JP6698040B2 - Generation device, generation method, and generation program - Google Patents

Description

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

  In recent years, with the rapid spread of the Internet, various contents have been actively provided via networks. In addition, with the increase in the amount of contents provided, the summary processing is also performed on the contents in order to quickly convey the contents of each content to the user.

  Here, as the technology related to content summarization, technology for outputting customer service information that meets customer requirements, technology for extracting summary sentences that reflect the meaning of words in word order, and user content for content such as broadcast programs A technique for generating a content summary sentence that reflects a preference tendency is known.

JP, 2003-196547, A JP, 2016-1889089, A JP, 2016-11813, A

  However, the above-mentioned conventional technique cannot always contribute to the purchasing activity of the user. Specifically, in the above-mentioned conventional technique, even if the contents can be summarized, it is not always possible to efficiently provide the information desired by the user regarding the purchasing activity of the user.

  The present application has been made in view of the above, and an object thereof is to provide a generation device, a generation method, and a generation program that can contribute to a purchasing activity of a user.

  The generating device according to the present application is an information indicating a characteristic of the commercial product from an acquisition unit that acquires text data associated with a commercial product that is a target of the commercial transaction on the network and the text data acquired by the acquisition unit. An extracting unit that extracts the characteristic information that is, and a generating unit that, based on the characteristic information extracted by the extracting unit, generate the content to which the advertising information that is the information for advertising the product is attached. It is characterized by having.

  According to the aspect of the embodiment, it is possible to contribute to the purchasing activity of the user.

FIG. 1 is a diagram illustrating an example of a generation process according to the embodiment. FIG. 2 is a diagram illustrating a configuration example of the generation processing system according to the embodiment. FIG. 3 is a diagram illustrating a configuration example of the generation device according to the embodiment. FIG. 4 is a diagram illustrating an example of the learning information storage unit according to the embodiment. FIG. 5 is a diagram illustrating an example of the auction information storage unit according to the embodiment. FIG. 6 is a diagram illustrating an example of the product information storage unit according to the embodiment. FIG. 7 is a flowchart (1) showing a processing procedure according to the embodiment. FIG. 8 is a flowchart (2) showing the processing procedure according to the embodiment. FIG. 9 is a flowchart (3) showing the processing procedure according to the embodiment. FIG. 10 is a hardware configuration diagram illustrating an example of a computer that realizes the function of the generation device.

  Hereinafter, modes (hereinafter, referred to as “embodiments”) for carrying out the generation device, the generation method, and the generation program according to the present application will be described in detail with reference to the drawings. Note that the generation device, the generation method, and the generation program according to the present application are not limited by this embodiment. Further, the respective embodiments can be appropriately combined within the range in which the processing content is not inconsistent. Also, in each of the following embodiments, the same parts are designated by the same reference numerals, and duplicate description will be omitted.

[1. Example of generation process]
First, an example of the generation process according to the embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of a generation process according to the embodiment. FIG. 1 illustrates an example in which the generation apparatus 100 according to the present application performs a generation process of content that introduces a product in electronic commerce (for example, a product sales page on a shopping site).

  The generation device 100 illustrated in FIG. 1 is a model (learning device) for extracting feature information, which is information indicating the feature of the commercial product, from text data associated with the commercial product that is a target of the commercial transaction on the network. Is a server device that performs the generation process using the generated model.

  Here, it is assumed that a plurality of exhibitors put up various products (products) handled by each of them on an electronic commerce site such as a shopping site exemplified in the embodiment. For example, the electronic commerce site illustrated in the embodiment may be a shopping site that imitates a shopping mall in which a plurality of sellers have their own stores, or an auction site where a plurality of sellers arbitrarily sell products. But it's okay.

  In such an electronic commerce site, a seller usually creates a title and description of a product and gives it to the product. Then, the seller uploads the image of the product and the like together with the title and description of the product. The electronic commerce site side creates a sales page for the product based on the information uploaded by the seller. A user who uses the electronic commerce site accesses the electronic commerce site and refers to the title and the description provided on the sales page to make a purchase.

  Here, in such a commercial transaction site, the creation of the information (description of the product) to be given to the product may be entrusted to the seller. For this reason, there is a problem that appropriate product information is not always provided to the user on the commercial transaction site. That is, when the seller arbitrarily adds a description, an excessively long description is added or unrelated words are inserted into the description by, for example, implementing SEO measures (Search Engine Optimization). There is a case.

  In this case, when the user accesses the commercial transaction site and browses the sales page, it may be difficult to determine what characteristics the product has. This may reduce the usability of the entire commercial transaction site. Here, at the request of the commercial transaction site, it is possible to obligate the seller to give an appropriate explanation, but for businesses that sell a large number of products, etc. Providing an appropriate explanation is a heavy work load. For this reason, the business operator may hesitate to put up an item on the commercial transaction site. In this case, the commercial transaction site side loses a large amount of sellers, which is not desirable for the operation of the commercial transaction site. Further, it is not realistic that the commercial transaction site side artificially adds a description to all the exhibited products when the number of exhibited products is huge.

  Therefore, the generation device 100 according to the embodiment extracts feature information, which is information indicating the feature of a product, from the description associated with the product by the seller by the following method. Furthermore, the generation device 100 newly generates a page that introduces a product, to which is added advertising information that is information for promoting a product, based on the extracted characteristic information. By such processing, the generation device 100 can generate a sales page with good visibility for the user without burdening the seller or the commerce site. Thereby, the generation device 100 can contribute to the purchasing activity of the user. Hereinafter, with reference to FIG. 1, an example of the generation process performed by the generation device 100 will be described along the flow.

  As illustrated in FIG. 1, the generation device 100 performs learning for performing a process of extracting feature information from a description as a learning phase. First, the generation device 100 acquires learning data to be used for learning processing (step S11).

  Specifically, the generation device 100 acquires learning data B01, B02, B03, and the like, which are existing product introduction pages uploaded to the commercial transaction site. The learning data B01 and the like are, for example, an introduction page of a product put up by the seller (for example, a sales page for accepting purchase of a product from a user), and include texts for explaining each product. For example, the learning data B01 includes the explanatory note F01. Similarly, the learning data B02 includes the explanatory note F02, and the learning data B03 includes the explanatory note F03.

