JP7822915B2 - Information processing device, information processing method, and information processing program - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and an information processing program.

Technologies have been known in the past to provide assistance with search queries, which are character strings entered by users when conducting internet searches, etc. One example of such technology is a technique that, when an entered search query corresponds to a character string that could be entered as a typographical error, converts it into a character string associated with that character string that could be entered as a typographical error, and then uses the converted character string as a search query to search for information.

Patent No. 6529456

However, the above-mentioned techniques may not be able to identify the appropriate space separator position in a search query.

For example, the above-mentioned technology only converts typographical errors into corresponding correct strings, and cannot be said to be able to identify the appropriate space separators in a search query.

This application was made in consideration of the above, and aims to identify appropriate space delimiter positions in a search query.

The information processing device according to the present application is characterized by having a generation unit that generates a character string indicated by a search query entered by a user by deleting space separators included in the search query; a classification unit that classifies the search queries based on the character strings generated by the generation unit; and an identification unit that identifies the position of space separators in the character strings generated from each search query based on the number of times each search query classified into the same group has been entered.

According to one aspect of the embodiment, it is possible to identify the appropriate space delimiter position in a search query.

FIG. 1 is a diagram illustrating an example of information processing according to an embodiment. FIG. 2 is a diagram illustrating an example of the configuration of the information processing device 10 according to the embodiment. FIG. 3 is a diagram illustrating an example of the search query information database 31 according to the embodiment. FIG. 4 is a diagram showing an example of the group information database 32 according to the embodiment. FIG. 5 is a flowchart illustrating an example of a procedure for information processing according to the embodiment. FIG. 6 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing device 10.

Below, modes for implementing an information processing device, information processing method, and information processing program according to the present application (hereinafter referred to as "embodiments") will be described with reference to the drawings. Note that the information processing device, information processing method, and information processing program according to the present application are not limited to these embodiments. Furthermore, in the following embodiments, identical components will be designated by the same reference numerals, and duplicate explanations will be omitted.

1. Embodiment
Information processing implemented by an information processing device or the like according to this embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of information processing according to this embodiment. Note that in Fig. 1, the information processing according to this embodiment is implemented by an information processing device 10, which is an example of the information processing device according to the present application.

As shown in FIG. 1, the information processing system 1 according to the embodiment includes an information processing device 10 and a user terminal 100. The information processing device 10 and the user terminal 100 are connected to each other via a network N (see, for example, FIG. 2) via a wired or wireless connection so that they can communicate with each other. The network N is, for example, a wide area network (WAN) such as the Internet. Note that the information processing system 1 shown in FIG. 1 may include multiple information processing devices 10 and multiple user terminals 100.

The information processing device 10 shown in FIG. 1 is an information processing device that performs information processing, and is realized, for example, by a server device or a cloud system. For example, the information processing device 10 provides a search service and manages information indicating the history of search queries entered by users in the search service in a storage unit of the information processing device itself. Then, the information processing device 10 identifies appropriate space delimiter positions for the search query based on the search query history.

The services provided by the information processing device 10 may be, for example, a search service, shopping service, payment service, route search service, map provision service, travel service, restaurant introduction service, weather forecast service, schedule management service, news provision service, auction service, video content distribution service, financial transaction (stock trading, etc.) service, etc.

Here, conventionally, when a user enters a search query into a search service, etc., there are cases where spaces are not entered into the search query due to the user's input error, or where spaces are not entered because the search query is entered by voice. In such cases, a method can be considered that uses morphological analysis technology to automatically enter spaces into the search query entered by the user. However, when using such a method, for example, if the search query includes recently created proper nouns (e.g., the names of content such as manga, movies, or songs), the proper nouns may be excessively separated.

The information processing device of this embodiment therefore performs the following information processing to identify appropriate space delimiter positions in a search query.

The user terminal 100 shown in Figure 1 is an information processing device used by a user. The user terminal 100 may be realized, for example, by a smartphone, tablet terminal, notebook PC (Personal Computer), desktop PC, mobile phone, or PDA (Personal Digital Assistant). In the example shown in Figure 1, the user terminal 100 is a smartphone used by a user.

An example of information processing performed by information processing device 10 will be described below using FIG. 1. In the following description, user terminal 100 will be described as user terminals 100-1 to 100-N (N is any natural number) depending on the user using user terminal 100. For example, user terminal 100-1 is the user terminal 100 used by a user (user U1) identified by user ID "UID#1." In the following description, user terminals 100-1 to 100-N will be referred to as user terminal 100 when there is no particular distinction between them. In the following description, user terminal 100 may be considered the same as the user. In other words, user U1 may be read as user terminal 100-1.

First, the information processing device 10 generates a character string indicated by a search query entered by a user in a search service (step S1). For example, the information processing device 10 generates the character string by deleting space separators included in the search query.

