JP7615546B2 - KEYWORD RECOMMENDATION PROGRAM, KEYWORD RECOMMENDATION METHOD, AND INFORMATION PROCESSING APPARATUS - Google Patents

Description

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

In recent years, meetups have been held online as events aimed at meeting and interacting with people who share the same goals. As the number of participants in meetups has increased and they have become larger in scale, it has become possible to interact with people you have never met before.

However, when there are a large number of participants, the number of introductions for each participant becomes enormous, making it difficult to determine which participants to meet. One idea is to connect participants to each other based on the words contained in the introductions written by each meetup participant.

There are known techniques for creating software deliverables with uniform task granularity by using the same subject vocabulary, action vocabulary, and parameter vocabulary contained in software deliverables by different creators. In addition, a technique has been proposed that uses keywords that indicate vague relationships and thoughts unique to the searcher to realize information searches that match the searcher's intentions.

JP 2008-112246 A JP 2007-018350 A

The introductions of meetup participants vary in the words and phrases they use. Sometimes they are written using specific words, and sometimes they are written abstractly using general words.

If the level of detail in the words used in the introduction varies, it is not possible to accurately match participants with the same purpose. This is because many participants are not good at writing introductions.

So, one aspect is to create appropriate documentation.

According to one aspect, a keyword recommendation program is provided that causes a computer to perform the following process: extract a first keyword from a first sentence of a first user received from a terminal of the first user; based on data defining a semantic inclusion relationship between multiple keywords, identify a candidate keyword from second keywords extracted from second sentences of other users that, instead of the first keyword, more specifically expresses an object represented by the first keyword; and transmit the candidate keyword to the terminal of the first user.

Able to create appropriate documentation.

FIG. 1 illustrates an example of a network configuration of a system according to an embodiment of the present invention. FIG. 2 illustrates an example of a hardware configuration according to the present embodiment. FIG. 2 is a diagram illustrating an example of a functional configuration of an information processing device according to the present embodiment. FIG. 2 is a sequence diagram for explaining an overview of processing in the system according to the present embodiment. FIG. 11 is a diagram showing a time series of keyword detection states. FIG. 11 is a diagram illustrating an example of ontology information. FIG. 13 is a diagram illustrating an example of updating a keyword DB. FIG. 13 is a diagram showing an example of an input screen for inputting an introduction message and displayed on the terminal in the present embodiment. FIG. 4 is a diagram showing a specific example of input on a terminal in the present embodiment. 11 is a flowchart illustrating an overall keyword recommendation process. FIG. 11 is a flowchart illustrating an extraction process. FIG. 11 is a flowchart for explaining a determination process. FIG. 11 is a flowchart illustrating a recommendation process. FIG. 11 is a sequence diagram when the number of users is increased midway in the system according to the embodiment. 15 is a diagram showing an example of data transition of a keyword DB in the case of the operation example of FIG. 14; FIG. 13 is a diagram showing an example of a screen when creation support information is provided by push notification.

The following describes an embodiment of the present invention with reference to the drawings. In this embodiment, among the multiple words included in an introduction, a word used for matching is called a keyword. The level of detail with which the keyword expresses the object represented is called granularity. The participants of a meetup are called users. In this embodiment, a keyword recommendation process is disclosed below that recommends keywords when participants write introductions so that the granularity of keywords is unified in similar categories. The keyword recommendation process in this embodiment can be realized by an information processing device in a system such as that shown in FIG. 1.

Figure 1 is a diagram showing an example of the network configuration of a system in this embodiment. The system 1000 shown in Figure 1 has an information processing device 100 and multiple terminals 3. The information processing device 100 is a computer that provides a meetup service over the Internet 2 and executes a program that realizes the keyword recommendation process in this embodiment.

The multiple terminals 3 are terminals used by users 1 who participate in the meetup by connecting to the information processing device 100 via the Internet 2. The terminals 3 may be mobile terminals or standalone computers.

In the system 1000, after the terminal 3 is connected to the information processing device 100, when the user 1 inputs an introduction 3a to the terminal 3, the introduction 3a being input is transmitted to the information processing device 100 as needed. An example of the introduction 3a is a text in which the user 1 introduces himself/herself (self-introduction), but this embodiment is not limited to this example.

