JP7086661B2 - Information processing systems, information processing methods, and programs - Google Patents

Description

The present invention relates to information processing systems , information processing methods, and programs.

A technique is known in which a degree of similarity between a news article and a user is obtained, and content suitable for the user's hobbies and tastes is distributed based on the degree of similarity (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2016-0388222

However, in the conventional technique, the process of obtaining the similarity between the content such as a news article and the user is complicated, and the processing load often becomes large.

The present invention has been made in view of the above problems, and an object of the present invention is to provide an information processing system , an information processing method, and a program capable of reducing the processing load by making the processing flow simple. There is.

In one aspect of the present invention, a processing reservation unit for storing a message specifying a processing content in a storage unit of a queue structure and a plurality of the messages stored in the storage unit have priority over the previously stored message. A plurality of processing execution units that are read from the storage unit and perform processing based on the read message in parallel, and an output unit that outputs the processing results of each of the plurality of processing execution units are provided. The reservation unit stores the query related to the content input by the creator of the content in the storage unit as the message, and the processing execution unit stores the query in the input query when the query is read from the storage unit. Based on this, the read query is input to the external device that selects and outputs the user to be the distribution target of the content from among the plurality of users who are the candidates for the distribution target of the content, and the output unit receives the read query. , An information processing device that outputs the output result of the external device to the query to the terminal device of the creator.

According to one aspect of the present invention, the processing flow can be simplified and the processing load can be reduced.

It is a figure which shows an example of the information processing system 1 including the information processing apparatus 100 in embodiment. It is a figure which shows an example of the screen of the terminal apparatus 10 to which the content was delivered. It is a figure which shows an example of the structure of the information processing apparatus 100 in an embodiment. It is a flowchart which shows the flow of the series of batch processing by the information processing apparatus 100 in embodiment. It is a figure for demonstrating the generation process of a user vector. It is a figure which shows an example of the result of the status stored in the 1st storage part DB1. It is a flowchart which shows the flow of the series of on-demand processing by the information processing apparatus 100 in embodiment. It is a figure for demonstrating the selection process of the user who is interested in a query. It is a figure which shows an example of the aggregation result of the user. It is a figure which shows an example of the hardware composition of the information processing apparatus 100 of an embodiment.

Hereinafter, an information processing system , an information processing method, and a program to which the present invention is applied will be described with reference to the drawings.

[Overview]
The information processing device is realized by one or more processors. The information processing device stores the message specifying the processing content in the storage unit of the queue structure, and among the plurality of messages stored in the storage unit, the message stored earlier is preferentially read from the storage unit and read out. Processes based on the messages are performed in parallel. Then, the information processing apparatus outputs the processing result of the processing based on the message.

In the information processing device, a query related to the content may be input as a message by the creator of the content. For example, a content creator may input a keyword that simply expresses the content into an information processing device as a query in order to identify a user who is interested in the content created by the content creator. In this case, the information processing apparatus stores the keyword input as a query in the storage unit of the queue structure. When a plurality of processes are performed in parallel based on the message read from the storage unit, the information processing device inputs the query to a certain external device if the read message is a query.

When a query is input, the external device derives the similarity between the characteristics of the entered query and the characteristics of each of the plurality of users who are candidates for content distribution, and based on the similarity. , A device that selects and outputs a user whose content is to be distributed from a plurality of users. For example, the external device selects a user whose similarity with the query feature is equal to or higher than the threshold value as the user to whom the content is to be distributed. A user whose similarity to a query feature is equal to or greater than a threshold value is a user who is expected to be interested in the content created by the creator of the content.

The information processing device produces content based on the output results obtained by inputting a query to an external device, that is, the aggregated results of users who are expected to be interested in the content created by the creator of the content. Output to the terminal device used by the person. In this way, by using the message queuing system, the processing flow can be simplified and the processing load of the entire device can be reduced.

[overall structure]
FIG. 1 is a diagram showing an example of an information processing system 1 including an information processing apparatus 100 according to an embodiment. The information processing system 1 in the embodiment includes, for example, one or more terminal devices 10, a service providing device 20, an information processing device 100, and a distributed processing device 200. These devices are connected via the network NW. Further, a part of these devices may be included in another device as a virtual device. For example, a part or all of the functions of the service providing device 20 and the distributed processing device 200 may be included in the information processing device 100. It may be a virtual machine realized by the functions of the above, and conversely, a part or all of the functions of the information processing apparatus 100 may be realized by the functions of the service providing apparatus 20 and the distributed processing apparatus 200. It may be a machine.

