JP6679448B2 - 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.

  Conventionally, a technique for determining the evaluation of content based on various kinds of information has been provided. For example, there is provided a technique for generating a flexible electronic catalog matching the ambiguous product selection behavior of each user.

JP, 2002-133291, A

  However, in the above-mentioned conventional technique, it may be difficult to appropriately generate the model used for determining the evaluation of the content. For example, it is difficult to appropriately generate a model used for determining the evaluation of content only based on the selection behavior of each user. For example, in such a case, it may be difficult to appropriately reflect the evaluation from the viewpoint of the content provider in the content evaluation.

  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 appropriately generate a model used for determining the evaluation of content.

  The generation device according to the present application includes a positive example content that is content that is desired to be provided in electronic commerce, and a negative example content that is content that is not desired to be provided in the electronic commerce. An acquisition unit for acquiring, and a generation unit for generating a model used for evaluating the content provided in the electronic commerce based on the positive example content and the negative example content acquired by the acquisition unit. It is characterized by

  According to the aspect of the embodiment, it is possible to appropriately generate the model used for determining the evaluation of the content.

FIG. 1 is a diagram illustrating an example of a generation process according to the embodiment. FIG. 2 is a diagram illustrating an example of the determination process according to the embodiment. FIG. 3 is a diagram illustrating a configuration example of the providing device according to the embodiment. FIG. 4 is a diagram illustrating an example of the user information storage unit according to the embodiment. FIG. 5 is a diagram illustrating an example of the positive example information storage unit according to the embodiment. FIG. 6 is a diagram illustrating an example of the negative example information storage unit according to the embodiment. FIG. 7 is a diagram illustrating an example of the model information storage unit according to the embodiment. FIG. 8 is a diagram illustrating an example of the product information storage unit according to the embodiment. FIG. 9 is a flowchart showing an example of the generation process according to the embodiment. FIG. 10 is a flowchart showing an example of the determination process according to the embodiment. FIG. 11 is a diagram illustrating an example of provision of the reason why publication is not possible according to the embodiment. FIG. 12 is a flowchart illustrating an example of a process of providing the reason why publication is not possible according to the embodiment. FIG. 13 is a hardware configuration diagram illustrating an example of a computer that realizes the function of the providing apparatus.

  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. Also, in each of the following embodiments, the same parts are designated by the same reference numerals, and duplicated description will be omitted.

(Embodiment)
[1. 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 shows a case where the providing apparatus 100 generates a model based on the content labeled by a predetermined editor ED1. For example, if the predetermined editor ED1 referred to here has knowledge about guidelines for displaying information on an electronic commerce site (hereinafter also referred to as “target site”), and if an appropriate label can be given to the content, You can be any editor. For example, the predetermined editor ED1 may be a provider that provides electronic commerce. Note that the content that is labeled according to the guidelines of the target site is not limited to the content posted on the target site, as long as it can be labeled whether the guidelines of the target site are satisfied. May be. In the following, the case where the target site is the predetermined shop site A is shown as an example.

  Further, the providing device 100 uses the generated model to determine the evaluation of the product information (content) to be posted on the target site. It should be noted that the product information mentioned here means information on a predetermined product or service (hereinafter, collectively referred to as “product”). The product information may include various information such as image information (hereinafter, also simply referred to as “image”) and character information as long as it is information about a target product. Also, the providing device 100 provides a service to the user based on the evaluation of the determined content, which will be described with reference to FIG.

  As shown in FIG. 1, the providing system 1 includes a terminal device 10 (see FIG. 2), an editor device 50, and a providing device 100. The terminal device 10, the editor device 50, and the providing device 100 are communicably connected by wire or wirelessly via a predetermined communication network (not shown). The providing system 1 illustrated in FIGS. 1 and 2 may include a plurality of terminal devices 10, a plurality of editor devices 50, and a plurality of providing devices 100. For example, the providing device 100 is a generation device that generates a model used for content evaluation.

  The editor device 50 is an information processing device used by a predetermined editor ED1 that gives a label to content. The editor device 50 is realized by, for example, a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like. FIG. 1 shows a case where the editor device 50 is a notebook PC.

  For example, the predetermined editor ED1 uses the editor device 50 to give a label to each content. For example, the predetermined editor ED1 attaches a positive label to the content that meets the guidelines of the target site based on the information about the guidelines of the site of the electronic commerce, and gives a negative label to the content that does not meet the guidelines of the target site. Label the example. In the following, the content labeled with the positive example may be referred to as positive example content, and the content labeled with the negative example may be referred to as negative example content.

  In the example shown in FIG. 1, the predetermined editor ED1 gives a positive label to the content CT101 including the image IM101 and the character information CH101 (step S11-1). For example, the predetermined editor ED1 assigns a positive label to the content CT101, assuming that the image IM101, the character information CH101, and the like included in the content CT101 satisfy the guidelines of the target site.

  Note that, as described above, when described as “content CT * (* is an arbitrary numerical value)”, it indicates that the product is content identified by the content ID “CT *”. For example, when described as “content CT101”, the product is content identified by the content ID “CT101”.

  Further, the predetermined editor ED1 gives a positive label to the content CT102 including the image IM102 and the character information CH102 (step S11-2). For example, the predetermined editor ED1 assigns a positive label to the content CT102 because the image IM102, the character information CH102, and the like included in the content CT102 satisfy the guidelines of the target site.

  Further, the predetermined editor ED1 gives a negative label to the content CT201 including the image IM201 and the character information CH201 (step S12-1). Note that the hatching of the image IM201 in FIG. 1 indicates that the background color of the image IM201 is a color not recommended for publication on the target site. For example, the predetermined editor ED1 gives a negative label to the content CT201 because the background color of the image IM201 is a color that is not recommended for posting on the target site based on the guideline of the target site.

  Further, the predetermined editor ED1 gives a negative label to the content CT202 including the image IM202 and the character information CH202 (step S12-2). For example, based on the guideline of the target site, the predetermined editor ED1 recognizes that the “world's highest” character string included in the character information CH202 is a character string prohibited from being posted on the target site, so Give a negative label.

  Note that steps S11-1 and S11-2 are for explaining the process, and either step S11-1 or S11-2 may be performed first, and hereinafter, steps S11-1 and S11-. When the description will be made without distinguishing between the two, they will be collectively referred to as step S11. Further, steps S12-1 and S12-2 are for explaining the process, and either step S12-1 or S12-2 may be performed first, and hereinafter, S12-1 and S12-2. Will be collectively referred to as step S12. Further, either step S11 or step S12 may be performed first, or may be performed alternately.

  Then, the providing device 100 acquires the content to which the positive example label is attached in step S11, as shown in the positive example group PG1 (step S13-1). Specifically, the providing apparatus 100 acquires the content CT101, the content CT102, etc. to which the positive label is added in step S11.

  Further, the providing device 100 acquires the content labeled with the negative example in step S12, as shown in the negative example group NG1 (step S13-2). Specifically, the providing apparatus 100 acquires the content CT201, the content CT202, etc. to which the negative label is added in step S12. Note that steps S13-1 and S13-2 are for explaining the process, and either step S13-1 or S13-2 may be performed first, and steps S13-1 and S13-2 may be performed. It may be performed multiple times.

  Then, the providing device 100 adds the content included in the positive example group PG1 acquired in step S13-1 as the positive example content (step S14-1). Specifically, the providing device 100 adds the content CT101, the content CT102, and the like to the positive example information storage unit 122.

  Further, the providing device 100 adds the content included in the negative example group NG1 acquired in step S13-2 as the negative example content (step S14-2). Specifically, the providing device 100 adds the content CT201, the content CT202, and the like to the negative example information storage unit 123. Note that steps S14-1 and S14-2 are for explaining the process, and either step S14-1 or S14-2 may be performed first, and steps S14-1 and S14-2 may be performed. It may be performed multiple times.

  Then, the providing device 100 generates a model based on the positive example content and the negative example content (step S15). For example, the providing apparatus 100 generates a model by using the contents CT101, CT102 and the like in the positive example information storage unit 122 as positive examples and the contents CT201 and CT202 and the like in the negative example information storage unit 123 as negative examples.