  For example, the generation device 100 acquires the learning data B01 and the like by crawling on the network using a known method. The generation device 100 does not necessarily have to acquire the learning data B01 and the like from the same commercial transaction site. That is, the generation device 100 may acquire any page as long as it is a page provided with a product and a text for explaining the product.

  Subsequently, the generation device 100 performs a learning process on the acquired text (description) used for introducing the product (step S12). That is, the generation device 100 generates a predetermined model (learning device) for extracting the features included in the text (document).

  Although details will be described later, the generation device 100 is not limited to one method, and may generate a model using various methods. For example, the generation device 100 uses, together with the explanation sentence F01, a summary sentence obtained by manually summarizing the explanation sentence F01 as learning data. That is, the generation device 100 machine-learns, as the correct data, the summary sentence that summarizes the explanation sentence F01, and the words and phrases manually extracted when the explanation sentence F01 is summarized as the characteristic information of the explanation sentence F01. In this case, the generation device 100 can learn the characteristics of the words and phrases extracted during the manual summarization process by using sufficient samples for learning. Therefore, when a new explanatory sentence is input to the model, the generation device 100 can imitate a process in which a summary sentence is manually created, and extract words that are supposed to be important from the explanatory sentence. . This extracted information can be the characteristic information of the description.

  Further, the generation device 100 may generate the model by learning by statistical processing. For example, the generation device 100 performs learning for extracting a word/phrase included in a certain description and characterizing the description. Such processing is performed, for example, when the description is regarded as one document, by calculating the importance of each word in the document and generating a rule for extracting a phrase from the document based on the importance. .. For example, when a certain description (document) is input based on an index value such as tf-idf (Term Frequency-Inverse Document Frequency), the generation device 100 is a phrase that characterizes the description among the description. A model for extracting (outputting) may be generated.

  Note that the generation device 100 may generate the model using a method of deep learning, for example. That is, the generation device 100 uses a deep learning method to perform learning for extracting a part that can be a feature in the explanatory note F01 or the like. For example, the model generated by the generation device 100 is a model in which a plurality of nodes that output the calculation result of the input data are connected in multiple layers, and is a model for learning the features included in the text data. For example, the model is a neural network in which layers having a plurality of nodes are connected in multiple stages, and is a DNN (Deep Neural Network) realized by a so-called deep learning technique.

  For example, the model is generated by the following learning method by the deep learning technique. In the model, the connection coefficient between each node is initialized, and texts having various phrases are input. Then, the model is generated by a process such as backpropagation (backpropagation method) that corrects the parameter (connection coefficient) so that the error between the output and the input in the model is reduced. For example, the model may be trained so as to output the characteristic information included in the description when the text data forming the description is input. Note that the output of the model may be manually adjusted, for example, by returning appropriate feedback.

  The model learning method and the generated model are not limited to the above-described methods, and any known technique can be applied. That is, the generation device 100 can use any model as long as it can output the feature information indicating the feature abstracted from the text data. The generation device 100 stores the model generated in the learning phase in a predetermined storage unit. The above is the learning phase according to the embodiment.

  After that, the generation device 100 executes the generation phase according to the embodiment. First, when the seller attempts to put up a product or adds a description to a product, the generation device 100 determines the features in the description based on the description associated with the product. Is performed. Then, the generation device 100 generates a new product page based on the extracted characteristics.

  For example, on the commercial transaction site, an exhibitor who wants to newly sell an item uploads the item information via the exhibitor terminal 20 (step S21). The product information includes a product name of the product, price information of the product, image data of the product, and a description F10 that is a description of the product.

  Here, the generation device 100 acquires the explanatory note F10 uploaded from the seller (step S22). Then, the generation device 100 extracts feature information, which is information indicating the feature of the product, from the acquired description F10 (text data). Specifically, the generation device 100 inputs the acquired description F10 into the model and outputs the feature information in the description F10. More specifically, the generation device 100 generates the summary sentence A10 of the explanation sentence F10, which is generated by extracting only the characteristic information from the redundant explanation sentence F10 (step S23).

  Then, the generation device 100, based on the extracted characteristic information, is a page that introduces a product, and a page to which advertising information (here, summary sentence A10) that is information for advertising the product is added. To generate. Specifically, the generation device 100 generates a product page W10 which is a new page in which the summary sentence A10 is added to the original product page generated based on the information uploaded by the seller (step S10). S24).

  FIG. 1 shows a mode in which the product page W10 generated by the generation device 100 is displayed on the browser of the terminal device used by the user. As shown in FIG. 1, on the product page W10, a summary sentence A10 is displayed immediately below the title "Wine: Product XXX". That is, when browsing the product page W10, the user can understand the features of the product introduced on the product page W10 by browsing only the summary sentence A10 without reading the redundant description F10. it can.

  As described above, the generation device 100 according to the embodiment acquires the text data (description) associated with the commercial material that is the target of the commercial transaction on the network. Then, the generation device 100 extracts feature information, which is information indicating the feature of the commercial material, from the acquired text data. Furthermore, the generation device 100 generates a page that introduces a commercial material, to which is added advertising information that is information for promoting the commercial material, based on the extracted characteristic information.

  For example, the generation device 100 extracts the feature of the product from the description set by various sellers on the product introduction page or the like on the commercial transaction site, and generates advertisement information such as a summary based on the extracted feature. To do. Then, the generation device 100 newly generates a product page to which the generated advertisement information is added. Thereby, the generation device 100 can generate a product page with excellent visibility while maintaining the characteristics of the appeal points of the product given by the seller. Further, according to the generation device 100, instead of simply summarizing the explanation sentence, by learning and extracting the characteristic information of the explanation sentence, the advertisement information using words and phrases important for appealing as a product is generated. be able to. As a result, the generation device 100 can improve the usability of the commercial transaction site and contribute to the purchasing activity of the user.

  Note that the generation device 100 may use various information in order to complement the learning process. For example, in the learning phase described above, the generation device 100 may acquire result information (user reaction) such as whether or not the user has made a purchasing activity via the page generated by the generation device 100. Then, the generation device 100 may feed back the learning information to the advertising information provided to the page in which the user's reaction is good, such as a high purchase rate, as the correct answer data. Thereby, the generation device 100 can proceed with learning so as to optimize the process of extracting points appealing to the user from the description. Hereinafter, the configuration and the like of the generation device 100 that performs such processing and the generation processing system 1 including the generation device 100 will be described in detail.