Each device shown in FIG. 1 transmits and receives various information via the network NW. The network NW includes, for example, the Internet, a WAN (Wide Area Network), a LAN (Local Area Network), a provider terminal, a wireless communication network, a wireless base station, a dedicated line, and the like. It should be noted that not all combinations of the devices shown in FIG. 1 need not be able to communicate with each other, and the network NW may include a local network in part.

The terminal device 10 is a terminal device including an input device, a display device, a communication device, a storage device, and an arithmetic unit, such as a mobile phone such as a smartphone, a tablet terminal, and various personal computers. The communication device includes a network card such as a NIC (Network Interface Card), a wireless communication module, and the like. In the terminal device 10, a UA (User Agent) such as a web browser or an application program is activated, and a request according to the content input by the user is transmitted to the service providing device 20. Further, the terminal device 10 in which the UA is activated causes the display device to display various images based on the information acquired from the service providing device 20.

The service providing device 20 is, for example, a web server that provides a web page to the terminal device 10 in response to a request from a web browser activated as a UA. The web page may be, for example, a web page constituting a website such as a shopping site, an auction site, or a free market site that sells products on the Internet. Further, the service providing device 20 may provide the terminal device 10 with a web page that provides various services such as a search site, an SNS (Social Networking Service), and a mail service. Further, the service providing device 20 may be an application server that provides the same services as the various websites described above by providing the content to the terminal device 10 in response to a request from the application started as the UA. .. The service providing device 20 is an example of a “content distribution unit”.

For example, when new content is registered in the service provided by the service providing device 20, the service providing device 20 shows more interest or interest in the newly registered content among a plurality of users who use the service. Deliver the content to the expected users.

FIG. 2 is a diagram showing an example of a screen of the terminal device 10 to which the content is distributed. For example, when a news article is newly submitted as one of the contents, the service providing device 20 distributes the news article to the terminal device 10 of the user who is more likely to be interested in the news article by push notification or the like. (R1 in the figure).

The information processing device 100 is a device that uses a message queuing system. The message is data transmitted and received by each component (each functional unit realized by the processor) included in the information processing apparatus 100 to and from each other via a storage device having a queue structure. Since each component included in the information processing apparatus 100 sends and receives a message via a storage device having a queue structure, it is possible to independently perform processing at different cycles from each other. The message includes a message requesting processing from one component to the other component (hereinafter referred to as a processing request message), and a component of the person who received the processing request is a component of the person who requested the processing. A message for notifying that the processing is completed (hereinafter referred to as a processing completion message) is included. The processing completion message is an example of "processing completion information".

The information processing apparatus 100 performs, for example, batch processing in which predetermined processing is repeated in a predetermined cycle (for example, one day) and on-demand processing in which processing is performed in response to a predetermined user's request. The information processing apparatus 100 instructs the distributed processing apparatus 200 to generate a user vector, for example, as a batch process. The user vector is, for example, a vector showing the characteristics of each of a plurality of users who are candidates for distribution of newly registered contents.

Further, for example, when a query regarding a content newly created by a content creator is input as on-demand processing, the information processing device 100 has a plurality of candidates for content distribution to the distributed processing device 200. Instructs the process of selecting a user who is expected to be interested in the newly created content from the users. The query may be, for example, one word (word) or a phrase (phrase) containing a plurality of words.

The distributed processing device 200 uses, for example, a distributed file system called Hadoop to divide unstructured metadata such as images and text into a plurality of data, and parallel to each of the divided data. It is a device that performs processing. For example, when a query is input from the information processing device 100, the distributed processing device 200 generates a query vector indicating the characteristics of the input query, and the generated query vector and a plurality of users generated in advance. Derivation of similarity with the vector by parallel processing. The distributed processing device 200 extracts a user vector whose similarity with the query vector is equal to or higher than the threshold value from a plurality of user vectors, and selects the user who is the source of the extracted user vector as the user to be the content distribution target. Then, the result is output to the information processing apparatus 100.