  For example, the providing apparatus 100 generates a model such that when positive example content is input, a score that is equal to or greater than a predetermined threshold value is output, and when negative example content is input, a score that is less than the predetermined threshold value is output. To do. For example, the providing device 100 may generate a model that outputs a score of 0 to 1 when content is input. For example, the providing device 100 may generate a model that outputs a score of 0 to 1 as a probability value indicating whether the content is content according to a predetermined guideline when the content is input. For example, when 50 is output when content is input to the model generated by the providing apparatus 100, it indicates that the probability that the content complies with the predetermined guideline is 50%. Further, for example, when 0.8 is output when the content is input to the model generated by the providing apparatus 100, it indicates that the content has a probability of 80% in accordance with the predetermined guideline. Note that the above is an example, and the providing apparatus 100 may generate any model as long as it can be used to determine whether the input content complies with a predetermined guideline.

  In the example of FIG. 1, the providing device 100 generates a model M1 as shown in the model information storage unit 124. Further, as shown in the model information storage unit 124, the model information regarding the model M1 has a feature 1 weight of “0.5”, a feature 2 weight of “−0.4”, and a feature 3 weight of “0. 2 "and so on. For example, the providing device 100 determines (calculates) a score of 0 to 1 indicating whether the input product information is in accordance with a predetermined guideline by inputting the product information (content) to the model M1.

  As described above, the providing device 100 evaluates the content by learning using the positive example content and the negative example content labeled by the predetermined editor ED1 based on the information about the guideline of the target site of the electronic commerce. A model can be generated that is used to properly determine As described above, the providing apparatus 100 generates a model by using the information according to the guideline of the target site of the electronic commerce, thereby generating a model that appropriately determines whether or not it is adapted to the posting policy or the like on the target site. be able to. Therefore, the providing apparatus 100 can appropriately determine, for example, the evaluation of the content when displayed on the target site by using the model generated as described above.

  Note that, in the example of FIG. 1, the case where the providing apparatus 100 generates a model using the positive example content and the negative example content labeled by the predetermined editor ED1 is shown. A model may be generated using the attached positive example content and negative example content.

  For example, the providing apparatus 100 may use content selected by a user who satisfies a predetermined criterion (also referred to as “model user”) as the normal content. For example, the providing apparatus 100 may use a user whose demographic attribute or psychographic attribute satisfies a predetermined criterion as an exemplary user, and use the content selected by the exemplary user as the positive example content. For example, the providing apparatus 100 may use a user having a user attribute (for example, a woman in the twenties, etc.) that is a target of a service provided at the target site as a model user, and use the content selected by the model user as the positive example content. .

  In addition, for example, the providing apparatus 100 may use content selected by a user who does not satisfy a predetermined criterion (also referred to as “non-model user”) as negative example content. For example, the providing apparatus 100 may use a user whose demographic attribute or psychographic attribute does not satisfy a predetermined criterion as a non-model user, and use the content selected by the non-model user as the negative example content. For example, the providing apparatus 100 uses a user having a user attribute (for example, male, sixties, etc.) outside the target of the service provided at the target site as a non-model user, and uses content selected by the non-model user as negative example content. Good.

  Further, for example, the providing device 100 may use the content selected by the user who purchases the product in a predetermined electronic commerce (hereinafter, also referred to as “user with sense”) as the positive example content. For example, the providing device 100 uses a user who purchases a product in the electronic commerce transaction X that satisfies a predetermined condition regarding a purchasing user layer, sales, etc., as a sense-enabled user, and uses the content selected by the sense-enabled user as a positive example content. May be. For example, the providing apparatus 100 may use, as the positive example content, the content selected by the user corresponding to the target user attribute of the service provided at the target site among the sensed users. For example, the electronic commerce X is assumed to be an electronic commerce service different from the electronic commerce of the target site.

  Further, for example, the providing device 100 may use content selected by a user who has not purchased a product in a predetermined electronic commerce (hereinafter, also referred to as “senseless user”) as a negative example content. For example, the providing apparatus 100 uses a user who has not purchased a product in the electronic commerce X satisfying a predetermined condition regarding the purchasing user layer, sales, etc., as a senseless user, and uses the content selected by the senseless user as a negative example content. May be. For example, the providing apparatus 100 may use, as the negative example content, the content selected by the user corresponding to the user attribute that is the target of the service provided at the target site among the senseless users.

  Further, for example, the providing device 100 may use, as the positive example content, the content selected by the user who has performed the search using the predetermined keyword (hereinafter, also referred to as “high sensitivity user”). For example, the providing apparatus 100 may use a user who has performed a search using a predetermined keyword related to trends such as buzzwords and buzzwords as a high-sensitivity user, and use the content selected by the high-sensitivity user as a positive example content. For example, the providing apparatus 100 may use, as the positive example content, the content selected by the user corresponding to the target user attribute of the service provided at the target site among the high-sensitivity users.

  Further, for example, the providing device 100 may use content selected by a user who has not performed a search using a predetermined keyword (hereinafter, also referred to as “low sensitivity user”) as a negative example content. For example, the providing device 100 may use, as a low sensitivity user, a user who has not searched using a predetermined keyword related to trends such as buzzwords and buzzwords, and use the content selected by the low sensitivity user as a negative example content. . For example, the providing apparatus 100 may use, as the negative example content, the content selected by the user corresponding to the user attribute that is the target of the service provided at the target site among the low-sensitivity users.

  The labeling of the positive example content and the negative example content described above is an example, and the providing apparatus 100 generates a model using the positive example content and the negative example content labeled based on what information. Good. For example, if the content is input to the generated model, the providing apparatus 100 can output a score indicating the possibility that the content having the model input is in line with the guidelines of the target site. The model may be generated using positive example content or negative example content labeled based on such information.

[2. Decision process)
An example of the determination process according to the embodiment will be described with reference to FIG. FIG. 2 is a diagram illustrating an example of the determination process according to the embodiment. FIG. 2 shows a case where the providing apparatus 100 determines the score of the product information as the evaluation of the product information that is the content to be posted on the target site. In addition, in FIG. 2, the providing device 100 determines whether or not each product information is allowed to be posted on the target site based on the determined score of the product information, and the ranking of the product information that is allowed to be posted (hereinafter, “rank Also referred to as “”). Note that, in the example of FIG. 2, the product information about the product is shown as an example of the content, but the content is not limited to the product information and may be any content as long as it is the content to be posted on the target site.

  The terminal device 10 is an information processing device used by a user. The terminal device 10 is realized by, for example, a smartphone, a tablet terminal, a notebook PC, a desktop PC, a mobile phone, a PDA, or the like. In the example shown in FIG. 2, the terminal device 10 is a smartphone used by the user. In addition, below, the terminal device 10 may be described as a user. That is, in the following, the user can be read as the terminal device 10. Specifically, FIG. 2 illustrates a case where the terminal device 10 is a smartphone used by a user identified by the user ID “U1” (hereinafter, sometimes referred to as “user U1”).

  Further, in the example shown in FIG. 2, the terminal device 10 will be described as the terminal devices 10-1 and 10-2 according to the display of the screen of the terminal device 10. The terminal devices 10-1 and 10-2 are the same terminal device 10. Further, in the following, the terminal devices 10-1 and 10-2 will be referred to as the terminal device 10 when they are described without distinction.

  First, the providing device 100 acquires a list request from the terminal device 10-1 used by the user U1 (step S21). For example, the user U1 transmits a list request to the providing apparatus 100 by performing a predetermined operation on the page W11 of the shop site A displayed on the terminal device 10-1. Further, in the example of FIG. 2, the terminal device 10-1 transmits, to the providing device 100, information indicating that the user U1 who uses the terminal device 10 is a female in their 20s. When the providing device 100 stores the user attribute information regarding the user U1, the terminal device 10-1 transmits to the providing device 100 information indicating that the user U1 who uses the terminal device 10 is a female in their 20s. You don't have to.

  The providing device 100 that has obtained the list request from the terminal device 10 determines the evaluation of the product information that is the content using the model for calculating the evaluation of the content (step S22). In the example of FIG. 2, the providing device 100 determines the score of each item of product information stored in the product information storage unit 125 based on the model M1 stored in the model information storage unit 124.

