belonging to the field and the relationship between the technical elements are linked for each technical field. The graph information indicating that, that is, the technical graph is generated (step S2).

For example, the publishing device 10 uses a patented invention or a dissertation belonging to a predetermined first category as a technical element, and specifies a relationship between the technical elements. Then, the publishing device 10 fits the technical elements to the nodes and generates a technical graph in which links indicating the relationships between the technical elements are set. Any graph generation technique can be adopted for the generation of such a technical graph.

Subsequently, the publishing device 10 estimates an increasing mode of future technical elements and relationships based on the Polyer's jar algorithm (step S3). For example, the publishing device 10 sets a value according to the speed at which a new link is generated with respect to the parameter of the probability that the relevance occurs in the polya jar algorithm, and is new with respect to the parameter of the probability that a new technical element is born. Set the value according to the speed at which the technical element occurs. The value set for such a parameter may be set based on, for example, a technical graph generated at a plurality of past time points in the target field and a comparison result between the generated technical graphs.

Subsequently, the public device 10 estimates the growth tendency of the target field based on the change in the number of nodes and links in each field (step S4). For example, the publishing apparatus 10 is a state of the technical graph corresponding to the first category, that is, the state N1 (T1) at the future timing T1, the state N1 (T2) at the future timing T2, and N1 at the future timing T3. (T3) is estimated. Then, the publishing device 10 estimates the growth rate # 1 of the first category based on the changes of the states N1 (T1), N1 (T2), and N1 (T3). For example, the publishing device 10 is based on changes in the number of nodes and links in state N1 (T1), the number of nodes and links in N1 (T2), and the number of nodes and links in N1 (T3). Estimate the growth rate # 1 of the first category. For example, the public device 10 may estimate the growth rate # 1 based on the average value of the increase in the number of nodes and the average value of the increase in the number of links.

Similarly, the publishing device 10 is a state of the technical graph corresponding to the second category, in the state N2 (T1) at the future timing T1, the state N2 (T2) at the future timing T2, and the future timing T3. Estimate N2 (T3). Then, the publishing device 10 estimates the growth rate # 2 of the second category based on the changes of the states N2 (T1), N2 (T2), and N2 (T3).

Then, the publishing device 10 selects a category having a high growth rate, and extracts a word set having a predetermined structure from the selected category, that is, a document belonging to a technical field (step S5). For example, as shown in FIG. 1 (A), the sentence showing the characteristics of the idea is "(noun) {to} (state expression word, adverbial phrase) {with, by, to} (verb) {. It is expected to have a structure of "can}". Therefore, the publishing device 10 uses a technique such as morphological analysis from the literature belonging to the selected technical field to "(noun) {to} (state expression word, adverbial phrase) {by, by, to} (verb). Extract sentences with the structure "{do, do}".

Then, as shown in FIG. 1B, the publishing device 10 extracts the noun included in the extracted sentence as word # 1, the state expression word or adverbial phrase as word # 2, and the verb as word # 3. Then, a word set including each of the extracted words # 1 to # 3 is generated. As shown in FIG. 1 (C), the word set generated in this way includes the feature of the idea shown by the extracted sentence, that is, the concept cε.

Then, the publishing device 10 automatically generates a program by learning the neural Turing machine using the information indicated by the words included in the extracted word set as learning data (step S6). That is, the publishing device 10 generates a program that realizes the concept cε indicated by the extracted word set. For example, the publishing device 10 learns the neural Turing machine M1 so as to output the information corresponding to the verb included in the word set when the information corresponding to the noun included in the word set is input. Automatically generate a program.

Subsequently, the publishing device 10 publishes the generated program to the user (step S7). Then, the public device 10 accepts a usage registration for using the published program from the user (step S8). That is, the public device 10 accepts user registration for the program.

A pair of such a program and a user list of users who have registered as a user for the program can be regarded as a service that can be sold. Therefore, the public device 10 discloses the set of the program and the list of registered users to the business operator as service information indicating the service contents that can be sold (step S9). As a result of such processing, the publishing device 10 can automate a series of flows from the automatic generation of a program for providing a predetermined service to the sale of the service and the sale of the company.