The information processing device 100 transmits creation support information 3b to the terminal 3, which includes more detailed keyword candidates to replace the input words, in order to standardize the granularity of the keywords included in the introductions 3 transmitted from each terminal 3.

User 1 of terminal 3 edits introduction 3a to include candidate keywords displayed on the screen during input. By editing introduction 3a using the keywords displayed on the screen, user 1 can roughly standardize the granularity of keywords used for matching between introductions 3a of other users 1.

The creation support information 3b sent from the information processing device 100 to the terminal 3 includes one or more keywords, a message encouraging the use of the keywords, and information indicating words recommended for use as specific expressions in the introduction text 3a being input or their positions within the introduction text 3a.

Candidates based on the creation support information 3b are displayed on the screen of the terminal 3 while the introduction 3a is being entered, prompting the user 1 to enter specific content. The method of notifying the candidates is not limited to displaying them on the screen during entry. The terminal 3 may receive the creation support information 3b by push notification. In this case, when the user 1 selects the notification on the terminal 3, the input screen for the introduction 3a is displayed, and the introduction 3a entered by the user 1, as well as keywords and a message, are displayed.

Fig. 2 is a diagram showing an example of a hardware configuration in this embodiment. In Fig. 2, the information processing device 100 is a computer having a CPU 11, a main memory device 12, an auxiliary memory device 13, an input device 14, a display device 15, a communication I/F 17, and a drive device 18, and is connected to a bus B1. The main memory device 12, the auxiliary memory device 13, and an external storage device accessible by the information processing device 100 correspond to the storage unit 130 in the functional block diagram (Fig. 3) described later.

The CPU 11 corresponds to a processor that controls the information processing device 100, and executes programs stored in the storage unit 130 to realize various processes according to the present embodiment described below. The input device 14 is operated by a user to input data in response to the operation, and the display device 15 displays various screens as a user interface. The communication I/F 17 controls communication with external devices.

The keyword recommendation program according to this embodiment stored in a storage medium 19 (e.g., a CD-ROM (Compact Disc Read-Only Memory)) is installed in the storage unit 130 via the drive device 18 and becomes executable by the CPU 11.

The storage medium 19 that stores the program according to this embodiment is not limited to a CD-ROM, but may be one or more non-transitory, tangible media that have a structure that can be read by a computer. In addition to a CD-ROM, the computer-readable storage medium may be a DVD (Digital Versatile Disk), a portable recording medium such as a USB memory, or a semiconductor memory such as a flash memory.

In FIG. 2, terminal 3 is an information processing terminal such as a tablet or mobile phone controlled by a computer. Terminal 3 has a CPU 311, a main memory device 312, a user I/F 316, a communication I/F 317, and a drive device 318, and is connected to bus B3.

The CPU 311 corresponds to a processor that controls the terminal 3, and executes programs stored in the main memory device 312 or the like to realize various processes related to this embodiment described below. The user I/F 316 displays various necessary information under the control of the CPU 311, and is a touch panel or the like that allows the user to input operations. Communication via the communication I/F 317 is not limited to being wireless or wired.

The drive device 318 provides an interface between the terminal 3 and a storage medium 319 (e.g., a Secure Digital (SD) memory card, etc.) set in the drive device 318. The storage medium 319 may be one or more non-transitory, tangible media having a structure that can be read by a computer.

The terminal 3 may be an information processing terminal of a desktop type, a notebook type, a laptop type, etc., and the hardware configuration thereof is similar to that of FIG. 2, so a description thereof will be omitted.

Figure 3 is a diagram showing an example of the functional configuration of an information processing device in this embodiment. In Figure 3, the processing units related to this embodiment are shown, and other processing units are omitted. As an example, a processing unit that provides a meetup service is omitted.

The information processing device 100 in FIG. 3 has a keyword recommendation unit 40. In response to receiving an introduction 3a being input to a terminal 3 connected via the Internet 2, the keyword recommendation unit 40 generates creation support information 3b and transmits it to the terminal 3 if the introduction 3a contains coarse-grained (highly abstract) words. The keyword recommendation unit 40 further has an extraction unit 41, a determination unit 42, and a recommendation unit 43. Each of the processing units 41, 42, and 43 is realized by a process executed by the CPU 11 of the information processing device 100 by a program installed in the information processing device 100.