  In the example of FIG. 2, the providing device 100 determines the score of the product information of each product GD1 to GD7 as shown in the score list SL1. Note that, as described above, when "Product GD * (* is an arbitrary numerical value)" is described, the product is identified by the product ID "GD *". For example, when "Product GD1" is described, the product is a product identified by the product ID "GD1". In the following, the product information of “product GD * (* is an arbitrary numerical value)” may be referred to as “product GD * (* is an arbitrary numerical value)”. For example, in the following, the product GD1 can be read as the product information of the product GD1.

  For example, the providing device 100 determines the score of the product GD1 to be “0.6” based on the product information of the product GD1 and the model M1. Further, for example, the providing device 100 determines the score of the product GD2 to be “0.7” based on the product information of the product GD2 and the model M1. Further, for example, the providing device 100 determines the score of the product GD3 to be “0.3” based on the product information of the product GD3 and the model M1. Further, for example, the providing device 100 determines the score of the product GD4 to be “0.5” based on the product information of the product GD4 and the model M1. Further, for example, the providing device 100 determines the score of the product GD5 to be “0.2” based on the product information of the product GD5 and the model M1. Further, for example, the providing device 100 determines the score of the product GD6 to be “0.8” based on the product information of the product GD6 and the model M1. Further, for example, the providing device 100 determines the score of the product GD7 to be “0.65” based on the product information of the product GD7 and the model M1. As described above, the providing apparatus 100 can appropriately determine the evaluation by using the model M1.

  After that, the providing device 100 excludes the product information whose score is less than a predetermined threshold value (step S23). For example, the providing device 100 narrows down the product information that is a candidate to be provided to the user U1 by excluding the product information whose score is less than “0.5”. In the example of FIG. 2, the providing device 100 excludes the products GD3 and GD5 having a score of less than “0.5” as shown in the score list SL2. Note that, as will be described later, if the provision device 100 prohibits the submission of content whose evaluation is less than a predetermined threshold when the content is submitted in advance, the process of step S23 may not be performed.

  After that, the providing device 100 determines the ranking of the product information based on the score determined in step S22 and the attribute information of the user who is the provision destination (step S24). For example, the providing device 100 determines the ranking of the product information excluding the product information whose score is less than “0.5” in step S23. For example, the providing device 100 calculates a correction score by correcting the score of each product information based on the correlation (click rate) between each product information and the user attribute and the like, and the respective products in order from the highest correction score. Determine the ranking of information. For example, the providing device 100 corrects the score of the product information in which the click rate of the user having the user attribute of the user U1 stored in the user information storage unit 121 is relatively higher than the click rates of the other user attributes. Use the score. Further, for example, the providing apparatus 100 lowers the score of the product information in which the click rate of the user having the user attribute of the user U1 stored in the user information storage unit 121 is relatively lower than the click rates of the other user attributes. Corrected score is used. In the example of FIG. 2, the providing device 100 determines that the rank of the product information of the product GD2 having the highest correction score is 1st. Further, the providing device 100 determines that the rank of the product information of the product GD4 having the second highest correction score after the product information of the product GD2 is second. Further, the providing device 100 determines that the rank of the product information of the product GD7 having the second highest correction score after the product information of the product GD4 is 3rd.

  As described above, in the example of FIG. 2, the providing device 100 determines the ranking of each product information as shown in the ranking information RK11. The determination of the above ranking is an example, and the providing apparatus 100 may determine the ranking of the product information by appropriately using various information. For example, the providing device 100 may determine the ranks in descending order from the highest score determined in step S22. In this case, the providing device 100 may determine the ranking in the order of the product GD6, the product GD2, the product GD7, the product GD1, and the product GD4. In addition, for example, when a tag of a target user attribute is attached to the product information, the providing apparatus 100 identifies the tag corresponding to the user attribute (20s, female, etc.) of the user U1 from the product information in the score list SL2. You may extract the product information provided with. Then, the providing device 100 may determine the ranking of the extracted product information in descending order of score.

  After that, the providing device 100 provides the terminal device 10 with the search result based on the ranking determined in step S24 (step S25). In the example of FIG. 2, the providing device 100 provides the ranking information RK11 to the terminal device 10-1.

  The terminal device 10 provided with the ranking information RK11 from the providing device 100 displays the ranking information RK11 (step S26). In the example of FIG. 2, the display is transited from the terminal device 10-1 to the terminal device 10-2.

  A page W12 displaying the search result is displayed on the screen of the terminal device 10-2. In the example of FIG. 2, the product information of the product GD2 and the product information of the product GD4 are displayed on the screen of the terminal device 10-2. For example, the screen of the terminal device 10-2 displays the image IM2, the title TL2, the explanatory note DC2, and the like included in the product information of the product GD2. Further, for example, the image IM4, the title TL4, the description DC4, and the like included in the product information of the product GD4 are displayed on the screen of the terminal device 10-2. The user U1 who uses the terminal device 10 touches the screen of the terminal device 10 and performs a scroll operation or the like, whereby the product information of the product GD7 and the product information of the product GD1 arranged below the product information of the product GD4. , The product information of the product GD6 and the like are sequentially displayed.

  As described above, the providing apparatus 100 appropriately determines the evaluation of each product information by the model M1 generated using the positive example content and the negative example content labeled according to the guideline of the target site of the electronic commerce. can do. As described above, the providing apparatus 100 uses the model generated in accordance with the guideline of the target site of the electronic commerce to determine whether each product information (content) is based on whether or not it is adapted to the posting policy or the like on the target site. Appropriately determine the evaluation. Therefore, the providing apparatus 100 can appropriately determine, for example, the evaluation of the content displayed on the target site by using the model M1 generated as described above. The process of step S22 may be performed in advance before the list request is acquired in step S21. For example, the providing apparatus 100 determines the evaluation of the product information, which is the content, using a model for calculating the evaluation of the content before step S21, and ranks each content based on the evaluation (score) of each content. You may decide. For example, the providing apparatus 100 may determine the ranking (rank) of each content as shown in the ranking information RK11 before step S21. In this case, when the providing apparatus 100 acquires the list request in step S21, the providing apparatus 100 may provide information based on a predetermined ranking of each content.

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

(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 by wire or wirelessly, and transmits / receives information to / from the terminal device 10.

(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. As shown in FIG. 3, the storage unit 120 according to the embodiment has a user information storage unit 121, a positive example information storage unit 122, a negative example information storage unit 123, a model information storage unit 124, and a product information storage unit. 125 and.

(User information storage unit 121)
The user information storage unit 121 according to the embodiment stores various kinds of information regarding the user. For example, the user information storage unit 121 stores various information regarding user attributes. FIG. 4 is a diagram illustrating an example of the user information storage unit according to the embodiment. The user information storage unit 121 shown in FIG. 4 includes items such as “user ID”, “age”, and “sex”.

  “User ID” indicates identification information for identifying a user. For example, the user identified by the user ID “U1” corresponds to the user U1 shown in the example of FIG. The "age" indicates the age of the user identified by the user ID. The “age” may be a specific age of the user identified by the user ID, such as 35 years old. Further, “sex” indicates the sex of the user identified by the user ID.

  For example, in the example illustrated in FIG. 4, the age of the user identified by the user ID “U1” is “20s” and the gender is “female”. Further, for example, in the example illustrated in FIG. 4, the age of the user identified by the user ID “U2” is “30s” and the gender is “male”.

  The user information storage unit 121 is not limited to the above, and may store various information according to the purpose. For example, the user information storage unit 121 may store other demographic attribute information or psychographic attribute information in addition to age and sex. For example, the user information storage unit 121 may store information such as name, home, work place, interests, family structure, income, and lifestyle.

(Regular example information storage unit 122)
The positive example information storage unit 122 according to the embodiment stores various kinds of information regarding positive examples. FIG. 5 is a diagram illustrating an example of the positive example information storage unit according to the embodiment. For example, the positive example information storage unit 122 stores the positive example content used to generate the model. The positive example information storage unit 122 illustrated in FIG. 5 includes items such as “content ID”, “image”, and “character information”.

  “Content ID” indicates identification information for identifying the content. For example, the content identified by the content ID “CT101” corresponds to the content CT101 shown in the example of FIG. "Image" indicates an image of the content. Although FIG. 5 shows an example in which conceptual information such as “IM101” is stored in “image”, in reality, image information or a file path name indicating the storage location is stored.