As a specific example, the information processing device 10 generates the character string "AABBCC" from the search query "AA BB CC" by deleting the space separator between "AA" and "BB" and the space separator between "BB" and "CC" in the search query "AA BB CC." Furthermore, the information processing device 10 generates the character string "AABBCC" from the search query "AA BBCC" by deleting the space separator between "AA" and "BBCC" in the search query "AA BBCC." Note that, because there are no space separators in the search query "AABBCC," the information processing device 10 generates the character string "AABBCC" that directly represents the search query "AABBCC."

Next, the information processing device 10 categorizes the search queries based on the character strings generated from the search queries (step S2). For example, the information processing device 10 categorizes search queries that have matching character strings after removing spaces into the same group. To give a specific example, the information processing device 10 categorizes the search queries "AABBCC," "AA BB CC," and "AA BBCC" into group #1 because the character string generated from each of the search queries is "AABBCC." Furthermore, the information processing device 10 categorizes the search queries "DDEEFF," "DDEE FF," and "DD EE FF" into group #2 because the character string generated from each of the search queries is "DDEEFF."

Next, the information processing device 10 identifies the positions of space separators in the character string based on the search query that has been most frequently entered by users in the search service among the search queries classified into the same group (step S3). For example, since the search query "AA BB CC" has been entered most frequently in group #1, the information processing device 10 identifies that appropriate positions of space separators in the character string "AABBCC" are between "AA" and "BB" and between "BB" and "CC."

Furthermore, for example, since the search query "DDEE FF" is entered most frequently in group #2, the information processing device 10 determines that the appropriate space delimiter position for the string "DDEEFF" is between "DDEE" and "FF."

Next, the information processing device 10 trains a model that, when a string is input, adds a space separator or deletes the space separator based on the position of the space separator in the string identified in step S3, and outputs the string (step S4). For example, when a search query entered by a user in a search service is input, the information processing device 10 trains a model that adds a space separator to the search query or deletes the space separator from the search query.

Note that the information processing device 10 may generate and learn models using various conventional machine learning techniques, as appropriate. For example, models may be generated using supervised machine learning techniques such as SVM (Support Vector Machine). Models may also be generated using unsupervised machine learning techniques. Models may also be generated using deep learning techniques. Models may also be generated using various deep learning techniques, such as DNN (Deep Neural Network), RNN (Recurrent Neural Network), and CNN (Convolutional Neural Network). Note that the above descriptions of model generation and learning are merely examples, and models may be generated using a learning method selected appropriately depending on the available information, etc.

Next, the information processing device 10 receives from the user terminal 100 a search query entered by the user in the search service (step S5). Next, the information processing device 10 edits the search query entered by the user using the positions of the space separators identified in step S3 and the model trained in step S4, and provides search results based on the edited search query to the user terminal 100 (step S6). For example, if user U1 enters the search query "AA BBCC" (in other words, if the user enters a search query that would be classified as group #1 if the space separators were deleted), the information processing device 10 edits the search query to "AA BB CC" and provides search results based on the edited search query. Also, if user U2 enters the search query "DD EE FF" (in other words, if the user enters a search query that would be classified as group #2 if the space separators were deleted), the information processing device 10 edits the search query to "DDEE FF" and provides search results based on the edited search query.

Furthermore, if user U3 inputs a search query (e.g., "GG HHII") indicating the string "GGHHII" without identifying the appropriate space separator, the information processing device 10 inputs the search query into the model and provides search results based on the output string "GGHH II."

Next, the information processing device 10 acquires behavioral information related to the user's behavior after the search results were provided from the user terminal 100 (step S7). For example, the information processing device 10 acquires behavioral information of user U3, who was provided with search results based on the character string "GGHH II" output from the model, from the user terminal 100-3. As a specific example, the information processing device 10 acquires behavioral information indicating whether user U3 selected the content provided as a search result based on "GGHH II."

Next, the information processing device 10 retrains the model based on the user's behavioral information (step S8). For example, if the behavioral information of user U3 indicates that user U3 selected content provided as a search result based on "GGHH II", the information processing device 10 trains the model using the edited string "GGHH II" of the string "GGHHII" as a positive example. Also, if the behavioral information of user U3 indicates that user U3 did not select content provided as a search result based on "GGHH II", the information processing device 10 trains the model using the edited string "GGHHII" of the string "GGHH II" as a negative example.

As described above, the information processing device 10 according to the embodiment identifies the position of space separators in the character string indicated by each search query based on the search query most frequently entered by users among the search queries classified into the same group. In other words, the information processing device 10 according to the embodiment can identify the appropriate position of space separators in the search query.

[2. Other processing examples]
The above-described process is merely an example, and the information processing device 10 may perform various processes using various information. In this regard, examples are listed below.

[2-1. Identifying space delimiters]
In the example of Figure 1, the information processing device 10 may calculate the proportion of times each search query was input within the group, and identify the appropriate position of a space separator in a character string based on the position of the space separator in the search query with the highest proportion.

In addition, the information processing device 10 may identify the position of space separators in the character string based on the search query that has been entered by the largest number of unique users among the search queries classified into the same group.