[Information processing device configuration]
FIG. 3 is a diagram showing an example of the configuration of the information processing apparatus 100 in the embodiment. As shown in the figure, the information processing apparatus 100 includes, for example, a job management unit 110, a job status monitoring unit 120, a job execution unit 130, a job execution status output unit 140, an on-demand processing reservation unit 150, and a user ID. It includes a conversion unit 160, a first storage unit DB1, a second storage unit DB2, and a third storage unit DB3. The combination of the job management unit 110, the on-demand processing reservation unit 150, and the user ID conversion unit 160 is an example of the "processing reservation unit". Further, the job management unit 110 is an example of the "processing management unit", and the job execution unit 130 is an example of the "processing execution unit".

The job management unit 110, the job status monitoring unit 120, the job execution unit 130, the job execution status output unit 140, the on-demand processing reservation unit 150, and the user ID conversion unit 160 include, for example, a processor such as a CPU (Central Processing Unit). It is realized by executing the programs stored in various storage units. Further, a part or all of the job management unit 110, the job status monitoring unit 120, the job execution unit 130, the job execution status output unit 140, the on-demand processing reservation unit 150, and the user ID conversion unit 160 are LSI (Large Scale). It may be realized by hardware (circuitry) such as Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or GPU (Graphics Processing Unit), or software and hardware. It may be realized by collaboration.

The first storage unit DB1, the second storage unit DB2, and the third storage unit DB3 are, for example, HDD (Hard Disc Drive), flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), ROM (Read Only Memory), RAM. It is realized by a storage device such as (Random Access Memory). Various programs such as firmware and application programs are stored in these storage devices.

The processing result of the job status monitoring unit 120 is stored in the first storage unit DB1.

The second storage unit DB2 includes a storage area (hereinafter, referred to as a message queue) having a queue structure in which messages are stored. The messages stored in the message queue are read out in order from the one stored first, unless a priority described later is provided. Of the various messages described above, at least the processing request message is stored in the message queue.

In the third storage unit DB3, identification information (hereinafter, referred to as a user ID) of each of a plurality of users who are candidates for content distribution is stored in advance. The user ID is an ID assigned to a user who uses the service provided by the service providing device 20, and can be arbitrarily set by the user.

The job management unit 110 refers to the execution status of the process (job) monitored by the job status monitoring unit 120, and if there is a process that has stopped in the middle without being completed, the stopped process is transmitted to the job execution unit 130. Re-execute.

The job status monitoring unit 120 monitors the execution status of processing by the job execution unit 130, which will be described later, and stores the monitoring result in the first storage unit DB1.

The job execution unit 130 includes a plurality of workers from the first worker 132 (1) to the k-th worker 132 (k: k is an arbitrary natural number). Each worker 132 is preferentially read from the previously stored processing request message among one or more processing request messages stored in the message queue of the second storage unit DB2, and is designated by the read processing request message. Perform processing. When the processing specified by the processing request message is completed, each worker 132 stores a processing completion message indicating that the processing is completed in the second storage unit DB2. The processing completion message stored in the second storage unit DB2 is referred to by the job status monitoring unit 120 described above. Each worker 132 repeats a series of processes of reading a processing request message from the above-mentioned message queue and storing the processing completion message in the second storage unit DB 2 when the processing is completed, in a predetermined cycle. Hereinafter, the state in which the worker 132 is executing the process is referred to as an “operating state”, and the state in which the worker 132 is not executing the process is referred to as a “standby state”.

The job execution status output unit 140 reads information indicating the result of the processing execution status of the job execution unit 130 from the first storage unit DB1, and reads the information read from the first storage unit DB1 via the network NW. It is provided (transmitted) to the terminal device 10 used by the creator of the above. For example, the job execution status output unit 140 reads from the first storage unit DB1 from the web browser started by the terminal device 10 via the interface (WebUI) for confirming the processing execution status of the job execution unit 130. Information may be provided to the terminal device 10.

Further, the job execution status output unit 140 transmits to the job management unit 110 information indicating the result of the processing execution status of the job execution unit 130, which is read from the first storage unit DB1.