  The “text information” indicates text information of the content. In FIG. 5, an example in which conceptual information such as “CH101” is stored in the “character information” is shown, but in reality, various character information related to the content such as the title of the content and a description is stored.

  For example, in the example shown in FIG. 5, the content identified by the content ID “CT101” (content CT101) includes the image IM101 and the character information CH101. Further, for example, in the example shown in FIG. 5, the content identified by the content ID “CT102” (content CT102) includes the image IM102 and the character information CH102.

  Note that the positive example information storage unit 122 is not limited to the above, and may store various information according to the purpose. For example, the positive example information storage unit 122 may store information regarding the date and time when the positive example content is added. Further, for example, the positive example information storage unit 122 may store information indicating by what determination process each positive example content is added. For example, the positive example information storage unit 122 may store information indicating whether or not each positive example content is determined by the administrator's selection.

(Negative information storage unit 123)
The negative example information storage unit 123 according to the embodiment stores various information regarding negative examples. FIG. 6 is a diagram illustrating an example of the negative example information storage unit according to the embodiment. For example, the negative example information storage unit 123 stores the negative example content used to generate the model. The negative example information storage unit 123 illustrated in FIG. 6 includes items such as “content ID”, “image”, and “character information”.

  “Content ID” indicates identification information for identifying the content. For example, the content identified by the content ID “CT201” corresponds to the content CT201 shown in the example of FIG. "Image" indicates an image of the content. Although FIG. 6 shows an example in which conceptual information such as “IM201” is stored in “image”, actually, image information or a file path name indicating the storage location is stored.

  The “text information” indicates text information of the content. Although FIG. 6 shows an example in which conceptual information such as “CH201” is stored in the “character information”, in reality, various character information related to the content such as the title of the content and a description is stored.

  For example, in the example shown in FIG. 6, the content identified by the content ID “CT201” (content CT201) includes the image IM201 and the character information CH201. Further, for example, in the example illustrated in FIG. 6, the content identified by the content ID “CT202” (content CT202) includes the image IM202 and the character information CH202.

  Note that the negative example information storage unit 123 is not limited to the above, and may store various information according to the purpose. For example, the negative example information storage unit 123 may store information regarding the date and time when the negative example content is added. Further, for example, the negative example information storage unit 123 may store information indicating by what determination process each negative example content is added. For example, the negative example information storage unit 123 may store information indicating whether or not each negative example content is determined by the selection of the administrator.

(Model information storage unit 124)
The model information storage unit 124 according to the embodiment stores information about the model. For example, the model information storage unit 124 stores the model information generated by the generation process. FIG. 7 is a diagram illustrating an example of the model information storage unit according to the embodiment. The model information storage unit 124 illustrated in FIG. 7 has items such as “feature 1” to “feature 3” corresponding to each model M1 and the like as model information. Although FIG. 7 illustrates only the model M1, a plurality of model information such as the models M2 and M3 may be stored.

  For example, in the example shown in FIG. 7, the model information about the model M1 has a feature 1 weight of “0.5”, a feature 2 weight of “−0.4”, a feature 3 weight of “0.2”, and the like. Is shown. For example, when the features (features) of the model are represented by an m-dimensional vector, the number of features is m, and the weights of the features 1 to m are stored.

  The model information storage unit 124 is not limited to the above, and may store various model information according to the purpose.

(Product information storage unit 125)
The product information storage unit 125 according to the embodiment stores various kinds of information about products. FIG. 8 shows an example of the product information storage unit 125 according to the embodiment. The product information storage unit 125 illustrated in FIG. 8 has items such as “product ID”, “image”, “title”, and “description”.

  “Product ID” indicates identification information for identifying a product. “Image” indicates an image of product information that is content. Although an example in which conceptual information such as "IM1" is stored in "image" is shown in FIG. 8, image information or a file path name indicating the storage location is actually stored.

  “Title” indicates the title included in the product information of the product identified by the product ID. Further, the “description” indicates an explanation included in the product information of the product identified by the product ID. In the product information storage unit 125 shown in FIG. 8, titles are indicated by abstract symbols such as the title “TL1” and the title “TL2”. However, for example, “new bag A ...”, “Free shipping ...”, etc. As described above, each title is a specific title of the corresponding product. Further, in the product information storage unit 125 shown in FIG. 8, the explanatory note is represented by an abstract symbol such as the explanatory note “DC1” or the explanatory note “DC2”. Each description is a concrete description of the corresponding product, such as "Free shipping will be delivered ...".

  For example, in the example shown in FIG. 8, the image of the product (product GD1) identified by the product ID “GD1” is the image “IM1”. Further, the title of the product GD1 indicates that the title is “TL1”. The description of the product GD1 indicates that the description is "DC1".

  The product information storage unit 125 is not limited to the above, and may store various information according to the purpose. The product information storage unit 125 may store a specific name or the like of the product.

(Control unit 130)
Returning to the description of FIG. 3, the control unit 130 is a controller, and is stored in a storage device inside the providing apparatus 100 by, for example, a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). It is realized by executing various programs (corresponding to an example of the determination program) using the RAM as a work area. The control unit 130 is 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, a generation unit 132, a determination unit 133, and a provision unit 134, and 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.

(Acquisition unit 131)
The acquisition unit 131 acquires various kinds of information. For example, the acquisition unit 131 acquires various types of information from the user information storage unit 121, the positive example information storage unit 122, the negative example information storage unit 123, the model information storage unit 124, the product information storage unit 125, and the like. In addition, the acquisition unit 131 may acquire various information from an external information processing device.

  For example, the acquisition unit 131 acquires information on guidelines of a provider that provides electronic commerce and guidelines regarding information display on an electronic commerce site. In addition, for example, the acquisition unit 131 acquires content relating to a product or service provided in electronic commerce. For example, the acquisition unit 131 acquires information about the sense of unity of the content displayed on the site as information about the guideline. For example, the acquisition unit 131 acquires information regarding character information whose publication is restricted on the site as information regarding guidelines. In addition, for example, the acquisition unit 131 acquires information regarding images whose publication is restricted on the site as information regarding guidelines.

  Further, for example, the acquisition unit 131 includes positive example content that is content that is desired to be provided in electronic commerce, and negative example content that is content that is not desired to be provided in electronic commerce, To get. For example, the acquisition unit 131 acquires, as the normal content, the content that satisfies the guideline guidelines of the provider that provides electronic commerce. In addition, for example, the acquisition unit 131 acquires, as negative example content, content that does not satisfy the condition regarding the guideline of the provider that provides electronic commerce. For example, the acquisition unit 131 acquires positive example content and negative example content for each of a plurality of categories.

  Further, for example, the acquisition unit 131 acquires, as a positive example content, the content selected by the user who is purchasing the product in a predetermined electronic commerce transaction. Further, for example, the acquisition unit 131 acquires, as a negative example content, content selected by a user who has not purchased a product in a predetermined electronic commerce transaction.

  Further, for example, the acquisition unit 131 acquires, as a positive example content, the content selected by the user who has performed the search using the predetermined keyword. For example, the acquisition unit 131 acquires, as a negative example content, content selected by a user who has not performed a search using a predetermined keyword. Further, for example, the acquisition unit 131 acquires the submitted content submitted as content desired to be displayed on the site.

  In the example of FIG. 1, the acquisition unit 131 acquires the content to which the positive example label is given by the predetermined editor ED1 as shown in the positive example group PG1. For example, the acquisition unit 131 acquires the content CT101, the content CT102, or the like to which the positive label is attached.

  Further, in the example of FIG. 1, the acquisition unit 131 acquires the content to which the label of the negative example is given by the predetermined editor ED1 as shown in the negative example group NG1. For example, the acquisition unit 131 acquires the content CT201, the content CT202, or the like to which the negative label is attached.

  Further, for example, the acquisition unit 131 may acquire the content selected by the model user who satisfies the predetermined criterion as the canonical content. For example, the acquisition unit 131 may acquire, as a model user, a user whose demographic attribute or psychographic attribute satisfies a predetermined criterion, and acquire the content selected by the model user as the canonical content. For example, the acquisition unit 131 may acquire a user having a user attribute (for example, a woman in the twenties, etc.) that is a target of the service provided on the target site as a model user and the content selected by the model user as the normal content. Good.