In addition, the information processing device 10 may identify the position of the space separator in the character string based on the search query that has been entered by the largest number of unique browsers among the search queries classified into the same group.

In addition, the information processing device 10 may identify the position of a space separator in a character string based on the number of times that content indicating search results for each search query classified into the same group was selected. For example, the information processing device 10 identifies the position of a space separator in a character string based on the search query that, among search queries classified into the same group, has the most number of times that content provided to users as search results has been selected (e.g., CTR (Click Through Rate)).

The information processing device 10 may also identify the position of space separators in a character string based on the number of conversions of advertising content corresponding to each search query classified into the same group. For example, the information processing device 10 identifies the position of space separators in a character string based on the search query that, among search queries classified into the same group, has the highest number of times advertising content provided to users as search results has been selected (e.g., CVR (Conversion Rate)).

The information processing device 10 may combine multiple of the above processes to identify appropriate space delimiter positions in a character string. For example, the information processing device 10 may identify appropriate space delimiter positions in a character string based on the number of times each search query classified into the same group was entered, the percentage of times each search query was entered within the group, the number of unique users who entered each search query, the number of unique browsers that entered each search query, the number of times content showing search results for each search query was selected, the number of conversions of advertising content corresponding to each search query, etc.

[2-2. When there are multiple search queries with values equal to or greater than a certain threshold]
In the example of FIG. 1 , if there are multiple search queries classified into the same group that have been entered more than a predetermined threshold number of times (in other words, if there are multiple search queries entered by users at similar frequencies), the information processing device 10 may identify the positions of space separators in the character string based on the search query with the most space separators. For example, if the input frequency threshold is set to "100," group #2 includes the search queries "DDEE FF" and "DD EE FF." In such a case, the information processing device 10 identifies the positions of space separators in the character string "DDEEFF" based on the search query "DD EE FF" that has the most space separator positions.

Note that, when there are multiple search queries classified into the same group that have been entered a predetermined number of times or more, the information processing device 10 may identify the positions of space separators in the character string based on the search query with the fewest number of space separators. For example, of the search queries "DDEE FF" and "DD EE FF," the information processing device 10 may identify the positions of space separators in the character string "DDEEFF" based on the search query "DDEE FF" that has the fewest space separator positions.

When there are multiple search queries classified into the same group whose input counts are equal to or greater than a predetermined threshold, the information processing device 10 may identify the positions of space separators in the character string based on search queries extracted based on the number of times each substring separated by a space is input. For example, the information processing device 10 tallies the input counts of the partial search queries "DD," "EE," and "DDEE" contained in the search queries "DDEE FF" and "DD EE FF" (in other words, the partial search queries corresponding to the parts of the search queries "DDEE FF" and "DD EE FF" where the positions of the space separators differ). If the input count of the partial search query "DDEE" is greater than the input count of the partial search queries "DD" and "EE," the information processing device 10 identifies the positions of the space separators in the character string "DDEEFF" based on the search query "DDEE FF." On the other hand, if the number of times either the partial search queries "DD" or "EE" is entered is greater than the number of times the partial search query "DDEE" is entered, the information processing device 10 identifies the positions of space delimiters in the character string "DDEEFF" based on the search query "DD EE FF."

[2-3. When a single search query can be classified into multiple groups]
In the example of FIG. 1 , if a search query can be classified into multiple groups, the information processing device 10 may classify the search query based on the number of times the search query is input and the number of times each search query classified into the multiple groups is input. For example, if a search query can be classified into multiple groups when part of the search query is converted into synonyms or similar words using a thesaurus, the information processing device 10 classifies the search query into one of the possible groups in which the difference between the number of inputs of the classified search query (e.g., the average number of inputs) and the number of inputs of the search query is less than a predetermined threshold. Furthermore, if a search query can be classified into multiple groups, the information processing device 10 classifies the search query into one of the possible groups in which the difference between the number of inputs of the classified search query and the number of inputs of the search query is equal to or greater than a predetermined threshold.

[2-4. About the model]
In the example of Figure 1, the information processing device 10 may separately generate and train a model (hereinafter, sometimes referred to as "model #1") that outputs a string separated by spaces when a string is input, and a model (hereinafter, sometimes referred to as "model #2") that outputs a string with the space separators removed when a string is input.

In addition, the information processing device 10 may perform training using different training data for models #1 and #2. For example, the information processing device 10 trains model #1 using the positions of space separators in character strings identified based on the number of times each search query is entered in the same group (in other words, the positions of space separators in character strings identified using the number of times the search query is entered as an index) and information indicating the character strings. Then, the information processing device 10 trains model #2 using the positions of space separators in character strings identified using an index (e.g., CTR or CVR) different from the information used to train model #1 and information indicating the character strings.