The public device 10 may automatically sell the service by cooperating with a predetermined payment server or the like (not shown).

[1-9. Estimating growth potential according to popularity after publication]
Here, the public device 10 may estimate the growth potential in a state where the sales price after the program is published is reflected. For example, the publishing device 10 may estimate the growth potential of the technical field related to the program based on the sales price of the published program. More specifically, the publishing device 10 may estimate the optimum value of the value indicating the growth potential by Bayesian optimization.

For example, the publishing device 10 specifies the sales value of a program automatically generated from the contents of a certain technical field, and estimates the growth potential of the technical field based on the specified sales value. For example, the public device 10 estimates the growth potential of the technical field as the higher the sales value, the higher the growth potential of the technical field. For example, the publishing device 10 estimates the optimum value of the value indicating the growth potential of the technical field from the sales value by using the Bayesian optimization technique in Bayesian estimation. That is, the publishing device 10 may specify posterior probabilities in Bayesian estimation based on the sales value of the program, and estimate a more accurate probability distribution indicating growth potential from the specified posterior probabilities.

[2. Configuration of public equipment]
Hereinafter, an example of the functional configuration of the public device 10 that realizes the above-mentioned provision processing will be described. FIG. 2 is a diagram showing a configuration example of a public device according to an embodiment. As shown in FIG. 2, the public device 10 has a communication unit 20, a storage unit 30, and a control unit 40.

The communication unit 20 is realized by, for example, a NIC (Network Interface Card) or the like. Then, the communication unit 20 is connected to the network N by wire or wirelessly, and transmits / receives information to / from the management device 100, the user terminal 300, and the business device 200.

The storage unit 30 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory (Flash Memory), or a storage device such as a hard disk or an optical disk. Further, the storage unit 30 stores a document database 31, an extended word database 32, a neural Turing machine 33, a technical graph database 34, and a service information database 35.

Information on documents as contents is registered in the document database 31. For example, FIG. 3 is a diagram showing an example of information registered in the literature database according to the embodiment. As shown in FIG. 3, information such as "literature ID (Identifier)", "text data", and "field" is registered in the document database 31 in association with each other.

Here, the "literature ID" is an identifier indicating a document. Further, "text data" is text included in a document, that is, text data. Further, the "field" is information indicating a technical field (that is, a category) to which the document indicated by the associated "literature ID" belongs. For example, in the example shown in FIG. 3, the document ID “reference # 1”, the text data “data # 1”, and the field “field # 1” are registered in association with each other in the document database 31. In such information, the text included in the document indicated by the document ID "Reference # 1" is the text data "Data # 1", and the document indicated by the document ID "Reference # 1" belongs to the field "Field # 1". Indicates that it is a document.

In the example shown in FIG. 3, conceptual values such as "Reference # 1", "Data # 1", and "Field # 1" are described, but in reality, character strings and numerical values that identify the document are used. Character strings that make up sentences, numerical values and character strings that indicate technical fields, etc. will be registered. Further, the information shown in FIG. 3 is merely an example, and arbitrary information other than the information shown in FIG. 3 may be registered in the document database 31.

Returning to FIG. 2, the explanation will be continued. In the extended word database 32, operational verbs used when expanding a word set are registered. For example, in the extended word database 32, a list of operational verbs including operational verbs of Kobark & Bakunal is registered in advance.

The neural Turing machine 33 is a model for generating a program for generating output data from input data, and is composed of, for example, a neural network such as RNN or LSTM, and a memory accessible to the neural network.

A technical graph is registered in the technical graph database 34. For example, FIG. 4 is a diagram showing an example of information registered in the technical graph database according to the embodiment. In the example shown in FIG. 4, the technical graph database 34 has information having items such as "category", "parameter", "link ID (Identifier)", "first node ID", and "second node ID". Is registered for each technology graph.