  Further, for example, the acquisition unit 131 may acquire, as negative example content, content selected by a non-exemplary user who does not satisfy a predetermined criterion. For example, the acquisition unit 131 may acquire a user whose demographic attribute or psychographic attribute does not satisfy a predetermined criterion as a non-model user, and acquire the content selected by the non-model user as a negative example content. For example, the acquisition unit 131 acquires a user having a user attribute (for example, male, sixties, etc.) outside the target of the service provided at the target site as a non-model user, and acquires the content selected by the non-model user as a negative example content. May be.

  Further, for example, the acquisition unit 131 may acquire, as the normal content, the content selected by the user with a sense who purchases the product in a predetermined electronic commerce transaction. For example, the acquisition unit 131 acquires a user who purchases a product in the electronic commerce transaction X satisfying a predetermined condition regarding a purchasing user layer, sales, etc., as a sensed user, and acquires the content selected by the sensed user as a positive example content. You may. For example, the acquisition unit 131 may acquire, as the positive example content, the content selected by the user corresponding to the target user attribute of the service provided at the target site among the sensed users.

  Further, for example, the acquisition unit 131 may acquire, as the negative example content, the content selected by the senseless user who has not purchased the product in a predetermined electronic commerce transaction. For example, the acquisition unit 131 acquires, as a negative example content, a user who has not purchased a product in the electronic commerce transaction X satisfying a predetermined condition regarding a purchasing user group, sales, etc., as a senseless user, and content selected by the senseless user. You may. For example, the acquisition unit 131 may acquire, as the negative example content, the content selected by the user corresponding to the user attribute that is the target of the service provided on the target site among the senseless users.

  Further, for example, the acquisition unit 131 may acquire the content selected by the high-sensitivity user who has performed the search using the predetermined keyword, as the positive example content. For example, the acquisition unit 131 may acquire, as a high-sensitivity user, a user who has performed a search using a predetermined keyword related to trends such as buzzwords and buzzwords, and acquire the content selected by the high-sensitivity user as a positive example content. . For example, the acquisition unit 131 may acquire, as the positive example content, the content selected by the user corresponding to the target user attribute of the service provided on the target site among the high-sensitivity users.

  Further, for example, the acquisition unit 131 may acquire, as the negative example content, the content selected by the low-sensitivity user who has not performed the search using the predetermined keyword. For example, the acquisition unit 131 acquires a user who has not searched using a predetermined keyword related to a trend such as a buzzword or a buzzword as a low sensitivity user, and acquires the content selected by the low sensitivity user as a negative example content. Good. For example, the acquisition unit 131 may acquire, as a negative example content, the content selected by the user corresponding to the target user attribute of the service provided on the target site among the low-sensitivity users.

(Generator 132)
The generation unit 132 generates various information. For example, the generation unit 132 generates a model used for evaluating the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit 131. Further, for example, the generation unit 132 generates a plurality of models corresponding to each of the plurality of categories based on the positive example content and the negative example content corresponding to each of the plurality of categories acquired by the acquisition unit 131. .

  For example, the generation unit 132 generates a model that outputs a score equal to or higher than a predetermined threshold value when positive example content is input and outputs a score that is less than the predetermined threshold value when negative example content is input. To do. For example, the generation unit 132 may generate a model that outputs a score of 0 to 1 when the content is input. For example, the generation unit 132 may generate a model that outputs a score of 0 to 1 as a probability value indicating whether the content is content according to a predetermined guideline when the content is input. For example, when 50 is output when the content is input to the model generated by the generation unit 132, it indicates that the probability that the content follows the predetermined guideline is 50%. Further, for example, when 0.8 is output when the content is input to the model generated by the generation unit 132, the probability that the content follows the predetermined guideline is 80%.

  For example, the generation unit 132 generates a model based on the positive example content and the negative example content. In the example of FIG. 1, the generation unit 132 generates a model by using the contents CT101, CT102 and the like in the positive example information storage unit 122 as positive examples and the contents CT201 and CT202 and the like in the negative example information storage unit 123 as negative examples. To do. In the example of FIG. 1, the generation unit 132 generates the model M1 as shown in the model information storage unit 124. Further, for example, the generation unit 132, as shown in the model information storage unit 124, the model information about the model M1 has a feature 1 weight of “0.5”, a feature 2 weight of “−0.4”, and a feature A model M1 having a weight of 3 such as "0.2" is generated.

(Decision unit 133)
The determination unit 133 determines various information. For example, the determination unit 133 determines the content evaluation based on the guideline acquired by the acquisition unit 131. For example, when the evaluation of the content is less than a predetermined threshold value, the determination unit 133 determines that the content cannot be displayed on the electronic commerce site. Further, for example, when the evaluation of the submitted content based on the guideline is less than the predetermined threshold, the determination unit 133 determines that the submitted content cannot be submitted.

  For example, the determination unit 133 determines the evaluation of the product information that is the content using a model that calculates the evaluation of the content. In the example of FIG. 2, the determination unit 133 determines the score of each item of product information stored in the product information storage unit 125 based on the model M1 stored in the model information storage unit 124.

  For example, the determination unit 133 determines the score of the product GD1 to be “0.6” based on the product information of the product GD1 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD2 to be “0.7” based on the product information of the product GD2 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD3 to be “0.3” based on the product information of the product GD3 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD4 to be “0.5” based on the product information of the product GD4 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD5 to be “0.2” based on the product information of the product GD5 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD6 to be “0.8” based on the product information of the product GD6 and the model M1. Further, for example, the determination unit 133 determines the score of the product GD7 to be “0.65” based on the product information of the product GD7 and the model M1.

  For example, the determination unit 133 determines that the content cannot be displayed on the target site when the content evaluation is less than the predetermined threshold value. In the example of FIG. 1, the determination unit 133 determines to exclude the product information whose score is less than a predetermined threshold value. For example, the determining unit 133 narrows down the candidate product information to be provided to the user U1 by determining to exclude the product information whose score is less than “0.5”. In the example of FIG. 2, as illustrated in the score list SL2, the determination unit 133 excludes products GD3, products GD5, and the like whose scores are less than “0.5”.

  Further, for example, the determination unit 133 determines the rank of the content based on the evaluation (score) of the content. For example, the determination unit 133 determines the ranking of the product information based on the determined score and the attribute information of the user who is the provision destination. In the example of FIG. 2, the determination unit 133 determines the ranking of the product information excluding the product information whose score is less than “0.5”. For example, the determination unit 133 calculates a correction score by correcting the score based on the correlation (click rate) between each product information and the user attribute, and ranks each product information in order from the highest correction score. decide. In the example of FIG. 2, the determination unit 133 determines that the rank of the product information of the product GD2 having the highest correction score is the first place. Further, the determining unit 133 determines that the rank of the product information of the product GD4 having the second highest correction score after the product information of the product GD2 is second. In addition, the determination unit 133 determines that the rank of the product information of the product GD7 whose correction score is the second highest after the product information of the product GD4 is 3rd. In the example of FIG. 2, the determination unit 133 determines the ranking of each item of product information, as shown in the ranking information RK11.

  Further, for example, the determination unit 133 may determine the ranking of the product information by appropriately using various information. For example, the determining unit 133 may determine higher rankings in the descending order of the determined scores. In this case, in the example of FIG. 2, the determination unit 133 may determine the ranking in the order of the product GD6, the product GD2, the product GD7, the product GD1, and the product GD4. Further, in the example of FIG. 2, when the tag of the target user attribute is attached to the product information, the determination unit 133 determines the user attribute of the user U1 (20s, female, etc.) from the product information in the score list SL2. The product information to which the tag corresponding to is attached may be extracted. Then, the determining unit 133 may determine the ranking of the extracted product information in descending order of score.

(Providing section 134)
The providing unit 134 provides various information. For example, the providing unit 134 provides the terminal device 10 with various kinds of information. For example, the providing unit 134 provides the service regarding the content based on the evaluation of the content determined by the determining unit 133. For example, the providing unit 134 provides the list information of the contents displayed in the order of the contents determined by the determining unit 133. For example, the providing unit 134 provides the terminal device 10 with the search result based on the ranking determined by the determining unit 133. In the example of FIG. 2, the providing unit 134 provides the ranking information RK11 to the terminal device 10-1.