For example, when a word or a phrase is input to a query input window provided in a Web UI for inputting a query, the on-demand processing reservation unit 150 uses the word or phrase as a query. For example, the creator of the content may create a word or phrase that simply expresses the content from the content of the content created by himself / herself and input it into the query input window, or a word extracted from the text contained in the content. Or words may be entered in the query input window. For example, the creator of the content creates a news article about the online pick, and enters words and phrases related to the online pick such as "○○ Olympics", "○○ Olympics", and "figure skating_gold medal" in the query input window. Suppose you did. In this case, the on-demand processing reservation unit 150 queries for words and phrases such as "○○ Olympics", "○○ Olympics", and "figure skating_gold medal". Then, the on-demand processing reservation unit 150 stores the word or phrase input as a query as a message (hereinafter referred to as a query message) in the message queue. The query message is one of the processing request messages. The processing content based on the query message will be described later.

The user ID conversion unit 160 acquires account information of a plurality of users who are candidates for content distribution from the distributed processing device 200. The account information includes, for example, an e-mail address set as a user ID by the user, and personal information such as a password, date of birth, name, and gender.

The user ID conversion unit 160 extracts an e-mail address from the acquired account information, stores the extracted e-mail address as a user ID in the third storage unit DB3, and encrypts the extracted user ID. Then, the user ID conversion unit 160 stores the processing request message requesting the generation of the user vector of the user identified by the encrypted user ID in the message queue of the second storage unit DB2. The user identified by the encrypted user ID is an example of the "target user to be derived from the feature".

[Flowchart of batch processing]
Hereinafter, the flow of a series of batch processing by the information processing apparatus 100 in the embodiment will be described according to a flowchart. FIG. 4 is a flowchart showing a flow of a series of batch processing by the information processing apparatus 100 in the embodiment. The processing of this flowchart may be repeated in a predetermined cycle such as an hourly unit, a daily unit, or a weekly unit.

First, the user ID conversion unit 160 stores the processing request message requesting the generation of the user vector of the user identified by the encrypted user ID in the message queue of the second storage unit DB2 (S100). This processing request message includes an encrypted user ID.

Next, the information processing apparatus 100 determines whether or not the standby worker 132 exists among the plurality of workers from the first worker 132 (1) to the k-th worker 132 (k) (S102). If there is no worker 132 in the standby state (when all the workers 132 are in the operating state), the processing of this flowchart is terminated.

On the other hand, when the worker 132 in the standby state exists (when all the worker 132s are not in the operating state), the worker 132 in the standby state reads the processing request message from the message queue and the processing specified by the read processing request message. Is started (S104).

For example, when the read processing request message is a processing request message requesting the generation of a user vector, the worker 132 sends the encrypted user ID included in the processing request message to the distributed processing device 200. Then, the distributed processing apparatus 200 is made to generate a user vector.

FIG. 5 is a diagram for explaining a user vector generation process. For example, the distributed processing device 200 that has received the request decodes the encrypted user ID included in the processing request message, and creates a user vector based on the user's action history associated with the decrypted user ID. Generate. The user's behavior history includes, for example, the text included in the content viewed by the user and the log of the query entered at the time of content search. It is assumed that the user's action history has been collected in advance by the distributed processing device 200.

For example, when the content viewed by the user is a news article, the distributed processing device 200 divides the title and body of the news article into a plurality of words (morphemes) by morphological analysis, and each divided word is used as an element of the user vector. decide. At this time, the distributed processing apparatus 200 uses the value of each element as the weight of the word. For example, the distributed processing apparatus 200 uses a method called TF (Term Frequency) -IDF (Inverse Document Frequency) to evaluate the frequency of occurrence of words in each divided word (or a phrase that is a combination of a plurality of words). Weight is given to it. The TF represents the frequency of appearance of each word (morpheme) in the text (title and body), and the TF value increases as the number of times of occurrence in the target text increases. The IDF represents the frequency of the text in which a word appears, and the IDF value tends to appear in the text contained in the content viewed by the user and appears in the text contained in other content not viewed by the user. The harder the word, the bigger it gets. For example, the distributed processing apparatus 200 uses the product of the TF value and the IDF value as the weight of each word (hereinafter referred to as the first weight).

Further, the distributed processing device 200 derives the TF value by the above-mentioned method in order to reflect the difference in the viewing tendency of the content among the users in the weight of the word, and the target user (the user included in the processing request message). The IDF value (hereinafter, the IDF value) tends to be larger for other users (users with a user ID different from the user ID included in the processing request message), which are frequently viewed (ID users), and for text words included in the content, which are less frequently viewed. , Second IDF value) is derived. The distributed processing apparatus 200 uses the product of the TF value and the second IDF value as the second weight of each word.