Here, the "category" is information indicating a corresponding category corresponding to the technical graph. Further, the "parameter" is a parameter indicating how the number of nodes and links increases in the corresponding technical graph. Further, the "link ID" is information for identifying a link included in the technical graph. Further, the "first node ID" and the "second node ID" identify a node connected by a link indicated by the associated "link ID", that is, a node indicating two related technical elements. Information to do.

For example, in the example shown in FIG. 4, as the information of the technical graph shown by the "technical graph # 1", the category "category # 1", the parameter "parameter # 1", the link ID "link # 1", and the first node ID " "Node # 1" and the second node ID "node # 2" are registered in association with each other. For such information, the technical graph indicated by "technical graph # 1" is a technical graph corresponding to "category # 1", and the parameter "parameter # 1" is set as the value of ρ or v described above. The link indicated by the link ID "link # 1" indicates that the node indicated by the first node ID "node # 1" and the node indicated by the second node ID "node # 2" are connected.

In the example shown in FIG. 4, conceptual values such as "category # 1," "link # 1," "node # 1", etc. are described, but in reality, character strings, numerical values, and links indicating the categories are described. Character strings and numerical values indicating nodes will be registered. Further, the information shown in FIG. 4 is merely an example, and data in any format may be registered in the technical graph database 31 as long as the data has a graph structure.

Returning to FIG. 2, the explanation will be continued. In the service information database 35, a program and a user list of users who have registered to use the program are registered in association with each other. For example, FIG. 5 is a diagram showing an example of information registered in the service information database according to the embodiment. In the example shown in FIG. 5, information having items such as "service ID", "program", and "user list" is registered in the service information database 35 in association with each other.

Here, the "service ID" is an identifier for identifying service information. The "program" is a program published by the public device 10. Further, the "user list" is a list of users who have registered the use of the associated "program", and is, for example, various user information registered at the time of the use registration.

For example, in the example shown in FIG. 5, information such as the service ID “service # 1”, the program “program # 1”, and the user list “list # 1” is registered in the service information database 35 in association with each other. Such information indicates that the program "program # 1" and the user list "list # 1" are associated with each other as the service information indicated by the service ID "service # 1".

In the example shown in FIG. 5, conceptual values such as "service # 1," "program # 1," and "list # 1" are described, but in reality, service information is stored in the service information database 35. Character strings and numerical values for identification, program codes, user list information associated with a plurality of user information, and the like are registered. In addition to the information shown in FIG. 5, arbitrary information may be registered in the service information database 35.

Returning to FIG. 2, the explanation will be continued. The control unit 40 is a controller, and for example, various programs stored in a storage device inside the public device 10 by a processor such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit) store a RAM or the like. It is realized by being executed as a work area. Further, the control unit 40 is a controller, and may be realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 2, the control unit 40 includes a collection unit 41, an estimation unit 42, a selection unit 43, an extraction unit 44, a generation unit 45, a public unit 46, an acquisition unit 47, and a provision unit 48.

The collection unit 41 collects documents belonging to each technical field as contents that are the extraction source of the word set. For example, the collection unit 41 collects documents managed by the management device 100, and registers the collected documents in the document database 31.

The estimation unit 42 estimates the growth potential for each technical field based on the rate of increase in the content belonging to the technical field. For example, the estimation unit 42 uses a technical element indicated by a sentence belonging to a certain technical field as a node, and has an existing node with a probability based on a polya jar in a graph structure showing the relationship between each technical element as a link between the nodes. From, the tendency of change between the number of nodes and the number of links is estimated assuming that a new node connected to the node by a link is generated.