  For example, the providing unit 134 provides the provider who provides the content determined to be undisplayable by the determination unit 133 with information regarding the reason why the content is not displayable. In addition, for example, the providing unit 134 may determine that the content provider of the plurality of disapproved items does not correspond to the provider that provides the content determined to be undisplayable based on the scores of the disapproved items related to the content. Provide information about permission items. For example, the providing unit 134 applies the content of the plurality of disapproved items to the provider that provides the content determined to be undisplayable based on the model generated for each of the plurality of disallowed items related to the content. Providing information about disallowed items. This point will be described in detail in FIG.

[4. Flow of generation processing]
Next, a procedure of generation processing by the providing system 1 according to the embodiment will be described with reference to FIG. FIG. 9 is a flowchart showing an example of the generation process according to the embodiment.

  As shown in FIG. 9, the providing device 100 acquires a positive example (step S101). In the example of FIG. 1, the providing device 100 acquires the content to which the positive example label is given by the predetermined editor ED1 as shown in the positive example group PG1. For example, the providing device 100 acquires the content CT101, the content CT102, and the like to which the positive label is attached.

  Further, the providing device 100 acquires a negative example (step S102). In the example of FIG. 1, the providing apparatus 100 acquires the content labeled with the negative example label by a predetermined editor ED1 as shown in the negative example group NG1. For example, the providing device 100 acquires the content CT201, the content CT202, or the like to which the negative label is attached. Note that step S102 may be performed before step S101, or step S101 and step S102 may be performed multiple times.

  After that, the providing apparatus 100 generates a model based on the positive example and the negative example (step S103). In the example of FIG. 1, the providing apparatus 100 generates a model by using the contents CT101, CT102 and the like in the positive example information storage unit 122 as positive examples and the contents CT201 and CT202 and the like in the negative example information storage unit 123 as negative examples. To do. For example, the providing device 100 generates the model M1 based on the positive example content and the negative example content.

[5. Flow of decision processing]
Next, the procedure of the generation process by the providing system 1 according to the embodiment will be described with reference to FIG. FIG. 10 is a flowchart showing an example of the determination process according to the embodiment.

  As illustrated in FIG. 10, the providing apparatus 100 acquires a list request (step S201). In the example of FIG. 2, the providing device 100 acquires the list request from the terminal device 10-1 used by the user U1.

  Further, the providing device 100 determines the evaluation of each content item to be provided using the model (step S202). In the example of FIG. 2, the providing device 100 determines the score of each item of product information stored in the product information storage unit 125 based on the model M1 stored in the model information storage unit 124. In the example of FIG. 2, the providing device 100 determines the score of the product information of each product GD1 to GD7 as shown in the score list SL1.

  In addition, the providing device 100 excludes contents whose evaluation is less than the threshold value (step S203). In the example of FIG. 2, the providing device 100 excludes the products GD3 and GD5 having a score of less than “0.5” as shown in the score list SL2. As described above, when the providing apparatus 100 prohibits the submission of the content whose evaluation is less than the predetermined threshold when the content is submitted in advance, the providing apparatus 100 does not perform the process of step S203. May be.

  Then, the providing device 100 determines the ranking of each content based on the evaluation and the user information of the providing destination (step S204). In the example of FIG. 2, the providing device 100 determines that the rank of the product information of the product GD2 having the highest correction score is 1st. Further, the providing device 100 determines that the rank of the product information of the product GD4 having the second highest correction score after the product information of the product GD2 is second. Further, the providing device 100 determines that the rank of the product information of the product GD7 having the second highest correction score after the product information of the product GD4 is 3rd.

  Then, the providing device 100 provides the ranking information (step S205). In the example of FIG. 2, the providing device 100 provides the ranking information RK11 to the terminal device 10-1.

[6. Provision of reasons why publication is not possible
In the above-mentioned example, the example in which the providing apparatus 100 uses the model to determine the ranking when providing the product information to the user has been shown, but the providing apparatus 100 may provide various services. For example, the providing device 100 may acquire the content from the provider of the content desired to be posted on the target site, and preliminarily examine whether the content can be posted. In this case, for example, when it is determined that the content cannot be posted on the target site, the providing apparatus 100 may feed back the reason why the content cannot be posted on the target site to the content provider. This point will be described with reference to FIG.

  FIG. 11 is a diagram illustrating an example of provision of the reason why publication is not possible according to the embodiment. Specifically, in FIG. 11, the providing apparatus 100 acquires the content CT301 from the content creator CM1 who created the content CT301, and determines whether or not the content CT301 can be posted on the target site (for example, the shop site A). First, the model used in FIG. 11 will be described.

  In the example illustrated in FIG. 11, the providing device 100 determines whether or not to publish or not to publish by using the plurality of models M1 to M5 stored in the model information storage unit 124A. The model information storage unit 124A shown in FIG. 11 is the same as the model information storage unit 124 in FIG. 1 except that a plurality of models M1 to M5 are stored. The model M1 shown in the model information storage unit 124A in FIG. 11 is the same as the model M1 shown in the model information storage unit 124 in FIG.

  In the example of FIG. 11, as shown in the model information storage unit 124A, the model information regarding the model M2 has a feature 1 weight of “−0.3”, a feature 2 weight of “0.1”, and a feature 3 weight. Indicates “−0.2” or the like. Further, in the example of FIG. 11, the model M2 is a model used to determine whether the background of the image complies with the conditions regarding the background of the image in the guideline of the target site.

  For example, the providing apparatus 100 may acquire content (images) satisfying the condition related to the background of the image as a positive example content from the guidelines of the target site. Further, for example, the providing apparatus 100 may acquire, as the negative example content, content (image) that does not satisfy the condition regarding the background of the image among the guidelines of the target site. Further, the providing device 100 may store the positive example content related to the background of the image in the positive example information storage unit 122.

  For example, the providing apparatus 100 may generate a model used for determining whether or not a condition regarding the background of the image is satisfied among the guidelines of the target site by using the positive example content and the negative example content regarding the background of the image. In the example of FIG. 11, the providing apparatus 100 uses the positive example content and the negative example content regarding the background of the image to generate the model M2 regarding the background of the image.

  For example, when the content is input, the providing apparatus 100 outputs a score of 0 to 1 as a probability value indicating whether the content is content that meets the conditions regarding the background of the image among the guidelines of the target site. A model M2 to be generated is generated. For example, the providing apparatus 100 inputs an image (content) into the model M2 to determine a score of 0 to 1 indicating whether the input image complies with the condition regarding the background of the image among the guidelines of the target site ( calculate.

  In the example of FIG. 11, as shown in the model information storage unit 124A, in the model information about the model M4, the weight of the feature 1 is “0”, the weight of the feature 2 is “−1.4”, and the weight of the feature 3 is “. -2.2 "and so on. Further, in the example of FIG. 11, the model M4 is a model used for determining whether the character information is in accordance with the prohibited character condition in the guideline of the target site.

  For example, the providing apparatus 100 may acquire content (character information) satisfying the condition regarding prohibited characters as a positive example content from the guidelines of the target site. Further, for example, the providing apparatus 100 may acquire, as the negative example content, content (character information) that does not satisfy the condition regarding the prohibited character in the guideline of the target site. Further, the providing device 100 may store the canonical content regarding the prohibited characters in the canonical information storage unit 122.

  For example, the providing apparatus 100 may use a positive example content or a negative example content regarding prohibited characters to generate a model used for determining whether or not a condition regarding the prohibited characters is satisfied in the guidelines of the target site. In the example of FIG. 11, the providing apparatus 100 uses the positive example content and the negative example content regarding the prohibited character to generate the model M4 regarding the prohibited character.

  For example, when the content is input, the providing apparatus 100 outputs a score of 0 to 1 as a probability value indicating whether the content is the content according to the condition regarding the prohibited character in the guideline of the target site. Generate a model M4. For example, the providing apparatus 100 inputs the character information (content) into the model M4 to determine a score of 0 to 1 indicating whether the input character information complies with the prohibited character condition in the guideline of the target site. (calculate.

  In this way, the providing apparatus 100 may generate a plurality of models corresponding to each of the plurality of categories based on the positive example content and the negative example content for each of the plurality of categories. Further, the category is not limited to the content availability reason, and may be the category of the content itself. For example, the providing device 100 may generate a model corresponding to each of various contents such as product information and news articles.