Then, the distributed processing apparatus 200 determines the value of each element based on the first weight and the second weight, and generates a multidimensional vector having the plurality of element values as a user vector. Each time the distributed processing apparatus 200 receives a request for generating a user vector, the distributed processing apparatus 200 generates a user vector of the user identified by the user ID based on the user ID included in the processing request message. For example, when the distributed processing apparatus 200 generates a user vector, the user vector is stored in an internal storage device, and information indicating that the generation of the user vector is completed is returned to the worker 132. Upon receiving the information indicating that the generation of the user vector is completed from the distributed processing device 200, the worker 132 stores the processing completion message for the processing request message requesting the generation of the user vector in the second storage unit DB2.

Returning to FIG. 4, the job status monitoring unit 120 then determines whether or not the processing completion message is stored in the second storage unit DB2 (S106), and the processing completion message is stored in the second storage unit DB2. In this case, the status (execution status) of the process based on the process request message is set to the status indicating that the process is completed, and the result of the status is stored in the first storage unit DB1 (S108).

On the other hand, when the processing completion message is not stored in the second storage unit DB2, the job status monitoring unit 120 determines whether or not a predetermined time has elapsed since the worker 132 who read the processing request message from the message queue started processing. (S110), and when a predetermined time has elapsed since the processing was started, the status of the processing based on the processing request message is set to the status indicating that an abnormality (error) has occurred without completing the processing. The result of the status is stored in the first storage unit DB1 (S112). This ends the processing of this flowchart.

FIG. 6 is a diagram showing an example of the result of the status stored in the first storage unit DB1. As shown in the illustrated example, the job status monitoring unit 120 processes the job ID, which is the identification information of each processing request message, with respect to the worker ID, which is the identification information of the worker 132 that has read the processing request message. The information in which the date and time when the process was started, the date and time when the worker 132 completed the process, and the status may be stored in the first storage unit DB1 as the result of the status. In the illustrated example, the process of the job ID requested to the worker 132 (2) whose worker ID is “2” is not completed. In this case, the job status monitoring unit 120 sets the status of the processing of "DDD" requested to the worker 132 (2) to the error status.

The job execution status output unit 140 refers to the status result stored in the first storage unit DB1, and if an error status exists, extracts the job ID of the error status from the status result, and the extracted job ID. To the job management unit 110. The job execution status output unit 140 may transmit the status result stored in the first storage unit DB1 to the job management unit 110 as it is.

The job management unit 110 refers to the job ID of the error status, and stores the same processing request message as the job ID, that is, the processing request message for re-executing the processing, in the message queue of the second storage unit DB2. At this time, the job management unit 110 gives the highest priority to the processing request message. The priority is to rank which processing request message is read from among the processing request messages stored in the message queue. For example, the processing request messages stored in the message queue by the on-demand processing reservation unit 150 and the user ID conversion unit 160 have no priority, and the processing request messages are read out in the order stored in the message queue. .. On the other hand, the process request message stored in the message queue by the job management unit 110 has the highest priority, so even if another process request message is already stored in the message queue, it will be later. The stored process request message for re-execution is read first. As a result, any worker 132 in the standby state executes the process whose status is an error with the highest priority.

[Flowchart of on-demand processing]
Hereinafter, the flow of a series of on-demand processing by the information processing apparatus 100 in the embodiment will be described according to a flowchart. FIG. 7 is a flowchart showing a flow of a series of on-demand processing by the information processing apparatus 100 in the embodiment.

First, the on-demand processing reservation unit 150 determines, for example, whether or not a query has been input to the WebUI (S200), and when a query is input to the WebUI, a second storage of a query message including the input query. It is stored in the message queue of the unit DB2 (S202).

Next, the information processing apparatus 100 determines whether or not the standby worker 132 exists among the plurality of workers from the first worker 132 (1) to the k-th worker 132 (k) (S204). If there is no worker 132 in the standby state (when all the workers 132 are in the operating state), the processing of this flowchart is terminated.

On the other hand, when the worker 132 in the standby state exists (when all the worker 132s are not in the operating state), the worker 132 in the standby state reads the processing request message from the message queue and the processing specified by the read processing request message. Is started (S206).