For example, the estimation unit 42 generates an initial graph of a technical graph for each category. For example, the estimation unit 42 uses a paper, a patented invention, or the like belonging to a certain category as a technical element, and specifies a relationship between the technical elements. For example, the estimation unit 42 extracts a word group included in a paper or a specification of a patented invention by using various known techniques of morphological analysis and sentence analysis, and the technical element indicated by the extracted word group is synonymous with the technical element. Identify the relationship. Such identification may be realized by, for example, using various intention analysis techniques to identify a technique that is a unit of innovation. Further, the relationship between the estimated technologies may be specified based on the co-occurrence surname of the word group, and the relationship between the technical elements may be specified. That is, the estimation unit 42 uses various intention analysis techniques to identify technical elements and uses an estimation technique for estimating the relationship between the meanings of the word groups based on the similarity and co-occurrence of the word groups. Therefore, if the relationship between the technical elements is specified, any intention analysis technique or estimation technique may be adopted.

Then, the estimation unit 42 generates a technical graph in which the specified technical element is a node and the relationship between the technical elements is linked, and the generated technical graph is registered in the technical graph database 34. The estimation unit 42 may generate an initial state of graph information using the Barabashi-Albert model and register the generated graph information as a technical graph in the technical graph database 34.

Further, the estimation unit 42 sets the values of various parameters indicating the growth of the graph for each technical graph. For example, the estimation unit 42 may specify the increase amount of the node per unit time based on the change in the number of technical elements such as the actual number of papers and the number of patented inventions. Further, the estimation unit 42 may specify the amount of increase in the link per unit time based on the change in the relationship between the actual technical elements. Then, the estimation unit 42 registers various parameters indicating the growth of the graph in the technical graph database 34 based on the increase amount of these nodes and the increase amount of the links.

Subsequently, the estimation unit 42 uses the technical elements represented by the word groups belonging to the technical literature as nodes, and the number of nodes and the number of links included in the graph information showing the relationship between the elements as the links between the nodes. Estimate the trend of change in each technical field based on the trend of change in. For example, the estimation unit 42 increases the number of nodes and links of each technical graph based on the corresponding parameters for each technical graph registered in the technical graph database 31. That is, the estimation unit 42 grows the technical graph based on the parameters. Then, the estimation unit 42 identifies the amount of increase in the number of nodes and links by regarding the grown technology graph as a future technology graph, and based on the specified increase amount, determines the growth mode of the target field. presume.

For example, the estimation unit 42 may estimate whether or not the target field grows as a tendency of change in the target field. Further, the estimation unit 42 may estimate the development tendency of new technology belonging to the target field based on the increasing tendency of the number of nodes. Further, the estimation unit 42 may estimate the ease of development of the technology belonging to the target field based on the increasing tendency of the number of nodes and the increasing tendency of the number of links.

The estimation unit 42 estimates the growth rate of the target field and the mode in which the target field grows, based on the increasing tendency of the number of nodes and the increasing tendency of the number of links. You may. For example, the estimation unit 42 may estimate that the technology belonging to the target field is easy to develop when the number of links tends to increase as compared with the tendency of the number of nodes increasing.

Here, the estimation unit 42 estimates the tendency of change between the number of nodes and the number of links, assuming that a new node connected to the node by a link is generated from the existing node according to the power distribution. For example, the estimation unit 42 reads out the technical graph registered in the technical graph database 34 and the parameters corresponding to the technical graph. Then, when any element is selected from the set of a plurality of types of elements, the estimation unit 42 adds various types of elements to the set together with the selected element and then different types. Based on the probability that an element will be selected, we estimate the trend of change between the number of nodes and the number of links, assuming that an existing node will generate a new node connected to that node by a link. That is, the estimation unit 42 estimates the tendency of change between the number of nodes and the number of links based on the polya jar algorithm.

The selection unit 43 selects a technical field in which the estimated growth potential satisfies a predetermined condition. For example, the selection unit 43 selects a technical field in which the growth potential exceeds a predetermined threshold value. Further, the selection unit 43 selects a technical field in which the rate of increase between the number of nodes and the number of links exceeds a predetermined threshold value. For example, the selection unit 43 is presumed to be easily developed by the estimation unit 42, such as a technical field in which the increasing tendency of the number of nodes and the increasing tendency of the number of links exceeds a predetermined threshold value, a technical field in which the growth rate exceeds a predetermined threshold value, and the like. Select a technical field that has been selected.