The information processing device 10 also trains model #2 using the positions of space separators in character strings identified by a method different from that used for model #1 and information indicating the character strings. To give a specific example, the information processing device 10 trains model #2 using the positions of space separators in character strings identified based on the search query for which content provided to users as search results was selected the most frequently among search queries classified into the same group and information indicating the character strings. The information processing device 10 also trains model #2 using the positions of space separators in character strings identified based on the number of conversions of advertising content corresponding to each search query classified into the same group and information indicating the character strings.

[2-5. Generating a string by deleting specific characters]
1 , the information processing device 10 may generate a character string by deleting predetermined characters included in the search query in addition to space separators. For example, the information processing device 10 generates a character string indicated by the search query by deleting space separators and macrons from partial search queries separated by spaces. As a specific example, the information processing device 10 generates the character string "cherry juice in season" by deleting space separators and macrons at the end of the word "cherry" from the search queries "cherry juice in season" and "cherry juice in season."

In addition to the space separators included in the search query, the information processing device 10 may also delete the long note from the search query if deleting the long note included in the search query makes it possible to classify the search query into a predetermined group. For example, if deleting the long note at the end of the word "cherry" included in the search query "cherry juice in season" makes it possible to classify the search query into the group of the character string "cherry juice in season," the information processing device 10 deletes the long note at the end of the word "cherry."

The information processing device 10 may also generate a string by deleting all of the long notes included in the search query, in addition to the space separators included in the search query. For example, the information processing device 10 may delete the space separators and long notes from the search query "cherry juice shun" to generate the string "cherry juice shun." The information processing device 10 may also leave the space separators and generate the string "cherry juice shun."

3. Configuration of information processing device
Next, the configuration of the information processing device 10 will be described with reference to Fig. 2. Fig. 2 is a diagram showing an example of the configuration of the information processing device 10 according to the embodiment. As shown in Fig. 2, the information processing device 10 includes a communication unit 20, a storage unit 30, and a control unit 40.

(Regarding the communication unit 20)
The communication unit 20 is realized by, for example, a network interface card (NIC), etc. The communication unit 20 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the user terminal 100, etc.

(Regarding the storage unit 30)
The storage unit 30 is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. As shown in FIG. 2 , the storage unit 30 has a search query information database 31, a group information database 32, and a model database 33.

(Regarding the search query information database 31)
The search query information database 31 stores various information related to search queries entered by users. An example of information stored in the search query information database 31 will now be described with reference to FIG. 3. FIG. 3 is a diagram showing an example of the search query information database 31 according to an embodiment. In the example of FIG. 3, the search query information database 31 has items such as a "search query ID," a "group ID," a "search query," a "character string," a "number of inputs,""search result information," and "conversion information."

"Search query ID" indicates identification information for identifying a search query. "Group ID" indicates identification information for identifying a group. "Search query" indicates information (such as text information) that indicates the search query. "String" indicates a string generated from the search query. "Number of inputs" indicates the number of times the search query was input. "Search result information" indicates information about the search results when a search query was input, and stores information indicating, for example, the number of times content input as a search result was selected. "Conversion information" indicates information about conversions of advertising content corresponding to the search query.

In other words, Figure 3 shows an example in which a search query identified by search query ID "QID#1" is classified into a group identified by group ID "GID#1", the search query is "search query #1", the character string is "character string #1", the number of inputs is "number of inputs #1", the search result information is "search result information #1", and the conversion information is "conversion information #1".

Note that the information stored in the search query information database 31 is not limited to the above, and various types of information may be stored. For example, the search query information database 31 may store information indicating the number of unique users who have input search queries, the number of unique browsers that have input search queries, etc.

(Regarding the group information database 32)
The group information database 32 stores various information related to groups into which search queries are classified. An example of information stored in the group information database 32 will now be described with reference to FIG. 4. FIG. 4 is a diagram illustrating an example of the group information database 32 according to an embodiment. In the example of FIG. 4, the group information database 32 has items such as "group ID,""search query information,""characterstring," and "space delimiter information."

"Group ID" indicates identification information for identifying a group. "Search query information" indicates information about the search queries included in the group (such as the search query ID). "String" indicates the string generated from the search queries classified into the group. "Space separator information" indicates the position of the space separator in the string.

In other words, Figure 4 shows an example in which the search query information for a group classified by group ID "GID#1" is "search query information #1," the character string is "character string #1," and the space separator information is "space separator information #1."

(Regarding the model database 33)
The model database 33 stores a model that, when a character string is input, outputs a character string with a space delimiter added or deleted.

(Regarding the control unit 40)
The control unit 40 is a controller, and is realized, for example, by a CPU (Central Processing Unit) or an MPU (Micro Processing Unit) executing various programs stored in a storage device inside the information processing device 10 using RAM as a work area. The control unit 40 is also a controller, and is realized, for example, by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). As shown in FIG. 3 , the control unit 40 according to the embodiment has a generation unit 41, a classification unit 42, an identification unit 43, and a learning unit 44, and realizes or executes the functions and actions of information processing described below.