For example, when the read processing request message is a query message, the worker 132 sends the query included in the query message to the distributed processing device 200, so that the distributed processing device 200 has a plurality of candidates for content distribution. Have users select users who are interested in the query. Since words and phrases related to the content created by the content creator are input to the WebUI as a query, a user who is interested in the query is also a user who is interested in the content created by the content creator. be. A user who is interested in a query is an example of a "user to whom content is delivered".

FIG. 8 is a diagram for explaining a selection process of a user who is interested in a query. For example, when the distributed processing apparatus 200 acquires a query from the worker 132, it generates a query vector indicating the characteristics of the query. For example, the distributed processing apparatus 200 may generate a query vector from a query by using a method called a distributed expression or a word embedding expression (Word Embedding (s)) such as word2vec or GloVe. A distributed expression or word embedding expression is a method of vectorizing a word or phrase based on the co-occurrence of the word or phrase and the word or phrase that appears before or after the word or phrase, and is prepared in advance, for example. This is a method of finding the appearance probability of words that appear before and after a query in a certain context based on a corpus containing a plurality of words and phrases, and generating a multidimensional vector with the appearance probability as an element value.

For example, when the distributed processing device 200 generates a query vector, the distributed processing apparatus 200 derives the similarity (for example, cosine similarity) between the query vector and each of the plurality of user vectors generated in advance, and uses the query vector as a reference. A predetermined number of high-ranking user vectors (for example, 10) having a high degree of similarity are extracted, or all user vectors having a degree of similarity with a query vector equal to or higher than a threshold value are extracted. Then, the distributed processing apparatus 200 selects the user of the user ID that is the source of the extracted user vector as the user who is interested in the query.

When the distributed processing device 200 selects a user who is interested in a query from a plurality of users who are candidates for content distribution, the distributed processing device 200 shows an aggregation result (user selection result) of users who are interested in the query. The information and the information indicating that the user's selection process is completed are returned to the worker 132. The distributed processing device 200 may include users other than the users selected as the users who are interested in the query in the aggregation result of the users. When the worker 132 receives the information indicating the user's aggregation result and the information indicating that the user's selection processing is completed from the distributed processing device 200, the worker 132 receives the processing completion message for the query message and the information indicating the user's aggregation result. And are stored in the second storage unit DB2.

Returning to FIG. 7, the job status monitoring unit 120 then determines whether or not the processing completion message for the query message and the information indicating the user's aggregation result are stored in the second storage unit DB2 (S208). , When both are stored in the second storage unit DB2, the status (execution status) of the process based on the query message is set to the status indicating that the process is completed, and the result of the status is stored in the first storage unit DB1. At the same time, the information indicating the aggregation result of the user is stored in the first storage unit DB1 (S210).

Next, the job execution status output unit 140 provides information indicating the aggregated results of the users stored in the first storage unit DB1 to the terminal device 10 of the content creator via WebUI or the like (S212).

FIG. 9 is a diagram showing an example of the aggregated results of users. For example, the job execution status output unit 140 provides the number of users who are interested in the query or the users who are interested in the co-content as the "number of users targeted for targeting notification" as the aggregation result of the users. Further, the job execution status output unit 140 uses the user ID (encrypted user ID) of each user as the identification information (device ID in the figure) of the terminal device 10 used by the user or the terminal device 10. Information associated with the type of installed OS (Operating System) and the like may be provided as an aggregation result of users. Further, the job execution status output unit 140 may provide a graph or the like showing the distribution of the number of users with respect to the magnitude of the degree of similarity with the query vector as the aggregation result of the users. Since such information is provided to the content creator as the aggregated result of the users, the content creator can know how many users are interested in the content produced by the content creator. As a result, the creator of the content can reorganize the created content or replace it with other content.

Further, the job execution status output unit 140 provides services in place of or in addition to providing the information indicating the aggregated results of the users stored in the first storage unit DB1 to the terminal device 10 of the content creator. It may be provided to the device 20. In this case, the service providing device 20 is used by the content creator on the terminal device 10 of the user included in the user's aggregation result, that is, the user who is expected to be interested in the content produced by the content creator. The created content may be delivered by push notification or the like.