The extraction unit 44 extracts a word set that is included in a sentence indicating a predetermined process and has a predetermined structure. For example, the extraction unit 44 extracts a document belonging to the technical field selected by the selection unit 43 from the document database 31, and extracts a word set having a predetermined structure from the sentences included in the extracted document. For example, the extraction unit 44 extracts, as a word set, a word set including a word that becomes a noun, a word or a word set that becomes an adverbial phrase, and a word that becomes a verb related to the word that becomes the noun.

The generation unit 45 uses the extracted word set to generate a program that realizes a predetermined process. For example, the generation unit 45 uses a neural Turing machine to output the information indicated by the verb word when the information indicated by the noun word among the words included in the extracted word set is input. Generate a program. Further, when the generation unit 45 receives the word set extracted by the extraction unit 44, the generation unit 45 generates a new word set from the extracted word set by combining the extracted word set with a plurality of words indicating a predetermined operation. Then, a program is generated using the generated word set.

For example, the generation unit 45 generates a new word set from the word set extracted by the extraction unit 44 by replacing the word included in the word set with the word registered in the extended word database 32. Then, the generation unit 45 acquires the input data and the output data based on the generated word set. For example, the generation unit 45 collects the information corresponding to the noun of the word set or the information corresponding to the noun and the adverbial phrase of the word set as input data, and also collects the information corresponding to the verb of the word set.

Then, the generation unit 45 learns the neural Turing machine 33 by using the collected information. As a result of such processing, the memory of the neural Turing machine 33 stores a program that generates output data from the input data. Therefore, the generation unit 45 acquires the program stored in the memory of the neural Turing machine 33 as the generated program.

The public unit 46 publishes the program generated by the generation unit 45. For example, the public unit 46 distributes information indicating the contents of the program generated by the generation unit 45 to the user terminal 300. Then, for example, when the user registers the use of the program, the public unit 46 distributes the program to be registered for use to the user terminal 300.

The acquisition unit 47 acquires information on users who use the published program. More specifically, the acquisition unit 47 accepts the program usage registration from the user. For example, the acquisition unit 47 acquires user information such as a user's name, age, address, etc. from the user terminal 300 when the user registration is performed. Then, the acquisition unit 47 generates service information in which the program and the user information registered for use of the program are associated with each program, and registers the generated service information in the service information database 35.

The providing unit 48 provides the service information in which the information about the program and the information of the user who uses the program are associated with each other as the information about the service that can be sold. For example, the providing unit 48 refers to the service information database 35 and generates thumbnails showing various service information. In addition, such a thumbnail may include, for example, a part of the program included in the service information, a character string indicating the content of the process realized by the program, and the like. Then, the providing unit 48 transmits the generated thumbnail to the business operator device 200 as information indicating the service that can be sold.

Further, for example, when the providing unit 48 receives a notification from the business operator device 200 to purchase any of the services, or when the payment processing related to the purchase of the service is completed by the payment server (not shown), the purchase is performed. The service information of the target service, that is, the program to be purchased and the user list of the users who use the program are transmitted to the business apparatus 200.

[3. An example of the flow of processing executed by the public appliance]
Next, an example of the flow of processing executed by the publishing device 10 will be described with reference to FIG. FIG. 6 is a flowchart illustrating an example of a flow of processing executed by the public device according to the embodiment.

First, the publishing device 10 generates a technical graph (step S101), and estimates an increasing tendency of the number of nodes and links of each technical graph based on each parameter by the polya jar algorithm (step S102). Subsequently, the publishing device 10 identifies a technical field in which the growth potential exceeds a predetermined threshold value based on the estimation result (step S103), and extracts a word group having a predetermined structure from the sentences belonging to the specified technical field. (Step S104). Then, the publishing device 10 expands the word group using a predetermined operation word (step S105), and the publishing device 10 regards the words included in the word group as input data and output data, and performs neural Turing. By learning the machine, a program is generated (step S106).