[2. Generation processing system configuration]
Next, the configuration of the generation processing system 1 including the generation device 100 according to the embodiment will be described with reference to FIG. FIG. 2 is a diagram illustrating a configuration example of the generation processing system 1 according to the embodiment. As illustrated in FIG. 2, the generation processing system 1 according to the embodiment includes a user terminal 10, a seller terminal 20, and a generation device 100. These various devices are communicably connected to each other via a network N in a wired or wireless manner. Note that the generation processing system 1 illustrated in FIG. 2 may include a plurality of user terminals 10 and a plurality of exhibitor terminals 20.

  The user terminal 10 is, for example, a smart phone, a desktop PC (Personal Computer), a notebook PC, a tablet terminal, a mobile phone, a PDA (Personal Digital Assistant), an information processing device such as a wearable device (Wearable Device). Is. The user terminal 10 accesses a commercial transaction site according to an operation by the user, and performs processing for the user to browse information or purchase a product on the commercial transaction site.

  The seller terminal 20 is an information processing terminal used by a seller who sells a product on a commercial transaction site. The exhibitor terminal 20 is an information processing device such as a smartphone, a desktop PC, a notebook PC, a tablet terminal, a mobile phone, a PDA, or a wearable device.

  The seller terminal 20 executes various processes for selling a product on the commercial transaction site according to the operation by the seller. For example, the seller terminal 20 uploads the text data such as the title and description of the product to the server (the generation device 100 in the embodiment) that provides the commercial transaction site.

  The generation device 100 is a server device that extracts characteristic information from text data associated with a product and also generates a product page to which advertisement information of the product generated based on the characteristic information is added.

  Then, when the generation device 100 receives access from the user, the generation device 100 provides the generated product page to the user terminal 10. As described above, the generation device 100 according to the embodiment also serves as a web server that provides a commercial transaction site or a generated product page. In addition, in the generation processing system 1, the generation device 100 and the web server may actually be configured as separate devices.

[3. Configuration of generator]
Next, the configuration of the generation device 100 according to the embodiment will be described with reference to FIG. FIG. 3 is a diagram illustrating a configuration example of the generation device 100 according to the embodiment. As illustrated in FIG. 3, the generation device 100 includes a communication unit 110, a storage unit 120, and a control unit 130. The generation device 100 includes an input unit (for example, a keyboard and a mouse) that receives various operations from an administrator who uses the generation device 100, and a display unit (for example, a liquid crystal display) for displaying various information. You may have.

(About communication unit 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 110 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the user terminal 10 and the exhibitor terminal 20 via the network N.

(About storage unit 120)
The storage unit 120 is realized by, for example, a semiconductor memory device 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. The storage unit 120 includes a learning information storage unit 121, an auction information storage unit 122, and a product information storage unit 123.

(About learning information storage unit 121)
The learning information storage unit 121 stores information about learning for generating a model used for extraction processing. Here, FIG. 4 illustrates an example of the learning information storage unit 121 according to the embodiment. FIG. 4 is a diagram illustrating an example of the learning information storage unit 121 according to the embodiment. In the example shown in FIG. 4, the learning information storage unit 121 has items such as “learning data ID”, “category”, “product”, and “learning data”.

  “Learning data ID” indicates identification information for identifying learning data. In the following description, identification information such as the learning data ID may be indicated by a reference symbol. For example, the learning data having the learning data ID “B01” may be referred to as “learning data B01”.

  “Category” indicates the category to which the product on the commercial transaction site belongs. “Product” indicates the type of product on the commercial transaction site.

  The “learning data” indicates the contents of the learning data actually used in the learning process. For example, the learning data according to the embodiment is text data associated with a product, that is, a description provided for explaining the product.

  Although not shown in FIG. 4, the items of the learning data include not only explanatory texts used for learning but also correct answer data such as a summary text manually created based on the explanatory texts. Good. Further, when the deep learning method is used in learning, learning does not necessarily need to use only learning data for which a correct example (correct answer data) is clearly known. For example, the generation device 100 does not necessarily need the correct answer data as long as it can learn the feature of text data (document) and recognize the feature information included in the text data.

  That is, the example of the data shown in FIG. 4 indicates that the category of the product in the learning data B01 identified by the learning data ID “B01” is “food” and the product is “wine”. In addition, the learning data in the learning data B01 is “description sentence F01”.

  The learning information storage unit 121 may store the model itself learned based on the learning data. Further, the generation device 100 may appropriately update the learning data stored in the learning information storage unit 121 and continue learning the generated model.

(About the exhibition information storage unit 122)
The auction information storage unit 122 stores information on auction items (articles) uploaded from the seller terminal 20. Here, FIG. 5 shows an example of the auction information storage unit 122 according to the embodiment. FIG. 5 is a diagram showing an example of the auction information storage unit 122 according to the embodiment. In the example shown in FIG. 5, the exhibition information storage unit 122 has items such as “exhibitor ID”, “exhibition page ID”, “category”, “product”, “description”, and “image”.

  The “exhibitor ID” indicates identification information for identifying the exhibitor or the exhibitor terminal 20. The “exhibition page ID” indicates identification information for identifying the exhibition page corresponding to the product put up by the seller.

  “Category” and “product” correspond to the same item shown in FIG. The “description” is a description given to the product by the seller. The “image” indicates an image (for example, an image obtained by capturing an image of a product) associated with the auction page by the seller.

  That is, an example of the data shown in FIG. 5 indicates that the seller identified by the seller ID “20” has the exhibition page identified by the exhibition page ID “B10” on the commercial transaction site. ing. Further, it is indicated that the category of the product on the exhibition page B10 is “food” and the product is “wine”. Further, it is indicated that the explanatory note given to the exhibition page B10 is “explanatory note F10” and the image is “image P10”.

(About the product information storage unit 123)
The product information storage unit 123 stores information about the product page generated by the generation device 100 among the pages on the commercial transaction site. Here, FIG. 6 illustrates an example of the product information storage unit 123 according to the embodiment. FIG. 6 is a diagram illustrating an example of the product information storage unit 123 according to the embodiment. In the example shown in FIG. 6, the product information storage unit 123 stores items such as “product page ID”, “exhibition page ID”, “description”, “advertisement information”, “effect measurement information”, and “review information”. Have.