The memory unit 130 also stores an introduction DB 51, a keyword DB 52, ontology information 53, etc.

In response to receiving an introduction 3a for each user 1 of the terminal 3, the extraction unit 41 stores the introduction 3a in the introduction DB 51 and extracts keywords 3d from the introduction 3a. The extraction unit 41 stores the extracted keywords 3d and a user ID that identifies the user 1 in the keyword DB 52. When extracting keywords 3d, the extraction unit 41 extracts keywords 3d from the introduction 3a being input by the user 1 at any time. When storing keywords 3d in the keyword DB 52, the extraction unit 41 adds new keywords 3d that have not yet been stored to the keyword DB 52 so as not to duplicate keywords 3d.

In response to the addition of a keyword 3d to the keyword DB 52, the determination unit 42 determines the granularity relationship with the keyword 3d stored in the keyword DB 52 that is associated with another user 1, based on the ontology information 53. It is not limited to the ontology information 53, and any data that specifies a semantic inclusion relationship may be used.

When the recommendation unit 43 detects the addition of a keyword 3d to the keyword DB 52, it refers to the ontology information 53 and acquires from the keyword DB 52 a keyword 3d' that is associated with another user ID and is a word with a finer granularity than the added keyword 3d. The recommendation unit 43 generates creation support information 3b using the acquired keyword 3d' and transmits the generated creation support information 3b to the user terminal 3 used by the user 1 whose user ID is associated with the added keyword 3d.

In this embodiment, among words with finer granularity than the keyword 3d added to the keyword DB 52, keywords 3d' included in the introduction 3a of other users 1 are displayed as candidates on the terminal 3 of the user 1 who input the coarser granularity keyword 3d.

When words with finer granularity than the keywords 3d obtained by referencing the ontology information 53 are displayed as candidates on the terminal 3 of the user 1 who has entered the coarse keywords 3d, the number of candidates may be too large. The user 1 feels that it is time-consuming to determine which words to use from too many candidates and create a document, and so tends not to be specific about the keywords 3d in the introduction 3a that he is entering.

Figure 4 is a sequence diagram for explaining an overview of the processing in the system in this embodiment. In Figure 4, a case will be explained in which users u1, u2, and u3 connect to the information processing device 100 using their respective terminals 3 and create an introduction 3a for participating in a meetup. In this embodiment, the number of users 1 is not limited to three.

Each of users u1, u2, and u3 uses terminal 3 to input a self-introduction including the purpose of participating in the meetup or categories of interest on an input screen provided by information processing device 100 (step S101).

In the information processing device 100, when the extraction unit 41 receives the introduction 3a from each of the users u1 to u3, it extracts keywords 3d from each introduction 3a (step S102). The extracted keywords 3d are stored in the keyword DB 52.

When the determination unit 42 detects the addition of a keyword 3d to the keyword DB 52 at regular intervals using a timer or the like, it refers to the ontology information 53 and determines the granularity of the detected keyword 3d (step S103). If it determines that the granularity of the detected keyword 3d is coarse, the determination unit 42 notifies the recommendation unit 43 of the determined keyword 3d' and the user ID of the user u1. It is assumed that the keyword 3d included in the introduction 3a of the user u1 is detected and the granularity is determined.

The recommendation unit 43 generates creation support information 3b and transmits it to the terminal 3 of user u1 (step S104). The recommendation unit 43 extracts words (hereinafter, referred to as refined words) that are finer than the granularity of the keyword 3d from the ontology information 53. Then, the recommendation unit 43 determines, from among the refined words obtained by extraction, the keyword 3d' obtained from the introductions 3a of other users u2 and u3 that exist in the keyword DB 52, as a recommended candidate. The recommendation unit 43 generates creation support information 3b including information such as the keyword 3d', the message, and the position of the keyword 3d in the introduction 3a of user u1, and transmits the generated creation support information 3b to the terminal 3 of user u1.

The user u1 edits the introduction 3a being input by referring to the keywords 3d' and the message displayed on the terminal 3 based on the creation support information 3b (step S105). The edited introduction 3a is sent to the information processing device 100, and the same process as described above is repeated until the user u1 closes the input screen.