Further, the publishing device 10 publishes the generated program (step S107) and collects information on the user who uses the program (step S108). Then, the public device 10 provides a set of the user's information and the program as service information that can be sold (step S109), and ends the process.

[4. Modification example]
In the above, an example of the processing by the public device 10 has been described. However, the embodiments are not limited to this. Hereinafter, variations of the processing executed by the publishing device 10 will be described.

[4-1. Device configuration〕
In the above example, the publishing device 10 has executed the publishing process and the providing process in the publishing device 10. However, the embodiments are not limited to this. For example, the publishing device 10 may be realized by coordinating a generation server that automatically generates a program, a public server that executes a publishing process, and a providing server that executes a providing process. In such a case, the generation server has a collection unit 41, an estimation unit 42, a selection unit 43, an extraction unit 44, and a generation unit 45 shown in FIG. 2, and the public server has a public unit 46 and an acquisition unit shown in FIG. 47, and the providing server will have the providing unit 48 shown in FIG.

[4-2. others〕
Further, among the processes described in the above-described embodiment, all or a part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed can be performed. All or part of it can be done automatically by a known method. In addition, the processing procedure, specific name, and information including various data and parameters shown in the above text and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown in the figure.

Further, each component of each of the illustrated devices is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of them may be functionally or physically distributed / physically in any unit according to various loads and usage conditions. Can be integrated and configured.

In addition, the above-described embodiments can be appropriately combined as long as the processing contents do not contradict each other.

[4-3. program〕
Further, the public device 10 according to the above-described embodiment is realized by, for example, a computer 1000 having a configuration as shown in FIG. 7. FIG. 7 is a diagram showing an example of a hardware configuration. The computer 1000 is connected to the output device 1010 and the input device 1020, and the arithmetic unit 1030, the primary storage device 1040, the secondary storage device 1050, the output IF (Interface) 1060, the input IF 1070, and the network IF 1080 are connected by the bus 1090. Have.

The arithmetic unit 1030 operates based on a program stored in the primary storage device 1040 or the secondary storage device 1050, a program read from the input device 1020, or the like, and executes various processes. The primary storage device 1040 is a memory device that temporarily stores data used by the arithmetic unit 1030 for various operations such as RAM. Further, the secondary storage device 1050 is a storage device in which data used by the arithmetic unit 1030 for various calculations and various databases are registered, and is realized by a ROM (Read Only Memory), an HDD, a flash memory, or the like.

The output IF 1060 is an interface for transmitting information to be output to an output device 1010 that outputs various information such as a monitor and a printer. For example, USB (Universal Serial Bus), DVI (Digital Visual Interface), and the like. It is realized by a connector of a standard such as HDMI (registered trademark) (High Definition Multimedia Interface). Further, the input IF 1070 is an interface for receiving information from various input devices 1020 such as a mouse, a keyboard, a scanner, and the like, and is realized by, for example, USB.

The input device 1020 is, for example, an optical recording medium such as a CD (Compact Disc), a DVD (Digital Versatile Disc), a PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), or a tape. It may be a device that reads information from a medium, a magnetic recording medium, a semiconductor memory, or the like. Further, the input device 1020 may be an external storage medium such as a USB memory.

The network IF 1080 receives data from another device via the network N and sends it to the arithmetic unit 1030, and also transmits the data generated by the arithmetic unit 1030 to the other device via the network N.

The arithmetic unit 1030 controls the output device 1010 and the input device 1020 via the output IF 1060 and the input IF 1070. For example, the arithmetic unit 1030 loads a program from the input device 1020 or the secondary storage device 1050 onto the primary storage device 1040, and executes the loaded program.

For example, when the computer 1000 functions as the public device 10, the arithmetic unit 1030 of the computer 1000 realizes the function of the control unit 40 by executing the program or data loaded on the primary storage device 1040. The CPU 1100 of the computer 1000 reads these programs or data from the recording medium 1800 and executes them, but as another example, these programs may be acquired from another device via the network N.