  “Product page ID” indicates identification information for identifying the product page generated by the generation device 100. The "exhibition page ID" and the "description" correspond to the same item shown in FIG.

  The “advertisement information” is the advertisement information generated by the generation device 100 and indicates the information added (added) to the original exhibition page. For example, the advertising information is a summary sentence or the like that summarizes the explanation sentence.

  “Effect measurement information” indicates result information indicating that the product page is effective, out of the information on the reaction of the user to the product page. The “review information” indicates information regarding a review posted by the user on the sale item page, out of information regarding the user's reaction to the sale item page. Note that, in the example shown in FIG. 6, the “effect measurement information” and the “review information” are shown by the concept of “G01” and “H01”, but in reality, various specific information is stored. It

  For example, the “effect measurement information” may store the number of users who have accessed the product page, the number of users who have reached conversion on the product page, or the rate. Examples of conversion include purchase of a product on the product page, access to a product introduction page (for example, a web page of a product manufacturer), and the like. In addition, the effect measurement information may store, for example, the number of products selected by the user (the number of clicks or touches, the number of products entered in the cart, or the like) or the rate thereof even if the product has not been purchased. .. Further, the effect measurement information may store the frequency with which the user accesses the product page, the time spent in the product page, and the like. That is, as the effect measurement information, any information may be stored as long as it is information about the product page or the reaction of the user to the product.

  Further, the “review information” stores information related to user reviews of product pages. For example, when a system in which a user evaluates a product with a numerical value of 5 levels is adopted in the product page, the review information stores the score evaluated by the user. For example, the review information stores an average value of numerical values from “0” to “5” transmitted from the user. Further, the review information may store the number and rate of user reviews transmitted from users, or comments (text data) of specific user reviews. That is, as the effect measurement information, any information may be stored as long as it is information related to a product page or a user review of the product.

  That is, the example of the data illustrated in FIG. 6 indicates that the sale item page W10 identified by the sale item page ID “W10” is originally the sale item page identified by the sale item page ID “B10”. Also, it is indicated that the explanatory note associated with the product page W10 is “explanatory note F10” and the newly added advertising information is “summary statement A10”. Further, it is indicated that the effect measurement information of the product page W10 is “G01” and the review information is “H01”.

(Regarding the control unit 130)
The control unit 130 is, for example, a controller, and includes a CPU (Central Processing Unit), an MPU (Micro Processing Unit), and the like that store various programs (an example of a generation program is stored in a storage device inside the generation apparatus 100). (Corresponding) is executed by using the RAM as a work area. The control unit 130 is, for example, a controller, and is realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

  As illustrated in FIG. 3, the control unit 130 includes an acquisition unit 131, an extraction unit 132, a reception unit 133, a generation unit 134, a storage unit 135, a reception unit 136, and a provision unit 137, It realizes or executes the functions and actions of information processing described below. The internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 3 and may be another configuration as long as it is a configuration for performing information processing described later. Further, the connection relationship between the processing units included in the control unit 130 is not limited to the connection relationship illustrated in FIG. 3 and may be another connection relationship.

(About the acquisition unit 131)
The acquisition unit 131 acquires various kinds of information. For example, the acquisition unit 131 acquires information about an auction page on a commercial transaction site on the network N. Further, the acquisition unit 131 acquires text data (description) associated with a product that is a target of a commercial transaction on the network N.

  For example, the acquisition unit 131 performs morphological analysis on the explanatory note associated with the product and acquires the word included in the explanatory note. Further, the acquisition unit 131 acquires a word indicating the attribute of the product. More specifically, the acquisition unit 131 acquires, as the word indicating the attribute of the product, for example, the size, weight, manufacturing date, release date, or manufacturer name of the product. The acquisition unit 131 may acquire not only the description but also any information as long as it is text data associated with a product such as a title.

  Note that the acquisition unit 131, when performing a morpheme analysis of a title or a description, based on the number of appearances of words (phrases) included in a document (here, the title or the description), the frequency of occurrence of words or a plurality of words. The importance of each word in the description may be calculated. For example, the acquisition unit 131 may calculate the DF value of the word related to the acquired description. Further, the acquisition unit 131 may acquire an index value such as tf-idf instead of the DF value. Then, the acquisition unit 131 may acquire only words that exceed a predetermined threshold value. Accordingly, the acquisition unit 131 can acquire only words that are supposed to characterize the product without acquiring extremely common words and phrases.

  Further, the acquisition unit 131 may acquire information regarding the reaction to the product page from the user who actually browsed the product page. For example, the acquisition unit 131 may acquire at least one of the number or rate of items selected by the user or the number or rate of conversions regarding the item as the user's response to the item page or the item. In addition, the acquisition unit 131 may acquire information regarding the review transmitted from the user. Then, the extraction unit 132 and the generation unit 134, which will be described later, may appropriately tune the generated model and the generated advertisement information based on the information on the user reaction acquired by the acquisition unit 131.

  The acquisition unit 131 may acquire information about the user who browsed the product page generated by the generation unit 134. The information about the user is, for example, attribute information of the user (sex, age, place of residence, occupation, etc.) and behavior information of the user (past browsing history on the network, purchase history, etc.). For example, the acquisition unit 131 specifies a user based on a cookie transmitted from the user terminal 10 and acquires information about the specified user.

  The acquisition unit 131 appropriately stores the acquired information in each unit of the storage unit 120. The acquisition unit 131 may appropriately acquire information used by each processing unit, which will be described later, for processing from each unit of the storage unit 120.

(About the extraction unit 132)
The extraction unit 132 extracts feature information, which is information indicating the features of the product, from the text data acquired by the acquisition unit 131.

  For example, the extraction unit 132 extracts feature information based on a learning process (for example, supervised machine learning) using correct answer data set for text data. Alternatively, the extraction unit 132 may extract the feature information using a model learned to extract the information characterizing the text data (for example, a model learned by deep learning).

  That is, the extraction unit 132 generates a model through the above-described learning process and extracts feature information using the model. The characteristic information extracted by the extraction unit 132 may be a word that characterizes the product in the text data, a combination of a plurality of words, or a sentence.