Figure 5 is a diagram showing the state in which keywords are detected in a time series. In Figure 5, at time T1, keywords kw1, kw2, and kw3 are detected from the introduction 3a of user u1 and stored in keyword DB 52.

After that, at time T2, the keywords kw1a and kw1b are detected from the introduction 3a of user u2. At almost the same time, the keyword kw3b is detected from the introduction 3a of user u3. At time T3, a little after time T2, the keyword kw3ca is detected from the introduction 3a of user u3.

In this example, keywords kw1, kw2, and kw3 are the most abstract keywords 3d, and the more alphabetic characters in the subscript, the more specific the keywords 3d become. As an example, keywords kw1a and kw1b are vocabulary that more specifically express the meaning of keyword kw1. Keyword kw3b is vocabulary that more specifically expresses the meaning of keyword kw3. Keyword kw3ca is vocabulary that more specifically expresses the meaning of keyword kw3. The semantic relationships between such keywords 3d are represented by ontology information 53 as shown in FIG. 6.

Figure 6 is a diagram showing an example of ontology information. In Figure 6, ontology information 53 is represented as a tree structure with a base as the root, and various abstract vocabulary branches from the base. In other words, it branches from the base to keywords kw1, kw2, kw3, etc. located at the first level.

At the second level, keyword kw1 located at the first level branches into keywords kw1a, kw1b, and kw1c. Keyword kw2 located at the first level branches into keywords kw2a, kw2b, and kw2c. Keyword kw3 located at the first level branches into keywords kw3a, kw3b, and kw3c.

At the third level, keyword kw3a, located at the second level, is refined to keyword kw3aa. At the fourth level, keyword kw3aa, located at the third level, is refined to keyword kw3ab, and at the fifth level, keyword kw3ac is refined.

Similarly, the keyword kw3c located at the second level is refined to keyword kw3ca at the third level, then to keyword kw3cb at the fourth level, and then to keyword kw3cc at the fifth level. Such ontology information 53 is referenced by the recommendation unit 43.

Figure 7 is a diagram showing an example of updating the keyword DB. In Figure 7, the left side shows an example of data in the keyword DB 52 when the keywords are extracted in Figure 5. The keywords kw1, kw2, and kw3 extracted from the introduction 3a of user u1 are stored in association with the user ID "u1". The keywords kw1a and kw1b extracted from the introduction 3a of user u2 are also stored in association with the user ID "u2". Furthermore, the keywords kw3b and kw3ca extracted from the introduction 3a of user u3 are also stored in association with the user ID "u2".

When the ontology information 53 is referenced, it is determined that the granularity of the keyword kw1 of the user u1 is coarser than the keywords kw1a and kw1b of the user u2, and that the granularity of meaning is not uniform between the users. The keyword kw1 of the user u1 is notified to the recommendation unit 43 from the determination unit 42.

After that, the user u1 adds a more specific sentence based on the creation support information 3b generated by the recommendation unit 43, and the keyword DB 52 is updated. As a result of the update, a record 5a indicating the keyword kw1a associated with the user ID "u1" is added to the keyword DB 52.

Next, an example of the screen operated by user 1 will be described with reference to Figs. 8 and 9. Figs. 8 and 9 use an example in which user u1 in Fig. 5 is inputting an introductory text 3a.

Figure 8 is a diagram showing an example of an input screen for inputting an introductory text displayed on a terminal in this embodiment. In Figure 8, on the screen on which user u1 inputs "I'm interested in kw1. I'm also interested in kw2 and kw3," a message 4m such as "Why don't you specify kw1 to get other people more interested?" is displayed. Within message 4m, keywords kw1a and kw1b correspond to keyword 3d' selected by recommendation unit 43.

The message 4m area also points to the relevant keyword kw1 in the introduction 3a, and prompts the user to write the specific meaning of the keyword kw1. The message 4 shows selectable keywords kw1a and kw1b. If the user u1 selects the keyword kw1a, a standard sentence 7a such as "Specifically, regarding kw1a, . . . " is inserted. The user u1 completes the ". . . " portion. This operation adds a record of the keyword "kw1a" associated with the user ID "u1" to the keyword DB 52.