[5. effect〕
As described above, the publishing device 10 extracts a word set having a predetermined structure, which is a word set included in a sentence indicating a predetermined process. Further, the publishing device 10 uses the extracted word set to generate a program that realizes a predetermined process. Then, the publishing device 10 publishes the generated program. As a result, the public appliance 10 can automate the process from the drafting to the provision of the service.

Further, the public device 10 acquires information on a user who uses the published program. Then, the public device 10 provides service information in which information about the program and information of a user who uses the program are associated with each other as information about a service that can be sold. Therefore, the public appliance 10 can automate the sale of the automatically generated service.

Further, the publishing device 10 estimates the growth potential for each technical field based on the increase rate of the contents belonging to the technical field, and selects the technical field in which the estimated growth potential satisfies a predetermined condition. Then, the publishing device 10 extracts a word set included in a sentence extracted from the content belonging to the selected technical field. For example, the publishing device 10 selects a technical field in which the growth potential exceeds a predetermined threshold. Further, the public device 10 has a node as a technical element indicated by a sentence belonging to a certain technical field, and has an existing node with a probability based on a polya jar in a graph structure showing a relationship between each technical element as a link between the nodes. Therefore, assuming that a new node connected by a link to that node is generated, the tendency of change between the number of nodes and the number of links is estimated, and the rate of increase between the number of nodes and the number of links is a predetermined threshold. Select a technical field that exceeds. In addition, the public device 10 estimates the growth potential of the technical field related to the program based on the sales price of the published program. That is, the public device 10 estimates the growth potential in a state of reflecting the sales price after the public release. For example, the publishing device 10 estimates the optimum value of the value indicating the growth potential by Bayesian optimization. Therefore, the publishing device 10 can automatically generate a program that realizes a more advanced concept, that is, a program that the user is interested in.

Further, the publishing device 10 extracts, as a word set, a word set including a word that becomes a noun, a word or a word set that becomes an adverbial phrase, and a word that becomes a verb related to the word that becomes a noun. Then, the publishing device 10 generates a program using a neural Turing machine so that when the information indicated by the extracted noun word is input, the information indicated by the verb word is output. Further, the publishing device 10 generates a new word set from the extracted word set by combining the extracted word set and a plurality of words indicating a predetermined operation, and uses the generated word set to create a program. Generate. Therefore, the publishing device 10 can realize automatic generation of a program that realizes the concept belonging to the field from the contents belonging to the field.

Although some of the embodiments of the present application have been described in detail with reference to the drawings, these are examples, and various modifications are made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure column of the invention. It is possible to carry out the present invention in other modified forms.

Further, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the generation unit can be read as a generation means or a generation circuit.

10 Public equipment 20 Communication unit 30 Storage unit 31 Literature database 32 Extended word database 33 Neural Turing machine 34 Technical graph database 35 Service information database 40 Control unit 41 Collection unit 42 Estimating unit 43 Selection unit 44 Extraction unit 45 Generation unit 46 Public unit 47 Acquisition department 48 Providing department 100 Management equipment 200 Business equipment 300 User terminal

Claims (12)