  Note that the extraction unit 132 may perform a process of adjusting the characteristic information to be extracted based on the information regarding the reaction of the user. For example, the extraction unit 132 has extracted features on a product page for which the reaction from the user has exceeded a predetermined threshold (for example, a page for which conversion has exceeded a predetermined number, or a page for which user review scores have exceeded a predetermined score). The information is determined to be correct data and the model is updated. That is, the extraction unit 132 treats a product page that has a relatively good response from the user as a page for which the characteristic information has been extracted appropriately, and treats it as correct answer data in model generation. Thereby, the extraction unit 132 can optimize the extraction process.

  The extraction unit 132 may also extract different characteristic information for each user. For example, in advertisement information given to a product, there is a possibility that there is advertisement information that is highly effective depending on the attribute of the user. For example, even if the product page is the same, the conversion rate may be high for female users and low for male users. In such a case, the extraction unit 132 separately extracts the characteristic information for female users and the characteristic information for male users in the extraction processing of the characteristic information that is the source of the advertising information. Then, the generation unit 134 described later generates a different sale item page for each user attribute information. Thereby, the generation device 100 can generate different product pages to which the advertisement information that is expected to be more effective for each user is added, even if the pages correspond to the same product. As a result, the generation device 100 can increase the purchase rate of each product and contribute to the purchasing activity of the user and the operation of the commercial transaction site.

  Further, the extraction unit 132 may adjust the characteristic information to be extracted for each category of the product or each type of the product. For example, it is assumed that the term indicating the feature in the description changes depending on the category of the product or the type of the product. As an example, if the product is "wine", it is assumed that the words that follow "Chateau" in the description show characteristics, but if the product is something other than "wine", the words that follow "Chateau" are the products. Are not expected to exhibit the characteristics of. In this way, the extraction unit 132 can extract the characteristic information suitable for the product by learning for each product category and product type.

(Regarding the reception unit 133)
The accepting unit 133 accepts information related to auction in electronic commerce from the seller terminal 20. For example, the reception unit 133 receives an image of a product that the seller is going to sell. In addition, the reception unit 133 receives text data such as a title and a description provided to the product.

  When accepting an auction, the accepting unit 133 assigns an auction page ID for each auction and stores the information in the auction information storage unit 122.

(Regarding the generation unit 134)
Based on the characteristic information extracted by the extraction unit 132, the generation unit 134 generates a product page, which is a page that introduces a product and to which advertising information that is information for promoting the product is added.

  For example, the generation unit 134 generates a product page including, as advertising information, a summary sentence of a description sentence associated with a product. In this case, the generation unit 134 may generate the product page in which the summary text is arranged prior to the original description text associated with the product. Accordingly, the generation unit 134 can generate a product page in a mode in which the user can grasp the features of the product without reading the redundant description.

  Further, the generation unit 134 may generate the product page to which the advertisement information is added, based on the reaction from the user. That is, like the extraction unit 132 described above, the generation unit 134 also generates the advertising information generated based on the user's reaction as correct answer data (or as incorrect answer data if the user's reaction is relatively poor). The processing may be learned to optimize the product page to be generated.

  As a specific example, a case where a product is converted will be described. On the product page, if the information presented as advertising information (for example, a summary) is incorrect or the characteristics of the product are not properly reflected, it is assumed that the user will hesitate to purchase the product. To be done. On the other hand, when the product is purchased (converted) from the user, it is assumed that the advertising information attached to the product page was correct.

  Therefore, the generation unit 134 estimates the propriety of the advertising information associated with the product, based on such user reaction. For example, the generation unit 134 sets a predetermined threshold value in advance. For example, the generation unit 134 sets a conversion rate threshold for a certain product. Then, the generation unit 134 determines whether or not the conversion rate exceeding the threshold value is achieved within a predetermined period or during a fixed number of browsing.

  When the conversion rate exceeds the threshold value, the generation unit 134 determines that the advertising information associated with the product is the correct answer, and the description of the product and the advertising information (or the characteristic information) are the correct answer data. As one, it feeds back to the extraction part 132 and the learning information storage part 121. In addition, when the conversion rate does not exceed the threshold value, the generation unit 134 determines that the advertisement information associated with the product is incorrect, and the description of the product and the advertisement information (or characteristic information). As one of the negative examples, feedback is provided to the extraction unit 132 and the learning information storage unit 121. In this way, by continuing the learning by utilizing the reaction of the user, the generation unit 134 can improve the estimation accuracy. As a method of obtaining such a user reaction, for example, manual determination using crowdsourcing may be performed. By learning by using human hands, the generation unit 134 can generate advertisement information that is natural when viewed by a person.

  Further, the generation unit 134 may generate a product page to which advertisement information is added, which is different for each user, based on information about the user. That is, similar to the extraction unit 132 described above, the generation unit 134 also generates advertisement information that is expected to be more effective for the user based on the user attribute information, action history, etc., and adds such advertisement information. You may make it produce|generate the said product page.

  Note that the generation unit 134 may maintain the original auction page when the extraction unit 132 does not extract the characteristic information due to insufficient learning or the like. In this case, the original exhibition page provided with the description provided by the seller is provided to the user. Then, when the learning process progresses and the characteristic information is extracted from the information on the exhibition page, the generation unit 134 may generate a product page corresponding to the exhibition page. In addition, when the generation processing is suspended as described above, the generation unit 134 may send an alert to, for example, the administrator of the generation device 100 (that is, the administrator of the commercial transaction site). In this case, the administrator of the generation device 100 can take measures such as artificially adding the advertising information.

(About the storage unit 135)
The storage unit 135 stores the product and the advertisement information corresponding to the product estimated by the generation unit 134 in the product information storage unit 123 in association with each other. That is, the storage unit 135 builds a database relating to product pages on the commercial transaction site.

(About receiving unit 136)
The receiving unit 136 receives various requests. For example, the receiving unit 136 receives a search query for searching for products from the user terminal 10 that has accessed the commercial transaction site. Then, the receiving unit 136 receives an acquisition request for the product page desired by the user from the user who browsed the search result page. That is, the receiving unit 136 receives an access request to the product page desired by the user. The receiving unit 136 sends the received various requests to the providing unit 137 and the like.