(Regarding the generation unit 41)
The generation unit 41 generates a character string indicated by the search query entered by the user by deleting space separators included in the search query. For example, in the example of Fig. 1 , the generation unit 41 deletes the space separator between "AA" and "BB" and the space separator between "BB" and "CC" in the search query "AA BB CC" to generate a character string "AABBCC" from the search query "AA BB CC" and stores the character string in the search query information database 31.

The generation unit 41 may also generate a character string indicated by the search query by deleting predetermined characters included in the search query. For example, in the example of FIG. 1, the generation unit 41 generates a character string indicated by the search query by deleting space separators and macrons at the ends of partial search queries separated by spaces.

(Regarding the classification unit 42)
The classification unit 42 classifies the search queries based on the character strings generated by the generation unit 41. For example, in the example of FIG. 1 , the character string generated from each of the search queries "AABBCC,""AA BB CC," and "AA BBCC" is "AABBCC," so the classification unit 42 classifies these search queries into group #1. As a specific example, the classification unit 42 associates the search queries "AABBCC,""AA BB CC," and "AA BBCC" with group #1 and stores them in the group information database 32.

Furthermore, when a search query can be classified into multiple groups, the classification unit 42 may classify the search query based on the number of times the search query is input and the number of times each search query classified into the multiple groups is input. For example, in the example of FIG. 1, when a thesaurus is used to convert part of the search query into synonyms or similar words, if the search query can be classified into multiple groups, the classification unit 42 classifies the search query into one of the possible groups into which the difference between the number of times the search query is input and the number of times the search query is input is less than a predetermined threshold. Furthermore, when a search query can be classified into multiple groups, the classification unit 42 classifies the search query into one of the possible groups into which the difference between the number of times the search query is input and the number of times the search query is input is equal to or greater than a predetermined threshold.

(Regarding the identification unit 43)
The identification unit 43 identifies the positions of space separators in the character string generated from each search query classified into the same group based on the number of times each search query has been input. For example, in the example of Fig. 1, the identification unit 43 identifies that the appropriate positions of space separators in the character string "AABBCC" are between "AA" and "BB" and between "BB" and "CC" because the search query "AA BB CC" has been input the most in group #1.

The identification unit 43 may also identify the position of space separators in the character string generated from each search query based on the proportion of search queries that are input within the same group, that is, the proportion within the same group. For example, in the example of FIG. 1, the identification unit 43 calculates the proportion of search queries that are input within the group, and identifies the appropriate position of space separators in the character string based on the position of space separators in the search query with the highest proportion.

The identification unit 43 may also identify the position of space separators in the character string generated from each search query classified into the same group, based on the number of browsers that input each search query. For example, in the example of FIG. 1, the identification unit 43 identifies the position of space separators in the character string, based on the search query that has been input by the largest number of unique browsers, among the search queries classified into the same group.

The identification unit 43 may further identify the position of space separators in the character string generated from each search query based on the number of times that content representing search results for each search query classified in the same group was selected. For example, in the example of FIG. 1, the identification unit 43 identifies the position of space separators in the character string based on the search query for which content provided to users as search results was selected the most frequently, among the search queries classified in the same group.

The identification unit 43 may further identify the position of space separators in the character string generated from each search query based on the number of conversions of advertising content corresponding to each search query classified into the same group. For example, in the example of FIG. 1, the identification unit 43 identifies the position of space separators in the character string based on the search query for which advertising content provided to users as search results has been selected the most frequently, among the search queries classified into the same group.

The identification unit 43 may also identify the positions of space separators in the character string generated from each search query based on the positions of space separators in the search query that has been entered most frequently among the search queries classified into the same group. For example, in the example of FIG. 1, the identification unit 43 identifies that the appropriate positions of space separators in the character string "AABBCC" are between "AA" and "BB" and between "BB" and "CC" because the search query "AA BB CC" has been entered the most frequently in group #1.

Furthermore, when there are multiple search queries classified into the same group that have been input a predetermined number of times or more, the identification unit 43 may identify the positions of space separators in the character string generated from each search query based on the search query with the greatest number of space separators among the search queries that have been input a predetermined number of times or more. For example, in the example of FIG. 1, if the input count threshold is set to "100," group #2 would include the search queries "DDEE FF" and "DD EE FF." In such a case, the identification unit 43 identifies the positions of space separators in the character string "DDEEFF" based on the search query "DD EE FF" that has the greatest number of space separator positions.

Furthermore, when there are multiple search queries classified into the same group that have been input a predetermined number of times or more, the identification unit 43 may identify the positions of space separators in the character string generated from each search query based on the search query with the fewest number of space separators among the search queries that have been input a predetermined number of times or more. For example, in the example of FIG. 1, if the input count threshold is set to "100," group #2 corresponds to the search queries "DDEE FF" and "DD EE FF." In such a case, the identification unit 43 identifies the positions of space separators in the character string "DDEEFF" based on the search query "DDEE FF" that has the fewest space separators among the search queries "DDEE FF" and "DD EE FF."