An extraction unit that extracts a word set having a predetermined structure and is included in a sentence indicating a predetermined process.
Using the word set extracted by the extraction unit, a generation unit that generates a program that realizes the predetermined processing, and a generation unit.
A public unit that publishes the program generated by the generator, and a public unit that publishes the program .
The acquisition unit that acquires the information of the user who uses the program released by the disclosure unit, and the acquisition unit.
With a provider that provides service information that associates information about the program with information about users who use the program as information about services that can be sold.
A public device characterized by having. An extraction unit that extracts a word set having a predetermined structure and is included in a sentence indicating a predetermined process.
Using the word set extracted by the extraction unit, a generation unit that generates a program that realizes the predetermined processing, and a generation unit.
A public unit that publishes the program generated by the generator, and a public unit that publishes the program.
For each technology field, an estimation unit that estimates growth potential based on the rate of increase in content belonging to that technology field,
It has a selection unit for selecting a technical field in which the growth potential estimated by the estimation unit satisfies a predetermined condition.
The extraction unit extracts word sets included in sentences extracted from contents belonging to the technical field selected by the selection unit.
A public device characterized by having . The publishing device according to claim 2 , wherein the selection unit selects a technical field in which the growth potential exceeds a predetermined threshold value. The estimation unit uses a technical element indicated by a sentence belonging to a certain technical field as a node, and in a graph structure showing the relationship between each technical element as a link between the nodes, from an existing node with a probability based on a polya jar. Assuming that a new node connected to the node by a link will occur, the tendency of change between the number of the node and the number of links is estimated.
The publishing device according to claim 3 , wherein the selection unit selects a technical field in which the rate of increase between the number of nodes and the number of links exceeds a predetermined threshold value. According to any one of claims 2 to 4 , the estimation unit estimates the growth potential of a technical field related to the program based on the sales value of the program published by the public unit. The listed public device. The public device according to claim 5 , wherein the estimation unit estimates the optimum value of the value indicating the growth potential by Bayesian optimization. An extraction unit that extracts a word set having a predetermined structure and is included in a sentence indicating a predetermined process.
Using the word set extracted by the extraction unit, a generation unit that generates a program that realizes the predetermined processing, and a generation unit.
With the public unit that publishes the program generated by the generator
Have,
When the generation unit inputs the information indicated by the noun word included in the word set extracted by the extraction unit, the neural Turing unit outputs the information indicated by the verb word included in the word set. A public device characterized by generating a program using a machine. An extraction unit that extracts a word set having a predetermined structure and is included in a sentence indicating a predetermined process.
Using the word set extracted by the extraction unit, a generation unit that generates a program that realizes the predetermined processing, and a generation unit.
With the public unit that publishes the program generated by the generator
Have,
The generation unit generates a new word set from the word set extracted by the extraction unit by combining the word set extracted by the extraction unit and a plurality of words indicating a predetermined operation, and the generated word set. A public device characterized in that the program is generated using the above. The extraction unit is characterized in that, as the word set, a word set including a word as a noun, a word or a word set as an adverbial phrase, and a word as a verb related to the word as the noun is extracted. Item 5. The public device according to any one of Items 1 to 8 . It is a publishing method performed by a publishing device.
An extraction step of extracting a word set having a predetermined structure, which is a word set included in a sentence indicating a predetermined process,
A generation step of generating a program that realizes the predetermined process using the word set extracted by the extraction step, and a generation step.
With the publishing process that publishes the program generated by the generation process
The acquisition process for acquiring the information of the user who uses the program released by the disclosure process, and the acquisition process.
A providing process for providing service information in which information about the program is associated with information about a user who uses the program as information about a service that can be sold.
A publishing method characterized by including. An extraction procedure for extracting a word set that is included in a sentence indicating a predetermined process and has a predetermined structure, and
A generation procedure for generating a program that realizes the predetermined process using the word set extracted by the extraction procedure, and a generation procedure.
With the publishing procedure to publish the program generated by the above generation procedure
The acquisition procedure for acquiring the information of the user who uses the program published by the publication procedure, and the acquisition procedure.
A provision procedure for providing service information in which information about the program is associated with information about a user who uses the program as information about a service that can be sold.
A public program that lets your computer run. It is a publishing method performed by a publishing device.
An extraction step of extracting a word set having a predetermined structure, which is a word set included in a sentence indicating a predetermined process,
A generation step of generating a program that realizes the predetermined process using the word set extracted by the extraction step, and a generation step.
With the publishing process that publishes the program generated by the generation process
For each technical field, an estimation process that estimates growth potential based on the growth rate of content belonging to the technical field, and
A selection process for selecting a technical field in which the growth potential estimated by the estimation process satisfies a predetermined condition
Including
The extraction step extracts word sets included in sentences extracted from contents belonging to the technical field selected by the selection step.
A publishing method characterized by that.

Images