  From here, the flow of the content availability determination of the content CT301 and the notification of the reason will be described with reference to FIG.

  First, the content creator CM1 uses the creation device 51 to submit the content CT301 including the image IM301, the character information CH301, and the like to the providing device 100 (step S31). For example, the content creator CM1 uses the creating device 51 to create the content CT301 including the image IM301 and the character information CH301, and submits the created content CT301 to the providing device 100. Thereby, the providing apparatus 100 acquires the content CT301 from the creating apparatus 51 used by the content creator CM1.

  The providing device 100 that has acquired the content CT301 determines whether or not the content CT301 can be posted on the target site. For example, the providing apparatus 100 uses the model M1 to determine whether or not the content CT301 can be posted on the target site. In the example of FIG. 11, the providing apparatus 100 determines whether or not the content CT301 can be posted on the target site by performing the processing shown in the processing group PS31. Note that in the example of FIG. 11, the providing apparatus 100 determines that the content input to the model M1 cannot be posted on the target site when the score output by the model M1 is less than 0.5.

  For example, the providing apparatus 100 inputs the content CT301 into the model M1 (step S32). The model M1 to which the content CT301 is input outputs a score (step S33). In the example of FIG. 11, the model M1 to which the content CT301 is input outputs the score “0.35” as shown in the score information SC31.

  After that, the providing device 100 determines whether or not the content input to the model M1 can be posted on the target site based on the score output by the model M1 (step S34). In the example of FIG. 11, as shown in the determination information DT31, the score of the content CT301 is “0.35” and is less than the threshold value “0.5”, so it is determined that the content CT301 cannot be posted on the target site.

  Therefore, the providing apparatus 100 uses other models M2 to M5 to determine the reason why the content CT301 cannot be posted. If the score output by the model M1 is equal to or greater than the threshold value “0.5”, the providing device 100 provides the creating device 51 with information indicating that the content input to the model M1 can be posted on the target site. Then, the processing may be ended.

  For example, the providing apparatus 100 uses the model M2 to determine whether or not the guideline of the target site satisfies the condition regarding the background of the image. In the example of FIG. 11, the providing device 100 determines whether or not the image IM301 of the content CT301 satisfies the condition related to the background of the image in the guideline of the target site by the process shown in the process group PS32. In the example of FIG. 11, when the score output by the model M2 is less than 0.3, the providing apparatus 100 determines that the content (image) input to the model M2 is related to the background of the image among the guidelines of the target site. It is determined that the conditions are not met.

  For example, the providing apparatus 100 inputs the image IM301 into the model M2 (step S35-1). The model M2 to which the image IM301 is input outputs a score (step S36-1). In the example of FIG. 11, the model M2 to which the image IM301 is input outputs the score “0.12” as shown in the score information SC32.

  After that, the providing apparatus 100 determines whether the content (image) input to the model M2 has an unreasonable reason regarding the background of the image in the guideline of the target site based on the score output by the model M2 (step S37). -1). In the example of FIG. 11, as shown in the determination information DT32, the score of the image IM301 is “0.12” and less than the threshold value “0.3”, so the image IM301 is the background of the image of the guidelines of the target site. It is determined that there is an unreasonable reason for.

  Further, for example, the providing apparatus 100 uses the model M3 to determine whether or not the guideline of the target site satisfies the condition regarding the exposure of the skin in the image. For example, it is assumed that the guidelines for the target site include a condition that publication is not possible when humans included in the image expose a predetermined proportion or more of the skin.

  In the example of FIG. 11, the providing apparatus 100 determines whether or not the image IM301 of the content CT301 satisfies the condition regarding the exposure of the skin in the image in the guideline of the target site by the processing as shown in the processing group PS33. In the example of FIG. 11, when the score output by the model M3 is less than 0.4, the providing apparatus 100 determines that the content (image) input to the model M3 is the skin of the image in the guideline of the target site. It is determined that the conditions regarding exposure are not satisfied.

  For example, the providing apparatus 100 inputs the image IM301 into the model M3 (step S35-2). The model M3 to which the image IM301 is input outputs a score (step S36-2). In the example of FIG. 11, the model M3 to which the image IM301 is input outputs the score “0.73” as shown in the score information SC33.

  After that, the providing apparatus 100 determines whether the content (image) input to the model M3 has an unreasonable reason for skin exposure in the image among the guidelines of the target site based on the score output by the model M3 ( Step S37-2). In the example of FIG. 11, since the score of the image IM301 is “0.73” and the threshold value is “0.4” or more as shown in the determination information DT33, the image IM301 is the skin of the image in the guideline of the target site. It is determined that there is no unreasonable reason for the exposure.

  Further, the providing device 100 determines the presence or absence of each reason by performing the above-described processing for each reason using the models M3 to M5 and the like.

  After that, the providing apparatus 100 provides the content creator CM1 with the reason why the content CT301 cannot be posted (step S38). The providing device 100 transmits the information indicating that the content CT301 cannot be posted to the background of the image or the use of the prohibited characters to the creating device 51, thereby providing the reason why the content CT301 cannot be posted. Provide to the content creator CM1.

  As a result, the content creator CM1 can change the content CT301 based on the provided impossibility of publication, and thus can create content in accordance with the guidelines of the target site. In this way, the providing device 100 can promote the creation of content in accordance with the guidelines of the target site.

[7. Flow of decision processing]
Next, the procedure of the determination process by the providing system 1 according to the embodiment will be described with reference to FIG. FIG. 12 is a flowchart illustrating an example of a process of providing the reason why publication is not possible according to the embodiment.

  As shown in FIG. 12, the providing device 100 acquires the content (step S301). In the example of FIG. 11, the providing apparatus 100 acquires the content CT301 from the creating apparatus 51 used by the content creator CM1.

  Then, the providing device 100 determines whether or not the content can be published (step S302). If the content can be posted (step S302: YES), the providing apparatus 100 ends the process. In the example of FIG. 11, the providing apparatus 100 determines whether or not the content input to the model M1 can be posted on the target site based on the score output by the model M1. Further, in the example of FIG. 11, as shown in the determination information DT31, the score of the content CT301 is “0.35” and less than the threshold value “0.5”, so it is determined that the content CT301 cannot be published on the target site. .

  On the other hand, if the content is not available for publication (step S302: NO), the providing apparatus 100 uses the model for each reason to determine the evaluation for each reason (step S303). For example, the providing apparatus 100 determines the score regarding the background of the image in the guideline of the target site by inputting the image IM301 to the model M2. Further, for example, the providing apparatus 100 determines the score regarding the exposure of the skin in the image in the guideline of the target site by inputting the image IM301 to the model M3.

  Then, the providing apparatus 100 determines the reason why the evaluation is less than the threshold as the reason why the content cannot be published (step S304). In the example of FIG. 11, the providing apparatus 100 determines that the score is less than the threshold value and the use of the background of the image or the prohibited characters is the reason why the content CT301 cannot be posted.

  After that, the providing device 100 provides the reason why publication is not possible (step S305). In the example of FIG. 11, the providing apparatus 100 transmits the content CT301 by transmitting information indicating that the reason why the content CT301 is not allowed to be displayed is the background of the image or the use of prohibited characters to the content CT301. The content creator CM1 is provided with the reason why the content is unavailable.

[8. effect〕
As described above, the providing device 100 according to the embodiment has the acquisition unit 131 and the generation unit 132. The acquisition unit 131 acquires positive example content, which is content that is desired to be provided in electronic commerce, and negative example content, which is content that is preferably not provided in electronic commerce. Further, the generation unit 132 generates a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit 131.

  As a result, the providing apparatus 100 according to the embodiment has a positive example content that is desirable to be provided in electronic commerce and a negative example content that is not desirable to be provided in electronic commerce. Based on the content and the content, a model used for determining the evaluation of the content can be appropriately generated.

  In addition, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as a positive example content, a content that satisfies a guideline of a provider that provides electronic commerce.

  With this, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determining the evaluation of the content based on the positive example content that satisfies the condition regarding the guideline of the provider that provides the electronic commerce.

  In addition, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as a negative example content, content that does not satisfy the condition regarding the guideline of the provider that provides electronic commerce.