Furthermore, when there are multiple search queries classified into the same group that have been input a predetermined number of times or more, the identification unit 43 may identify the position of space separators in the character strings generated from each search query based on search queries extracted based on the number of times each substring separated by a space was input from the search queries that have been input a predetermined number of times or more. For example, in the example of Figure 1, if the input count threshold is set to "100," group #2 corresponds to the search queries "DDEE FF" and "DD EE FF." In such a case, the identification unit 43 tallies the input counts of the partial search queries "DD," "EE," and "DDEE" included in the search queries "DDEE FF" and "DD EE FF," respectively. If the number of times the partial search query "DDEE" is entered is greater than the number of times each of the partial search queries "DD" and "EE" is entered, the information processing device 10 identifies the positions of space separators in the character string "DDEEFF" based on the search query "DDEE FF". On the other hand, if the number of times either the partial search queries "DD" or "EE" is entered is greater than the number of times the partial search query "DDEE" is entered, the information processing device 10 identifies the positions of space separators in the character string "DDEEFF" based on the search query "DD EE FF".

(Regarding the learning unit 44)
The learning unit 44 learns a model that outputs a character string separated by spaces when a character string is input, based on the positions of space separators in the character string identified by the identification unit 43. For example, in the example of Fig. 1, the learning unit 44 learns a model that outputs a character string separated by spaces when a character string is input.

The learning unit 44 may also train the model based on information regarding search results using the character string output by the model. For example, in the example of FIG. 1, if the behavioral information of user U3 indicates that user U3 selected content provided as a search result based on "GGHH II", the learning unit 44 trains the model using the edited character string "GGHH II" of the character string "GGHHII" as a positive example. Also, if the behavioral information of user U3 indicates that user U3 did not select content provided as a search result based on "GGHH II", the learning unit 44 trains the model using the edited character string "GGHHII" of the character string "GGHH II" as a negative example.

The learning unit 44 may also train a model that, when a string is input, outputs a string from which space separators have been deleted, based on the positions of space separators in the string identified by the identification unit 43. For example, in the example of Figure 1, the learning unit 44 trains a model that, when a string is input, outputs a string from which space separators have been deleted.

[4. Information processing flow]
The procedure of information processing by the information processing apparatus 10 according to the embodiment will be described with reference to Fig. 5. Fig. 5 is a flowchart showing an example of the procedure of information processing according to the embodiment.

As shown in FIG. 5, the information processing device 10 generates a character string indicated by a search query entered by a user by deleting space separators included in the search query (step S101). Next, the information processing device 10 classifies the search queries based on the generated character strings (step S102). Next, the information processing device 10 identifies the positions of space separators in the character strings generated from each search query based on the number of times each search query classified into the same group has been entered (step S103). Next, the information processing device 10 determines whether a search query has been received from the user (step S104). If a search query has not been received (step S104; No), the information processing device 10 waits until a search query is received.

On the other hand, if a search query has been received (step S104; Yes), search results are provided based on the identified space separator position and the received search query (step S105), and the process ends.

5. Modifications
The above-described embodiment is merely an example, and various modifications and applications are possible.

[5-1. Regarding the period when the search query was entered]
In the above-described embodiment, the information processing device 10 may perform the above-described process using search queries input by a user in a predetermined period (e.g., the last month, the last six months, etc.). Note that if the amount of search queries input by a user in a predetermined period does not reach a predetermined amount, the information processing device 10 may extend the period (e.g., from the last month to the last three months, or from the last six months to the last year, etc.) so that the amount of search queries input by the user reaches the predetermined amount.

[5-2. Generating strings]
The method by which the information processing device 10 generates a character string from a search query is not limited to the above-described method, and any method may be used. For example, if a group of the character string "BBCCAA" exists, the information processing device 10 may delete the space separator from the search query "AA BB CC" and change the order of the partial search queries to generate the character string "BBCCAA" from the search query.

5-3. Processing Mode
Of the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, and conversely, all or part of the processes described as being performed manually can be performed automatically using known methods. In addition, the information, including the processing procedures, specific names, various data, and parameters shown in the above text and drawings, can be changed as desired unless otherwise specified. For example, the various information shown in each drawing is not limited to the information shown in the drawings.

Furthermore, the components of each device shown in the figure are functional concepts and do not necessarily have to be physically configured as shown. In other words, the specific form of distribution and integration of each device is not limited to that shown, and all or part of them can be functionally or physically distributed and integrated in any unit depending on various loads, usage conditions, etc.

Furthermore, the above-mentioned embodiments can be combined as appropriate to the extent that the processing content is not contradictory.

The information processing device 10 may combine the above-described process with various other processes to build a model for spelling correction. For example, the information processing device 10 acquires spelling correction data such as "shop YY port town" as an uncorrected string and "shop YY port town" as a corrected string. This process is performed, for example, by comparing the log of the input search query with a pre-created database and extracting pairs of strings that are similar enough to or exceed a predetermined threshold in terms of the number of different characters, pronunciation, number of adjacent keyboard keys used during input, etc., as the uncorrected string and the corrected string. The information processing device 10 adds space separators to such spelling correction data using the above-described model. For example, the information processing device 10 may convert the corrected string "shop YY port town" in the spelling correction data into data such as "shop YY port town" and train a correction model that corrects "shop YY port town" to a string such as "shop YY port town." That is, the information processing device 10 may use the generated model to input a space delimiter into the corrected string used to train the correction model, and generate a correction model that corrects the pre-correction string to the corrected string.