On the other hand, when the job status monitoring unit 120 does not store either or both of the processing completion message for the query message and the information indicating the user's aggregation result in the second storage unit DB2, the job status monitoring unit 120 reads the query message from the message queue. It is determined whether or not a predetermined time has elapsed since the worker 132 started the process (S214), and if the predetermined time has elapsed since the process started, the status of the process based on the query message is set and the process is completed. Instead, a status indicating that an abnormality (error) has occurred is set, and the result of that status is stored in the first storage unit DB1 (S216). This ends the processing of this flowchart.

According to the embodiment described above, among the job management unit 110, the on-demand processing reservation unit 150, or the user ID conversion unit 160 that stores the processing request message in the message queue, and the processing request message stored in the message queue. , A job that preferentially reads the processing request message stored earlier from the message queue, performs processing based on the read processing request message in parallel, and outputs the processing results of each of the multiple worker 132s. The on-demand processing reservation unit 150 includes an execution status output unit 140, and stores a query related to the content input by the creator of the content in the message queue as a query message which is one of the processing request messages, and a plurality of queries are stored in the message queue. When the worker 132 in the standby state among the workers 132 reads the query message from the message queue, the worker 132 in the standby state inputs the query included in the query message to the distributed processing device 200. The distributed processing device 200 converts the input query into a query vector, and the plurality of users are based on the similarity between the query vector and each user vector of the plurality of users who are candidates for content distribution. The user to whom the content is to be distributed is selected from the above, and the result is output to the worker 132. The job execution status output unit 140 provides the output result of the distributed processing device 200 to the terminal device 10 of the content creator. By using such a message queuing system, each component of the information processing apparatus 100 can independently perform processing at different cycles, so that the processing flow can be simplified. It is possible to reduce the processing load of the entire device.

Further, according to the above-described embodiment, when there is a worker 132 that does not output a processing completion message within a predetermined time after reading the processing request message from the message queue among the plurality of workers 132, job management is performed. The unit 110 gives the highest priority to the processing request message for re-executing the processing performed by the worker 132, and stores it in the message queue. As a result, even if another processing request message is already stored in the message queue, the processing request message for re-execution stored later is read first, so the processing stopped in the middle is re-executed with the highest priority. be able to. As a result, the re-executability of the process can be improved.

Further, according to the above-described embodiment, since the user ID is encrypted in advance and various processes such as the user vector generation process are performed, the processing load of the entire device is increased while increasing the confidentiality of the user's personal information. Can be mitigated.

Further, according to the above-described embodiment, since the message queuing system is used, it is not necessary to synchronize the components, and the processing content of each component of the information processing apparatus 100 can be changed or newly. It is possible to easily add various components. As a result, the modularity of the information processing apparatus 100 can be improved.

<Hardware configuration>
The information processing apparatus 100 of the above-described embodiment is realized by, for example, a hardware configuration as shown in FIG. FIG. 10 is a diagram showing an example of the hardware configuration of the information processing apparatus 100 of the embodiment.

The information processing device 100 includes NIC100-1, CPU100-2, RAM100-3, ROM100-4, a secondary storage device 100-5 such as a flash memory or an HDD, and a drive device 100-6 on an internal bus or a dedicated communication line. It is configured to be interconnected by. A portable storage medium such as an optical disk is mounted on the drive device 100-6. The program stored in the portable storage medium mounted on the secondary storage device 100-5 or the drive device 100-6 is expanded into the RAM 100-3 by a DMA controller (not shown) or the like, and executed by the CPU 100-2. As a result, each component of the job management unit 110, the job status monitoring unit 120, the job execution unit 130, the job execution status output unit 140, the on-demand processing reservation unit 150, and the user ID conversion unit 160 is realized. The programs referenced by these components may be downloaded from other devices via the network NW.

Although the embodiments for carrying out the present invention have been described above using the embodiments, the present invention is not limited to these embodiments, and various modifications and substitutions are made without departing from the gist of the present invention. Can be added.