Figure 9 is a diagram showing a specific example of input on a terminal in this embodiment. In Figure 9, an input example of user 1 will be described. When user 1 inputs "I'm interested in vegetables. I'm also interested in meat and fish" on the input screen of terminal 3, a message 4m such as "Why don't you specify the vegetables so that other people will be more interested?" from recommendation unit 43 is displayed. Message 4m points to "vegetables" in the input sentence. Also, message 4m displays "tomato" and "onion" as keywords 3d' selected by recommendation unit 43 so that user 1 can select them.

When user 1 selects "tomatoes", a standard phrase 7b such as "Specifically, regarding tomatoes, ____" is inserted immediately after the sentence containing "vegetables". User 1 replaces the "____" portion of this inserted standard phrase with a sentence related to "tomatoes", such as "I'm growing cherry tomatoes". In this way, standard phrase 7b is input as user 1's sentence 7c. The completed introduction sentence 3a is created with the same level of granularity for other users, allowing for accurate matching.

In the above, the fixed phrase 7b is inserted to allow the user to input specific content, but the user 1 can also replace the word "vegetables" that the message 4m is pointing to with "tomatoes" as appropriate.

Next, each processing section will be explained using a flowchart.

Figure 10 is a flowchart for explaining the entire keyword recommendation process. In Figure 10, when the keyword recommendation unit 40 receives an introduction 3a (step S201), the extraction unit 41 performs an extraction process to extract keywords 3d from the introduction 3a (step S202). The extracted keywords 3d are added to the keyword DB 52 as long as they are not duplicated in the keyword DB 52. If there are no new keywords 3d to be added, the subsequent steps S203 to S205 are omitted, the keyword recommendation process is terminated, and the reception of the next introduction 3a is awaited.

In response to the addition of the keyword 3d to the keyword DB 52, the determination unit 42 performs a determination process to determine the granularity of the added keyword 3d (step S203). The determination unit 42 refers to the ontology information 53 and determines whether the keyword 3d is coarser than the words contained in the introduction 3a of the other user 1 (step S203). If the determination unit 42 does not determine that the keyword 3d is coarse (YES in step S203), the keyword recommendation unit 40 ends this keyword recommendation process and waits for the next introduction 3a to be received.

If the determining unit 42 determines that the granularity is coarse (YES in step S203), the recommending unit 43 performs a recommendation process to recommend a candidate keyword 3d' (step S205).
Creation support information 3b is generated and transmitted to the terminal 3. Upon completion of the recommendation process, the keyword recommendation unit 40 ends this keyword recommendation process and waits for reception of the next introduction 3a.

FIG. 11 is a flowchart for explaining the extraction process. In FIG. 11, the extraction unit 41 determines whether the received introduction 3a is a new introduction or an update of the introduction 3a (step S211). If the received introduction 3a is neither a new introduction nor an update of the introduction 3a (NO in step S211), the extraction unit 41 ends the extraction process.

On the other hand, if the received testimonial 3a is a new testimonial or an update to the testimonial 3a (YES in step S211), the extraction unit 41 associates the received testimonial 3a with the user ID, stores it in the testimonial DB 51, and updates the testimonial DB 51 (step S212).

Then, the extraction unit 41 extracts keywords 3d from the received introduction 3a, and if there are no duplicates, adds a new record in which the extracted keywords 3d are recorded in association with the user ID to the keyword DB 52 and updates it (step S214). After that, the extraction unit 41 ends the extraction process and waits for the next introduction to be received.

FIG. 12 is a flow chart for explaining the determination process. In FIG. 12, the determination unit 42 determines whether the keyword DB 52 has been updated at regular intervals (step S211). If the keyword DB 52 has not been updated (NO in step S211), the determination unit 42 repeats step S211 at regular intervals.

If the keyword DB 52 has been updated (YES in step S211), the determination unit 42 refers to the ontology information 53 and compares the granularity of the added keyword 3d (step S221).

Then, the determination unit 42 determines whether the added keyword 3d is a target keyword for elaboration (step S223). The determination unit 42 determines whether any of the words in the lower layer than the position of the added keyword 3d in the ontology information 53 matches the keyword 3d extracted from the introduction 3a of another user 1 in the keyword DB 52. If a matching keyword 3d exists, the determination unit 42 determines that the added keyword 3d is a target keyword for elaboration. If it is determined that the added keyword 3d is not a target keyword for elaboration (NO in step S223), the determination unit 42 returns to step S221 and repeats the same process described above.