(About providing unit 137)
The providing unit 137 provides various information. For example, when the receiving unit 136 receives the product page acquisition request, the providing unit 137 provides (transmits) the product page to the user terminal 10. Further, when the receiving unit 136 receives the search query for searching the product, the providing unit 137 provides the user terminal 10 with the search result corresponding to the search query.

  That is, the providing unit 137 provides the user with a purchase opportunity regarding the product associated with the product page to which the advertisement information generated by the generating unit 134 is added by providing the product page and the search result.

[4. Processing procedure]
Next, a procedure of processing by the generation device 100 according to the embodiment will be described with reference to FIGS. 7, 8 and 9. First, the processing procedure for model generation will be described with reference to FIG. 7. FIG. 7 is a flowchart (1) showing a processing procedure according to the embodiment.

  As shown in FIG. 7, the generation device 100 acquires learning data for model generation (step S101). For example, the generation device 100 acquires an explanation sentence associated with a product as learning data.

  Then, the generation device 100 generates a model based on the acquired information (step S102). The generation device 100 stores the generated model in the storage unit 120 (step S103). With the above flow, the learning phase by the generation device 100 ends.

  Next, with reference to FIG. 8, a processing procedure regarding the generation processing using the model will be described. FIG. 8 is a flowchart (2) showing the processing procedure according to the embodiment.

  As illustrated in FIG. 8, the generation device 100 determines whether or not the exhibition page including the information on the exhibit (product) uploaded from the exhibitor terminal 20 has been received (step S201). If the auction page has not been received (step S201; No), the generation device 100 waits until it is received.

  On the other hand, when the exhibition page is accepted (step S201; Yes), the generation device 100 inputs the text data included in the accepted exhibition page into the model (step S202). Then, the generation device 100 determines whether or not the feature information is output from the model (step S203).

  When the characteristic information is output (step S203; Yes), the generation device 100 generates a product page including advertisement information based on the characteristic information (step S204). On the other hand, when the characteristic information is not output (step S203; No), the generation device 100 once suspends the generation process and maintains the original exhibition page (step S205). With the above flow, the generation phase by the generation device 100 ends.

  Next, the procedure of feedback regarding the generation process will be described with reference to FIG. FIG. 9 is a flowchart (3) showing the processing procedure according to the embodiment.

  As illustrated in FIG. 9, the generation device 100 determines whether or not a reaction to the sale item page has been received from the user terminal 10 (step S301). When the reaction has not been received (step S301; No), the generation device 100 waits until it receives the reaction.

  On the other hand, when the reaction is accepted (step S301; Yes), the generation device 100 acquires the reaction from the user (step S302). Then, the generation device 100 determines whether the reaction from the user has exceeded a predetermined threshold value (step S303).

  When the response from the user exceeds the predetermined threshold value (step S303; Yes), the generation device 100 determines that the advertising information (or the extracted characteristic information) given to the product is the correct answer ( Step S304). On the other hand, when the response from the user does not exceed the predetermined threshold value (step S303; No), the generation device 100 determines that the advertising information given to the product is incorrect (step S305).

  Then, the generation device 100 updates the model by obtaining learning data in which the data obtained in steps S304 and S305 is a positive example or a negative example (step S306). In other words, the generation device 100 more strongly learns the model based on the reaction of the user. By repeating the above flow, the generation device 100 optimizes the model.

[5. Modification example)
The generation device 100 described above may be implemented in various different forms other than the above-described embodiment. Therefore, other embodiments of the generation device 100 will be described below.

[5-1. Use for search results]
The generation device 100 may use the advertisement information for various processes as well as generating the product page using the generated advertisement information. For example, the generation device 100 may display advertisement information together with the search result when searching for a product on a commercial transaction site or the like.

  That is, the generation device 100 may generate the search result page including the advertisement information of the product when the product is displayed as the result of the predetermined search process. In this case, on the search result page, for example, advertisement information (for example, a summary) about the product is displayed near the product searched based on the search query. Therefore, when a user searches for a product, the user can grasp the feature of the product without displaying the product page. In this way, the generation device 100 can improve the convenience of the user who uses the commercial transaction site.

[5-2. Advertising information]
In the above embodiment, the generation device 100 has shown an example in which a summary sentence is generated as advertising information. Here, the advertisement information is not limited to the summary sentence, and may be any form as long as it is information for advertising the product and is information for transmitting the information characterizing the product to the user.

  For example, the generation device 100 may extract a word included in the text data as the characteristic information and generate a product page in which the extracted word is highlighted in the text data associated with the product. Good.

  As an example, the generation device 100 highlights a word extracted as a characterizing item of the original description. In this case, the advertising information means a highlight display added to the text data. Then, the generation device 100 generates a sale item page including the highlighted word. That is, the generation device 100 generates a product page in which the characteristic words included in the description are highlighted while the layout of the original exhibition page is maintained. Thereby, the generation device 100 can generate the product page to which the information desired to appeal to the user is added while maintaining the form of the page intended by the seller. That is, the content generated by the generation device 100 is not limited to new content generated from the beginning, but also various types of content such as highlighting existing content or highlighting text to improve its visibility. The contents generated by the adding process and the converting process are also included.

[5-3. Use for recommendations)
The generation device 100 displays advertisement information not only on a commercial transaction site, but also on, for example, a product recommendation displayed with a web page or the like (for example, an advertisement frame of a web page or information displayed in a recommendation frame). Good.

  For example, a product put up for sale on a commercial transaction site may be displayed as a recommendation on a product page other than the product. In this case, the generation device 100 displays the advertising information together with the recommended product. Accordingly, the user can grasp the features of the product without accessing the product page of the recommended product (without changing the display from the page currently being browsed to the product page).

  Note that the generation device 100 may acquire information on whether the recommendation of the product has been clicked or the user has reached conversion. For example, when the advertising information displayed as the recommendation is inappropriate or the appeal effect is low, it is assumed that the user does not return the reaction of clicking or converting. On the other hand, when the user returns a reaction of clicking or converting, it is assumed that the advertising information displayed together with the recommendation was information appropriately showing the features of the product (correct answer).

  Then, as described in the above embodiment, the generation device 100 determines whether the processing of the advertising information or the characteristic information is correct according to the reaction to the recommendation. Thereby, the generation device 100 can improve the accuracy of the extraction process and the generation process (perform learning) by using the reaction from the user other than the product page.