6. Effects
As described above, the information processing device 10 according to the embodiment includes a generation unit 41, a classification unit 42, an identification unit 43, and a learning unit 44. The generation unit 41 generates a character string indicated by a search query entered by a user by deleting space separators included in the search query. The generation unit 41 also generates a character string indicated by the search query by deleting predetermined characters included in the search query. The classification unit 42 classifies the search query based on the character string generated by the generation unit 41. If a search query can be classified into multiple groups, the classification unit 42 classifies the search query based on the number of times the search query was entered and the number of times each search query classified into the multiple groups was entered. The identification unit 43 identifies the position of a space separator in a character string generated from each search query based on the number of times each search query classified into the same group was entered. The identification unit 43 also identifies the position of a space separator in a character string generated from each search query based on the position of a space separator in the search query that has been entered the most frequently among the search queries classified into the same group. The learning unit 44 trains a model that outputs a character string separated by spaces when a character string is input, based on the positions of space separators in the character string identified by the identification unit 43. The learning unit 44 also trains a model that outputs a character string from which the space separators have been deleted when a character string is input, based on the positions of space separators in the character string identified by the identification unit 43. The learning unit 44 also trains the model based on information related to search results using the character string output by the model.

As a result, the information processing device 10 according to the embodiment can identify the position of space separators in the character strings indicated by each search query based on the search query most frequently entered by users among the search queries classified into the same group, thereby identifying the appropriate position of space separators in the search query.

In the information processing device 10 according to the embodiment, for example, the identification unit 43 identifies the position of space separators in the character string generated from each search query based on the proportion of search queries entered within the same group. The identification unit 43 also identifies the position of space separators in the character string generated from each search query based on the number of browsers that entered each search query entered within the same group. The identification unit 43 further identifies the position of space separators in the character string generated from each search query based on the number of times that content indicating search results for each search query entered within the same group was selected. The identification unit 43 also identifies the position of space separators in the character string generated from each search query based on the number of conversions of advertising content corresponding to each search query entered within the same group. When there are multiple search queries entered within the same group that have been entered a predetermined number of times or more, the identification unit 43 identifies the position of space separators in the character string generated from each search query based on the search query with the greatest number of space separators among the search queries entered a predetermined number of times or more. Furthermore, when there are multiple search queries that have been input a predetermined number of times or more, among the search queries classified into the same group, the identification unit 43 identifies the positions of space separators in the character string generated from each search query based on the search query with the fewest number of space separators among the search queries that have been input a predetermined number of times or more. Furthermore, when there are multiple search queries that have been input a predetermined number of times or more, among the search queries classified into the same group, the identification unit 43 identifies the positions of space separators in the character string generated from each search query based on the search queries extracted based on the number of times each substring separated by a space was input, among the search queries that have been input a predetermined number of times or more.

As a result, the information processing device 10 according to the embodiment can use various indices to identify the positions of space separators in a search query, thereby improving the accuracy of identifying the positions of space separators.

7. Hardware Configuration
The information processing device 10 according to each of the above-described embodiments is realized, for example, by a computer 1000 configured as shown in Fig. 6. The following description will be given using the information processing device 10 as an example. Fig. 6 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing device 10. The computer 1000 has a CPU 1100, a ROM 1200, a RAM 1300, a HDD 1400, a communication interface (I/F) 1500, an input/output interface (I/F) 1600, and a media interface (I/F) 1700.

The CPU 1100 operates based on programs stored in the ROM 1200 or HDD 1400, and controls each component. The ROM 1200 stores a boot program executed by the CPU 1100 when the computer 1000 starts up, as well as programs that depend on the computer 1000's hardware.

HDD 1400 stores programs executed by CPU 1100 and data used by such programs. Communication interface 1500 receives data from other devices via communication network 500 (corresponding to network N in the embodiment) and sends it to CPU 1100, and also transmits data generated by CPU 1100 to other devices via communication network 500.

The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, via the input/output interface 1600. The CPU 1100 acquires data from input devices via the input/output interface 1600. The CPU 1100 also outputs data generated via the input/output interface 1600 to output devices.

Media interface 1700 reads programs or data stored on recording medium 1800 and provides them to CPU 1100 via RAM 1300. CPU 1100 loads the programs from recording medium 1800 onto RAM 1300 via media interface 1700 and executes the loaded programs. Recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase Change Rewritable Disc), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

For example, when the computer 1000 functions as the information processing device 10, the CPU 1100 of the computer 1000 executes programs loaded onto the RAM 1300 to realize the functions of the control unit 40. In addition, the HDD 1400 stores various data within the storage device of the information processing device 10. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be obtained from another device via a specified communications network.