If it is determined that the keyword is a target keyword for elaboration (YES in step S223), the determination unit 42 notifies the recommendation unit 43 that a keyword is a target keyword for elaboration (step S224). The information notified to the recommendation unit 43 includes at least the keyword to be elaborated and the user ID.

FIG. 13 is a flowchart for explaining the recommendation process. In FIG. 13, the recommendation unit 43 determines at regular intervals whether or not a notification has been received from the determination unit 42 (step S231). If no notification has been received (NO in step S231), the recommendation unit 43 repeats step S231.

When a notification is received (YES in step S231), the recommendation unit 43 refers to the ontology information 53 and identifies candidate keywords to be recommended for use instead of or incidental use in place of the target keyword specified in the notification (step S232). The recommendation unit 43 refers to the ontology information 53 to identify the hierarchy in which the target keyword for refinement is located, and acquires words if a hierarchy deeper than the identified hierarchy exists. The recommendation unit 43 further acquires keywords 3d' included in the introduction text 3a of other users 1 that match any of the acquired words from the keyword DB 52.

Then, the recommendation unit 43 creates creation support information 3b including the identified candidate keyword 3d' and transmits it to the terminal 3 (step S233). The creation support information 3b further includes location information of the keyword to be detailed or the keyword to be detailed within the introduction 3a, a message, etc. The location information may be the number of characters from the first character of the introduction 3a to the keyword to be detailed. The creation support information 3b may further include a standard phrase.

Next, a case where keywords 3d are extracted from the testimonials 3a of three users u1 to u3 as shown in FIG. 4, and then a testimonial 3a is received from user u4 will be described with reference to FIG. 14 to FIG. 16.

Figure 14 is a sequence diagram for when the number of users increases midway through the system in this embodiment. Figure 14 shows the sequence for when user u4 joins after keywords 3d have been extracted from the introductions 3a of three users u1 to u3.

After extracting keywords 3d from the introduction 3a of each of users u1 to u3 (step S102-1), extract keywords 3d from the introduction 3a of user u4 (step S102-2).

As with the sequence in FIG. 4, the granularity of the keywords 3d is determined (step S103), and in this example, the introduction 3a of the user u4 contains coarse-grained keywords 3d. Creation support information 3b is sent to the terminal 3 of the user u4 (step S105), the user u4 edits the introduction 3a, and the text being input is sent to the information processing device 10 (step S106), and the same process as described above is repeated.

Figure 15 shows an example of data transition in the keyword DB for the operation example of Figure 14. Figure 15 (A) shows an example of data at the end of step S102-1 in Figure 14, and Figure 15 (B) shows an example of data at the end of step S102-2 in Figure 14. Also, Figure 15 (C) shows an example of data at the end of step S106 in Figure 14.

In this way, even if a lower-level word is present when referring to the ontology information 53, if the introduction 3a of another user 1 does not contain a keyword 3d that expresses a specific meaning, it is not recommended to refine it. This mechanism allows for accurate matching.

Figure 16 is a diagram showing an example screen when creation support information is provided by push notification. In Figure 16, user 1 can select push notification 8p to display an editing screen for the created introduction 3a on terminal 3. The editing screen includes a message 4m and candidate keywords 3d'. By selecting one of the candidate keywords 3d', the introduction 3a can be made more specific by performing the operations described in Figure 9.

The user 1 is prompted by the message 4m displayed on the terminal 3 to select a candidate keyword 3d' and edit the template 7b (Figure 9), so that even if the user is not good at creating introductions 3a, they can create an introduction 3a that can be properly matched.

All examples and specific terms cited herein are intended for instructional purposes to aid the reader in understanding the concepts contributed by the inventors to the advancement of the present invention and the art, and should not be construed as limiting the scope of any example configuration herein, or any such specifically cited examples and conditions, with respect to demonstrating the superiority or inferiority of the present invention.

Although the present embodiment has been described above using the meetup introduction text 3a as an example, the present embodiment is not limited to the meetup introduction text 3a. The present embodiment can be effectively applied to support specific description in various writings.

Although embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions and alterations can be made thereto without departing from the spirit and scope of the present invention.