[5-4. commodity〕
In the above-described embodiment, the product such as wine handled at the shopping site is taken as an example of the product, but the product is not limited to these. In other words, the generation device 100 performs the generation process according to the embodiment on any product such as a product or a service as long as the product is a product handled at a commercial transaction site and information about the product can be acquired. It can be carried out.

[5-5. content〕
In the above embodiment, the web page is taken as an example of the content, but the content is not limited to this example. For example, the content may be advertising content, an information page provided via an application, or content related to recommendations. That is, the content generated by the generation device 100 may be any content as long as it is an information content provided via a network.

  For example, the generation device 100 may generate advertising information for a predetermined product and reflect the advertising information in the advertising content for promoting the product. Specifically, the generation device 100 generates advertisement information about a product corresponding to the advertisement content submitted by the advertiser (provider of the product). Then, the generation device 100 adds the generated advertisement information (for example, a summary sentence) to the advertisement content to generate new advertisement content. Specifically, the generation device 100 adds a summary sentence, which is advertising information, to a portion of the advertising content where text data for advertising a product is inserted. Thereby, the generation device 100 can generate, for example, the advertisement content that allows the characteristics of the product to be simply transmitted to the user.

[6. Hardware configuration]
The generation device 100, the user terminal 10, and the exhibitor terminal 20 according to the above-described embodiment are realized by, for example, a computer 1000 having a configuration illustrated in FIG. 10. Hereinafter, the generation device 100 will be described as an example. FIG. 10 is a hardware configuration diagram illustrating an example of a computer 1000 that realizes the functions of the generation device 100. The computer 1000 has a CPU 1100, a RAM 1200, a ROM 1300, an HDD 1400, a communication interface (I/F) 1500, an input/output interface (I/F) 1600, and a media interface (I/F) 1700.

  The CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400, and controls each unit. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 starts up, a program dependent on the hardware of the computer 1000, and the like.

  The HDD 1400 stores programs executed by the CPU 1100, data used by the programs, and the like. The communication interface 1500 receives data from another device via the communication network 500 (corresponding to the network N shown in FIG. 2) and sends the data to the CPU 1100, and the data generated by the CPU 1100 is transmitted via the communication network 500 to another device. Send to device.

  The CPU 1100 controls output devices such as a display and a printer and input devices such as a keyboard and a mouse via the input/output interface 1600. The CPU 1100 acquires data from the input device via the input/output interface 1600. The CPU 1100 also outputs the data generated via the input/output interface 1600 to the output device.

  The media interface 1700 reads a program or data stored in the recording medium 1800 and provides the program or data to the CPU 1100 via the RAM 1200. The CPU 1100 loads the program from the recording medium 1800 onto the RAM 1200 via the media interface 1700, and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. Etc.

  For example, when the computer 1000 functions as the generation device 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the function of the control unit 130 by executing the program loaded on the RAM 1200. Further, the HDD 1400 stores the data in the storage unit 120. The CPU 1100 of the computer 1000 reads these programs from the recording medium 1800 and executes them, but as another example, these programs may be acquired from other devices via the communication network 500.

[7. Other]
Further, of the processes described in the above embodiment, all or part of the processes described as being automatically performed may be manually performed, or the processes described as being manually performed may be performed. All or part of the process can be automatically performed by a known method. In addition, the processing procedures, specific names, information including various data and parameters shown in the above-mentioned documents and drawings can be arbitrarily changed unless otherwise specified. For example, the various kinds of information shown in each drawing are not limited to the illustrated information.

  Further, each constituent element of each illustrated device is functionally conceptual, and does not necessarily have to be physically configured as illustrated. That is, the specific form of distribution/integration of each device is not limited to that shown in the figure, and all or a part of the device may be functionally or physically distributed/arranged in arbitrary units according to various loads or usage conditions. It can be integrated and configured. For example, the acquisition unit 131 and the reception unit 133 illustrated in FIG. 3 may be integrated. Further, for example, the information stored in the storage unit 120 may be stored in a predetermined storage device provided outside via the network N.

  Further, in the above-described embodiment, an example has been shown in which the generation device 100 performs, for example, an acquisition process of acquiring text data, an extraction process of extracting characteristic information, and a generation process of generating a product page. However, the above-described generation device 100 may be separated into an acquisition device that performs an acquisition process, an extraction device that performs an extraction process, and a generation device that performs a generation process. In this case, the acquisition device has at least the acquisition unit 131. The extraction device has at least an extraction unit 132. The generation device has at least the generation unit 134. The processing performed by the generation device 100 is realized by the generation processing system 1 including the acquisition device, the extraction device, and the generation device.

  Further, the above-described embodiments and modified examples can be appropriately combined within a range in which the processing content is not inconsistent.

[8. effect〕
As described above, the generation device 100 according to the embodiment has the acquisition unit 131, the extraction unit 132, and the generation unit 134. The acquisition unit 131 acquires text data associated with a commercial material that is a target of a commercial transaction on the network. The extraction unit 132 extracts feature information, which is information indicating the feature of the commercial material, from the text data acquired by the acquisition unit 131. The generation unit 134 generates a page (an example of content) to which advertising information, which is information for advertising a commercial material, is added based on the characteristic information extracted by the extraction unit 132.

  As described above, the generation device 100 according to the embodiment extracts the features of the product from the description set by various sellers on the product introduction page of the commercial transaction site, and the summary based on the extracted features. Generate advertising information such as sentences. Thereby, the generation device 100 can generate a product page with excellent visibility while maintaining the feature of the appeal point of the product given by the seller. Therefore, the generation device 100 can improve the usability of the commercial transaction site and contribute to the purchasing activity of the user.

  The extraction unit 132 also extracts the characteristic information based on the learning process using the correct answer data set for the text data.

  As described above, the generating apparatus 100 according to the embodiment can extract feature information similar to a summary or the like manually generated by performing learning based on preset correct answer data. Therefore, the generation device 100 can accurately extract the information indicating the characteristics of the product.

  In addition, the extraction unit 132 extracts the characteristic information using a model (learning device) that has been learned to extract the information that characterizes the text data.