  Accordingly, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determining the evaluation of the content based on the negative example content that does not satisfy the condition regarding the guideline of the provider that provides the electronic commerce. .

  In addition, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as the normal content, the content selected by the user who purchases the product in the electronic commerce.

  As a result, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determining the evaluation of the content based on the positive example content selected by the user who purchases the product in the electronic commerce.

  Further, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as a negative example content, content selected by a user who has not purchased a product in electronic commerce.

  As a result, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determining the evaluation of the content based on the negative content selected by the user who has not purchased the product in the electronic commerce.

  In addition, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as a positive example content, the content selected by the user who performed the search using the predetermined keyword.

  As a result, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determining the evaluation of the content based on the positive example content selected by the user who performed the search using the predetermined keyword. .

  Further, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires, as a negative example content, content selected by a user who has not performed a search using a predetermined keyword.

  As a result, the providing apparatus 100 according to the embodiment can appropriately generate a model used for determination of content evaluation based on negative example content selected by a user who has not performed a search using a predetermined keyword. it can.

  Further, in the providing device 100 according to the embodiment, the acquisition unit 131 acquires positive example content and negative example content for each of a plurality of categories. The generation unit 132 generates a plurality of models corresponding to each of the plurality of categories based on the positive example content and the negative example content corresponding to each of the plurality of categories acquired by the acquisition unit 131.

  With this, the providing apparatus 100 according to the embodiment generates a plurality of models corresponding to each of the plurality of categories based on the positive example content and the negative example content corresponding to each of the plurality of categories, thereby The model used to determine the evaluation of can be appropriately generated.

[9. Hardware configuration)
The providing apparatus 100 according to the above-described embodiment is realized by, for example, a computer 1000 configured as shown in FIG. FIG. 13 is a hardware configuration diagram illustrating an example of a computer that realizes the function of the providing apparatus. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM (Read Only Memory) 1300, an HDD (Hard Disk Drive) 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 network N, sends the data to the CPU 1100, and transmits the data generated by the CPU 1100 to the other device via the network N.

  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 obtains data from an input device via the input / output interface 1600. Further, the CPU 1100 outputs the generated data to the output device via the input / output interface 1600.

  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 providing device 100 according to the embodiment, the CPU 1100 of the computer 1000 implements the function of the control unit 130 by executing a program loaded on the RAM 1200. 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 another device via the network N.

  As described above, some of the embodiments and modifications of the present application have been described in detail with reference to the drawings. However, these are merely examples, and various aspects based on the knowledge of those skilled in the art including the mode described in the row of the disclosure of the invention. It is possible to implement the present invention in other forms that are modified or improved.

[10. Other]
Further, of the processes described in the above-described embodiment and modified examples, all or part of the processes described as being automatically performed may be manually performed, or described as manually performed. It is also possible to automatically carry out all or part of the processing performed by a known method. In addition, the processing procedures, specific names, and information including various data and parameters shown in the above documents and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each drawing is not limited to the information shown.

  Further, each component of each device shown in the drawings is functionally conceptual, and does not necessarily have to be physically configured as shown. 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 and usage conditions. It can be integrated and configured.

  Further, the above-described embodiments and modified examples can be appropriately combined within a range that does not contradict processing contents.

  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.

1 Providing System 100 Providing Device (Generating Device)
121 user information storage unit 122 positive example information storage unit 123 negative example information storage unit 124 model information storage unit 125 product information storage unit 130 control unit 131 acquisition unit 132 generation unit 133 determination unit 134 provision unit 10 terminal device N network

Claims (15)

An acquisition unit that acquires positive example content that is content that is desired to be provided in electronic commerce and negative example content that is content that is not desired to be provided in electronic commerce,
A generation unit that generates a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit;
Equipped with
The acquisition unit is
A generating apparatus, characterized in that the content selected by a user who purchases a product in the electronic commerce is acquired as the positive example content . An acquisition unit that acquires positive example content that is content that is desired to be provided in electronic commerce and negative example content that is content that is not desired to be provided in electronic commerce,
A generation unit that generates a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit;
Equipped with
The acquisition unit,
A generation apparatus, characterized in that the content selected by a user who has not purchased a product in the electronic commerce is acquired as the negative content . An acquisition unit that acquires positive example content that is content that is desired to be provided in electronic commerce and negative example content that is content that is not desired to be provided in electronic commerce,
A generation unit that generates a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit;
Equipped with
The acquisition unit,
A generating apparatus, characterized in that content selected by a user who has performed a search using a predetermined keyword is acquired as the positive example content . An acquisition unit that acquires positive example content that is content that is desired to be provided in electronic commerce and negative example content that is content that is not desired to be provided in electronic commerce,
A generation unit that generates a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition unit;
Equipped with
The acquisition unit,
A generation apparatus, characterized in that content selected by a user who has not performed a search using a predetermined keyword is acquired as the negative content . Generating apparatus according to any one of claims 1-4, characterized in that to obtain the satisfying content related provider guidelines for providing the electronic commerce as the positive sample content. Generating apparatus according to any one of claims 1-5, characterized in that to obtain the content that does not satisfy the condition relating provider guidelines for providing the electronic commerce as the negative sample content. The acquisition unit,
Acquiring the positive example content and the negative example content for each of a plurality of categories,
The generator is
A plurality of models corresponding to each of the plurality of categories is generated based on the positive example content and the negative example content corresponding to each of the plurality of categories acquired by the acquisition unit. generating apparatus according to any one of claims 1-6. A generation method executed by a computer,
An acquisition step of acquiring a positive example content that is content that is desired to be provided in electronic commerce, and a negative example content that is content that is not desired to be provided in the electronic commerce,
A generation step of generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition step;
Only including,
The acquisition step is
A generation method, characterized in that content selected by a user who is purchasing a product in the electronic commerce is acquired as the positive example content . An acquisition procedure for acquiring a positive example content that is content that is desired to be provided in electronic commerce and a negative example content that is content that is not desired to be provided in electronic commerce,
A generation procedure for generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition procedure;
To the computer ,
The acquisition procedure is
Generator characterized that you get the content selected by the user that purchased a product in the electronic commerce as the positive sample content. A generation method executed by a computer,
An acquisition step of acquiring a positive example content that is content that is desired to be provided in electronic commerce, and a negative example content that is content that is not desired to be provided in the electronic commerce,
A generation step of generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition step;
Including,
The acquisition step is
A method of generating, wherein the content selected by a user who has not purchased a product in the electronic commerce is acquired as the negative example content. An acquisition procedure for acquiring a positive example content that is content that is desired to be provided in electronic commerce and a negative example content that is content that is not desired to be provided in electronic commerce,
A generation procedure for generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition procedure;
To the computer,
The acquisition procedure is
A generation program, wherein the content selected by a user who has not purchased a product in the electronic commerce is acquired as the negative content. A generation method executed by a computer,
An acquisition step of acquiring a positive example content that is a content that is desired to be provided in electronic commerce and a negative example content that is a content that is not desired to be provided in the electronic commerce,
A generation step of generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition step;
Including,
The acquisition step is
A generation method characterized in that content selected by a user who has performed a search using a predetermined keyword is acquired as the positive example content. An acquisition procedure for acquiring a positive example content that is content that is desired to be provided in electronic commerce and a negative example content that is content that is not desired to be provided in electronic commerce,
A generation procedure for generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition procedure;
To the computer,
The acquisition procedure is
A generation program, wherein the content selected by a user who has performed a search using a predetermined keyword is acquired as the positive example content. A generation method executed by a computer,
An acquisition step of acquiring a positive example content that is content that is desired to be provided in electronic commerce, and a negative example content that is content that is not desired to be provided in the electronic commerce,
A generation step of generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition step;
Including,
The acquisition step is
A generation method characterized in that content selected by a user who has not performed a search using a predetermined keyword is acquired as the negative content. An acquisition procedure for acquiring a positive example content that is content that is desired to be provided in electronic commerce and a negative example content that is content that is not desired to be provided in electronic commerce,
A generation procedure for generating a model used for evaluation of the content provided in the electronic commerce, based on the positive example content and the negative example content acquired by the acquisition procedure;
To the computer,
The acquisition procedure is
A generation program, characterized in that content selected by a user who has not performed a search using a predetermined keyword is acquired as the negative content.

Images