[8. Other]
Although some of the embodiments of the present application have been described in detail above with reference to the drawings, these are merely examples, and the present invention can be implemented in other forms that include the embodiments described in the Disclosure of the Invention section and that have been modified and improved in various ways based on the knowledge of those skilled in the art.

Furthermore, the configuration of the information processing device 10 described above can be flexibly changed, such as by calling external platforms using APIs (Application Programming Interfaces) or network computing, depending on the function.

Furthermore, the term "unit" in the claims can be read as "means" or "circuit." For example, a generation unit can be read as a generation means or a generation circuit.

REFERENCE SIGNS LIST 10 Information processing device 20 Communication unit 30 Storage unit 31 Search query information database 32 Group information database 33 Model database 40 Control unit 41 Generation unit 42 Classification unit 43 Identification unit 44 Learning unit 100 User terminal

Claims (15)

a generation unit that generates a character string indicated by a search query input by a user by deleting space separators included in the search query;
a classification unit that classifies the search queries based on the character strings generated by the generation unit;
an identification unit that identifies the position of a space delimiter in the character string generated from each search query classified into the same group based on the number of times each search query has been input ,
The identification unit
Furthermore, the position of the space delimiter in the character string generated from each search query is identified based on the number of times that content representing search results for each search query classified into the same group has been selected.
1. An information processing device comprising: The identification unit
The information processing device according to claim 1, characterized in that the position of a space separator in the character string generated from each search query is identified based on the ratio of input search queries classified into the same group, the ratio within the same group. The identification unit
The information processing device according to claim 1 , further comprising: identifying a position of a space delimiter in the character string generated from each search query classified into the same group based on the number of browsers that input each search query. The identification unit
The information processing device according to claim 1 , further comprising: identifying a position of a space delimiter in the character string generated from each search query based on the number of conversions of advertising content corresponding to each search query classified into the same group. The identification unit
The information processing device according to claim 1, characterized in that the position of a space separator in the character string generated from each search query is identified based on the position of a space separator in the search query that has been entered most frequently among the search queries classified into the same group. The identification unit
2. The information processing device according to claim 1, wherein, when there are multiple search queries among the search queries classified into the same group that have been input a number of times equal to or greater than a predetermined threshold, the position of space separators in the character strings generated from each of the search queries is identified based on the search query that has the largest number of space separators among the search queries that have been input a number of times equal to or greater than the predetermined threshold. The identification unit
2. The information processing device according to claim 1, wherein, when there are multiple search queries that have been entered a predetermined number of times or more among the search queries classified into the same group, the position of space separators in the character strings generated from each of the search queries is identified based on the search query that has the fewest number of space separators among the search queries that have been entered a predetermined number of times or more. The identification unit
2. The information processing device according to claim 1, wherein, when there are multiple search queries among the search queries classified into the same group that have been input a predetermined threshold or more in number of times, the information processing device identifies the position of space separators in the character strings generated from each of the search queries based on search queries extracted based on the number of times each substring separated by a space has been input from the search queries that have been input a predetermined threshold or more in number of times. 2. The information processing device according to claim 1, further comprising: a first learning unit configured to learn a model that outputs a character string separated by spaces when a character string is input, based on the positions of space separators in the character string identified by the identification unit. The first learning unit
The information processing apparatus according to claim 9 , further comprising: learning the model based on information about a search result using a character string output by the model. 2. The information processing device according to claim 1, further comprising: a second learning unit configured to learn a model that outputs a character string from which space separators have been deleted when a character string is input, based on the positions of space separators in the character string identified by the identification unit. The second learning unit
The information processing apparatus according to claim 11 , further comprising: learning the model based on information about a search result using a character string output by the model. The generation unit
The information processing device according to claim 1 , further comprising: a processor configured to generate a character string indicated by a search query by deleting predetermined characters included in the search query. 1. A computer-implemented information processing method, comprising:
a generating step of generating a character string indicated by the search query by deleting space separators included in the search query entered by the user;
a classification step of classifying the search queries based on the character strings generated by the generation step;
and a specifying step of specifying the positions of space delimiters in the character strings generated from each of the search queries classified into the same group based on the number of times each of the search queries has been input,
The identifying step includes:
Furthermore, the position of the space delimiter in the character string generated from each search query is identified based on the number of times that content representing search results for each search query classified into the same group has been selected.
1. An information processing method comprising: a generation step of generating a character string indicated by a search query input by a user by deleting space separators included in the search query;
a classification step of classifying the search queries based on the character strings generated by the generation step;
and a step of identifying the positions of space delimiters in the character strings generated from each search query classified into the same group based on the number of times each search query has been input;
The identification procedure includes:
Furthermore, the position of the space delimiter in the character string generated from each search query is identified based on the number of times that content representing search results for each search query classified into the same group has been selected.
An information processing program characterized by :