  As described above, the generation device 100 according to the embodiment may perform learning by a method of extracting a feature amount characterizing text data by a method such as deep learning. Through such processing, the generation device 100 can accurately and efficiently extract the characteristic information of the product.

  In addition, the acquisition unit 131 acquires a reaction of the user who has browsed the page generated by the generation unit 134 to the product associated with the page. The generation unit 134 generates a page to which advertisement information is added based on the reaction from the user.

  In this way, the generation device 100 according to the embodiment can perform the generation process by feeding back the reaction of the user. Thereby, the generation device 100 can optimize the page to be generated.

  The acquisition unit 131 also acquires information about the user who browsed the page generated by the generation unit 134. The generation unit 134 generates, based on the information about the user, a page to which the advertising information is added and which corresponds to the information about the user.

  As described above, the generation device 100 according to the embodiment may generate different pages for each user. As a result, the generation device 100 can add advertisement information that matches the characteristics of each user, and thus the appeal effect of the generated page can be improved.

  In addition, the generation unit 134 generates a page including, as advertising information, a summary sentence of the text data associated with the commercial material.

  As described above, the generation device 100 according to the embodiment may generate, as the advertising information, a summary sentence that summarizes the characteristics of the product. Accordingly, the generation device 100 can present the characteristics of the product to the user with a short sentence instead of the redundant description, and thus can generate a page that can efficiently convey information.

  Further, the generation unit 134 generates a page in which a summary sentence is arranged prior to the text data associated with the commercial material.

  As described above, the generation device 100 according to the embodiment can convey the characteristics of the product to the user faster and more accurately by generating the page that presents the summary sentence before the original explanation sentence. Thereby, the generation device 100 can improve the usability of the user who uses the commercial transaction site.

  The extraction unit 132 also extracts a word included in the text data as the characteristic information. The generation unit 134 generates a page in a mode in which the words extracted by the extraction unit 132 are highlighted in the text data associated with the commercial material.

  In this way, the generation device 100 according to the embodiment may generate a page in a mode in which only the features are highlighted in the original description. Thereby, the generation device 100 can transmit the characteristics of the product to the user without breaking the layout such as the display of the original exhibitor's description.

  Further, when the commercial material is displayed as a result of the predetermined search processing, the generation unit 134 generates a search result page including commercial information of the commercial material as an example of the content.

  As described above, the generation device 100 according to the embodiment may present the advertising information as the information displayed together with the search result. Accordingly, the generation device 100 can convey the characteristics of the product to the user without the user having to bother to access the product page.

  As described above, the embodiments of the present application have been described in detail with reference to the drawings. However, this is an example, and various modifications and improvements are made based on the knowledge of those skilled in the art, including the modes described in the section of the disclosure of the invention. The present invention can be carried out in other forms described above.

  Also, the above-mentioned "section (module, unit)" can be read as "means" or "circuit". For example, the acquisition unit can be read as an acquisition unit or an acquisition circuit.

DESCRIPTION OF SYMBOLS 1 generation processing system 10 user terminal 20 exhibitor terminal 100 generation device 110 communication unit 120 storage unit 121 learning information storage unit 122 exhibition information storage unit 123 product information storage unit 130 control unit 131 acquisition unit 132 extraction unit 133 reception unit 134 generation unit 135 Storage 136 Reception 137 Providing

Claims (9)

An acquisition unit that acquires text data associated with a product that is a target of a commercial transaction on the network,
From the text data acquired by the acquisition unit, an extraction unit that extracts characteristic information that is information indicating the characteristics of the commercial material,
A generation unit that generates content to which advertising information, which is information for advertising the product, is added, based on the characteristic information extracted by the extraction unit ,
The extraction unit,
Based on a learning process using correct answer data set for the text data, as the characteristic information, a word included in the text data is extracted,
The generator is
Of the text data associated with the commercial item, the content extracted in a mode in which the word extracted by the extraction unit is highlighted is generated.
A generating device characterized by the above. The extraction unit,
Extracting the feature information using a learner trained to extract information characterizing the text data,
The generation device according to claim 1 , wherein The acquisition unit is
Acquiring the reaction of the user who browsed the content generated by the generation unit to the product associated with the content,
The generator is
Generate content to which the advertising information has been added based on the reaction from the user,
The generating device according to claim 1 or 2 , characterized in that. The acquisition unit is
Obtaining information about a user who browsed the content generated by the generation unit,
The generator is
Generating content corresponding to the information related to the user, which is the content to which the advertisement information is added, based on the information related to the user,
The generating device according to claim 1 , wherein the generating device is a generating device. The generator is
Generating the content including a summary of text data associated with the commercial material as the advertising information,
The generation device according to claim 1 , wherein The generator is
Generating the content in a manner in which the summary is arranged prior to the text data associated with the commercial material,
The generation device according to claim 5 , wherein The generator is
When the commercial product is displayed as a result of a predetermined search process, a search result page including advertisement information of the commercial product is generated as the content.
The generating device according to claim 1 , wherein A computer-implemented generation method,
An acquisition step of acquiring text data associated with a commercial material that is a target of a commercial transaction on the network,
An extraction step of extracting characteristic information, which is information indicating the characteristics of the commercial material, from the text data acquired by the acquisition step;
Based on the characteristic information extracted by the extraction step, a generation step of generating content to which advertising information, which is information for advertising the product, is added,
The extraction step is
Based on a learning process using correct answer data set for the text data, as the characteristic information, a word included in the text data is extracted,
The generation step is
Of the text data associated with the commercial material, the content extracted in the extraction step is generated to generate the content in a mode in which the word is highlighted.
A generation method characterized by the above. An acquisition procedure for acquiring text data associated with a product that is a target of a commercial transaction on the network,
From the text data acquired by the acquisition procedure, an extraction procedure for extracting characteristic information, which is information indicating the characteristics of the commercial material,
Based on the characteristic information extracted by the extraction procedure, a generation procedure for generating a content to which advertising information, which is information for advertising the commercial material, is generated, and causing the computer to execute:
The extraction procedure is
Based on a learning process using correct answer data set for the text data, as the characteristic information, a word included in the text data is extracted,
The generation procedure is
Of the text data associated with the commercial material, the content extracted in a manner in which the word extracted by the extraction procedure is highlighted is generated.
A generation program characterized by the above.

Images