The following supplementary notes are further disclosed regarding the embodiments including the above examples.
(Appendix 1)
Accepting a first keyword extracted from a first sentence of a first user;
Identifying a candidate keyword that is recommended for use instead of or in conjunction with the first keyword from second keywords extracted from second sentences of other users based on data defining a semantic inclusion relationship between a plurality of keywords;
A keyword recommendation program that causes a computer to perform a process of transmitting the candidate keywords to a terminal of the first user.
(Appendix 2)
The computer includes:
A keyword recommendation program as described in Appendix 1, which sends the first keyword and the candidate keywords to the terminal of the first user, specifies the first keyword in the first sentence displayed on the terminal, and displays the candidate keywords.
(Appendix 3)
The computer includes:
A keyword recommendation program as described in Appendix 1 or 2, which performs a process of determining that the granularity of the first keyword is coarse if there is a word with a more detailed meaning than the first keyword in the semantic inclusion relationship defined in the data.
(Appendix 4)
A keyword recommendation program as described in Appendix 3. If there is a word having a more detailed meaning than the first keyword in the semantic inclusion relationship defined in the data, and the second keyword does not match the word, the candidate keyword is not identified.
(Appendix 5)
Accepting a first keyword extracted from a first sentence of a first user;
Identifying a candidate keyword that is recommended for use instead of or in conjunction with the first keyword from second keywords extracted from second sentences of other users based on data defining a semantic inclusion relationship between a plurality of keywords;
The keyword recommendation method further comprises a process of transmitting the candidate keywords to a terminal of the first user by a computer.
(Appendix 6)
a receiving unit that receives a first keyword extracted from a first sentence of a first user;
an identification unit that identifies a candidate keyword that is recommended for use instead of or in conjunction with the first keyword from second keywords extracted from second sentences of other users based on data that defines a semantic inclusion relationship between a plurality of keywords;
A transmitting unit that transmits the candidate keywords to a terminal of the first user.
(Appendix 7)
The transmission unit transmits the first keyword and the candidate keywords to the terminal of the first user;
7. The information processing device according to claim 6, wherein the terminal specifies the first keyword in the first sentence displayed on the terminal and displays the candidate keywords.

REFERENCE SIGNS LIST 1 User 2 Internet 3 Terminal 3a Testimonial 3b Creation support information 3d, 3d' Keywords 40 Keyword recommendation unit 41 Extraction unit 42 Determination unit 43 Recommendation unit 51 Testimonial DB
52 Keyword DB
53 Ontology information 53
100 Information processing device

Claims (5)

Extracting a first keyword from a first sentence of a first user received from a terminal of the first user;
Identifying, from second keywords extracted from second sentences of other users based on data defining semantic inclusion relationships between a plurality of keywords, a candidate keyword that, instead of the first keyword, more specifically expresses an object represented by the first keyword;
A keyword recommendation program that causes a computer to perform a process of transmitting the candidate keywords to a terminal of the first user. The computer includes:
The keyword recommendation program according to claim 1 , further comprising a process for causing the terminal of the first user to display a notification including the first keyword, a message specifying the first keyword, and the candidate keyword. The computer includes:
If there is a word having a more detailed meaning than the first keyword in the semantic inclusion relationship defined by the data, the first keyword is determined to have a coarse granularity;
3. The keyword recommendation program according to claim 1, further comprising a step of specifying the candidate keywords when the granularity of the primary keywords is determined to be coarse. Extracting a first keyword from a first sentence of a first user received from a terminal of the first user;
Identifying, from second keywords extracted from second sentences of other users based on data defining semantic inclusion relationships between a plurality of keywords, a candidate keyword that, instead of the first keyword, more specifically expresses an object represented by the first keyword;
The keyword recommendation method further comprises a process of transmitting the candidate keywords to a terminal of the first user by a computer. an extraction unit that extracts a first keyword from a first sentence of a first user received from a terminal of the first user;
an identification unit that identifies, instead of the first keyword, a candidate keyword that more specifically expresses an object represented by the first keyword from second keywords extracted from second sentences of other users based on data that specifies a semantic inclusion relationship between a plurality of keywords;
A transmitting unit that transmits the candidate keywords to a terminal of the first user.

Images