1 ... Information processing system, 10 ... Terminal device, 20 ... Service providing device, 100 ... Information processing device, 110 ... Job management unit, 120 ... Job status monitoring unit, 130 ... Job execution unit, 132 ... Worker, 140 ... Job execution Situation output unit, 150 ... on-demand processing reservation unit, 160 ... user ID conversion unit, DB1 ... first storage unit, DB2 ... second storage unit, DB3 ... third storage unit, 200 ... distributed processing device

Claims (8)

An information processing system including an information processing device and a distributed processing device for distributed processing of information input from the information processing device.
The information processing device is
A processing reservation unit that stores a message that specifies the processing content in the storage unit of the queue structure,
Among the plurality of messages stored in the storage unit, a plurality of processing execution units that preferentially read the previously stored message from the storage unit and perform processing based on the read message in parallel.
It has an output unit that outputs the processing results of each of the plurality of processing execution units.
The processing reservation unit stores the query regarding the content input by the creator of the content in the storage unit as the message.
When the query is read from the storage unit, the processing execution unit inputs the read query to the distributed processing device.
When the query is input from the processing execution unit, the distributed processing device has a degree of similarity between the characteristics of the input query and the characteristics of each of the plurality of users who are candidates for distribution of the content. Among them, the user whose similarity is equal to or higher than the threshold is selected as the user to be the distribution target of the content, and the selection result of the user to be the distribution target of the content is output to the information processing apparatus.
The output unit outputs the output result of the distributed processing device to the query to the terminal device of the creator.
Information processing system. When the processing based on the read message is completed, each of the plurality of processing execution units outputs the processing completion information indicating that the processing is completed.
The information processing apparatus further includes a monitoring unit that monitors the execution status of each process of the plurality of process execution units based on the process completion information output by each of the plurality of process execution units.
The information processing system according to claim 1 . The information processing apparatus displays the processing completion information in the plurality of processing execution units whose processing execution status has been monitored by the monitoring unit from the time when the message is read from the storage unit until a predetermined time elapses. When there is a predetermined process execution unit that does not output, one of the plurality of process execution units further has a process management unit that re-executes the process based on the message read by the predetermined process execution unit.
The information processing system according to claim 2 . The processing management unit has the highest priority of the message read by the predetermined processing execution unit and stores it in the storage unit again, based on the message read by the predetermined processing execution unit. Re-execute the process with the highest priority,
The information processing system according to claim 3 . The processing reservation unit further stores a message designating a target user to be derived from the feature from the plurality of users in the storage unit.
When the processing execution unit reads a message designating the target user from the storage unit, the processing execution unit requests the distributed processing apparatus to derive the characteristics of the target user.
The distributed processing apparatus that has received the request derives the characteristics of the target user based on the behavior history of the target user.
The information processing system according to claim 1 . The information processing device further includes a content distribution unit that distributes the content of the creator to the user selected as the distribution target of the content by the distributed processing device.
The information processing system according to any one of claims 1 to 5 . Information processing equipment
A message that specifies the processing content is stored in the storage unit of the queue structure.
Of the plurality of the messages stored in the storage unit, the previously stored message is preferentially read from the storage unit.
Processing based on the read message is performed in parallel,
The processing result of the processing based on the read message performed in parallel is output.
A query related to the content input by the creator of the content is stored in the storage unit as the message.
When the query is read from the storage unit, the read query is input to the distributed processing device that performs distributed processing of the information input from the information processing device.
When the query is input from the information processing device, the distributed processing device has a degree of similarity between the characteristics of the input query and the characteristics of each of the plurality of users who are candidates for distribution of the content. Among them, the user whose similarity is equal to or higher than the threshold is selected as the user to be the distribution target of the content, and the selection result of the user to be the distribution target of the content is output to the information processing apparatus.
The information processing device outputs the output result of the distributed processing device to the query to the terminal device of the creator.
Information processing method. To the computer of the information processing device
A process to store a message that specifies the processing content in the storage unit of the queue structure,
Of the plurality of messages stored in the storage unit, the message stored earlier is preferentially read from the storage unit, and
The process of performing the process based on the read message in parallel and the process
A process to output the process result of the process based on the read message performed in parallel, and a process to output the process result.
A process of storing a query related to the content input by the creator of the content in the storage unit as the message.
When the query is read from the storage unit, the distributed processing device that performs distributed processing of the information input from the information processing device is made to execute the process of inputting the read query.
When the query is input from the information processing device to the computer of the distributed processing device, the characteristics of the input query are similar to the characteristics of each of the plurality of users who are candidates for distribution of the content. Of the degrees, a user whose similarity is equal to or higher than the threshold is selected as a user to be delivered to the content, and a process of outputting the selection result of the user to be delivered to the content to the information processing apparatus is executed. ,
A program for causing a computer of the information processing device to execute a process of outputting the output result of the distributed processing device to the query to the terminal device of the creator.

Images