JP7708412B2 - Recommendation device, recommendation method, program, and recording medium - Google Patents

Description

The present invention relates to a recommendation device, a recommendation method, a program, and a recording medium.

In human interaction, it is important to convey one's feelings to the other person. For example, Patent Document 1 reports a technology that determines the user's feelings based on the strength of the user's grip and communicates these determined feelings to other users.

JP 2000-066807 A

However, in Patent Document 1, the CG model is used to convey one's own feelings unilaterally, without considering the other person. This causes the problem that one's own feelings are not fully conveyed to the other person.

The present invention aims to provide a recommendation device, recommendation method, program, and recording medium that can recommend sentences that express one's feelings and are effective for the other person.

In order to achieve the above object, the recommendation device of the present invention comprises:
The method includes: acquiring qualification information; acquiring template information; selecting; and recommending;
The qualification information acquisition unit acquires qualification information of the other user based on identification information of the other user,
The qualification information is information indicating qualifications,
The fixed phrase information acquisition unit acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection unit selects at least one of the plurality of fixed phrases based on the degree of influence,
The recommendation unit is a device that recommends to one user that the fixed phrase be used as the feeling-expressing sentence.

The recommendation method of the present invention includes:
The method includes a qualification information acquisition step, a template information acquisition step, a selection step, and a recommendation step;
The qualification information acquisition step includes acquiring qualification information of the other user based on identification information of the other user;
The qualification information is information indicating qualifications,
The fixed phrase information acquiring step acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection step includes selecting at least one of the plurality of fixed phrases based on the degree of influence,
The recommending step is a method of recommending to one user that the template phrase be used as the feeling expressing sentence.

The present invention can recommend to users sentences that express their feelings and are effective for the other person.

FIG. 1 is a block diagram illustrating an example of a configuration of a recommendation server according to the first embodiment. FIG. 2 is a block diagram illustrating an example of a hardware configuration of the recommendation server according to the first embodiment. FIG. 3 is a flowchart illustrating an example of a process in the recommendation server according to the first embodiment. FIG. 4 is a schematic diagram showing an example of the qualification information database, the fixed phrase information database, and the dictionary information database. FIG. 5 is a schematic diagram showing an example of calculating the influence degree. FIG. 6 is a schematic diagram illustrating an example of a process performed by the extraction unit. FIG. 7 is a schematic diagram showing an example of processing by the co-occurrence network forming unit. FIG. 8 is a schematic diagram showing an example of processing performed by the fixed phrase information acquiring unit, the selecting unit, and the recommended sentence generating unit. FIG. 9 is a schematic diagram showing an example of updating the influence degree when negative evaluation information is acquired.

The recommendation device of the present invention, for example,
The system further includes an unsent sentence acquisition unit, an extraction unit, and a recommended sentence generation unit,
The unsent text acquisition unit acquires, from one user, an unsent feeling expression text and identification information of the other user to which the unsent feeling expression text is to be sent, in association with each other;
the template selected by the selection unit is an incomplete sentence that prompts for input of at least one word,
The extraction unit extracts at least one key word from the unsent sentence,
the recommended sentence generation unit inputs the key words into the template sentence to generate a recommended sentence;
The recommendation unit may be configured to recommend to the one user that the one user use the generated recommended sentence as the feeling expressing sentence.

In the recommendation device of the present invention, for example,
the extraction unit includes a syntax analysis unit, a co-occurrence network formation unit, and a dictionary information acquisition unit;
The parsing unit performs parsing of the unsent feeling expression sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming unit forms a co-occurrence network with each of the extracted words as a vertex,
The dictionary information acquisition unit acquires dictionary information including words related to feelings,
The extraction unit may extract, as the key word, a word that satisfies a preset condition based on the co-occurrence network and the dictionary information.

The recommendation device of the present invention, for example,
Further, the system includes an evaluation information acquisition unit and an impact update unit,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended by the recommendation unit,
The evaluation information acquisition unit acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
The influence updating section may update the influence using the evaluation information.

The recommendation device of the present invention, for example,
Further, the recording unit is included.
The recording unit executes at least one of: linking the user's identification information with the qualities of the user and recording the linked information as quality information; and linking the standard phrase, the quality information, and an influence level for each standard phrase related to a feeling expression sentence and recording the standard phrase information;
At least one of the qualification information acquisition unit and the fixed phrase information acquisition unit may acquire at least one of the recorded qualification information and the fixed phrase information.

In the recommendation device of the present invention, for example,
The feeling-expressing sentence may be a sentence expressing a positive feeling.

In the recommendation device of the present invention, for example,
The positive feeling may be gratitude.

The recommendation method of the present invention includes, for example,
The method further includes an unsent text acquisition step, an extraction step, and a recommended text generation step,
The unsent text acquisition step acquires, from one user, an unsent feeling-expressing text and identification information of the other user to which the unsent feeling-expressing text is to be sent, in association with each other;
the template selected by the selection step is an incomplete sentence that prompts input of at least one word,
The extraction step includes extracting at least one important word from the unsent text,
The recommended sentence generating step inputs the key words into the template sentence to generate a recommended sentence;
The recommending step may be a step of recommending to the one user that the generated recommended sentence be used as the feeling expressing sentence.

In the recommendation method of the present invention, for example,
The extraction step includes a syntax analysis step, a co-occurrence network formation step, and a dictionary information acquisition step;
The parsing step includes parsing the unsent emotion-expressing sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming step uses each of the extracted words to form a co-occurrence network having each of the words as a vertex;
The dictionary information acquisition step acquires dictionary information including words related to feelings,
The extracting step may be such that a word that satisfies a preset condition is extracted as the key word based on the co-occurrence network and the dictionary information.

The recommendation method of the present invention includes, for example,
Further, the method includes an evaluation information acquisition step and an impact degree update step,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended in the recommendation step,
The evaluation information acquisition step acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
The influence updating step may be configured to update the influence using the evaluation information.

The recommendation method of the present invention includes, for example,
Further, a recording step is included,
The recording step includes at least one of: linking the user's identification information with the qualities of the user and recording the linked information as quality information; and linking the standard phrase, the quality information, and an influence level for each standard phrase related to a feeling expression sentence and recording the standard phrase information;
At least one of the qualification information acquiring step and the fixed phrase information acquiring step may be configured to acquire at least one of the recorded qualification information and the fixed phrase information.

In the recommendation method of the present invention, for example,
The feeling-expressing sentence may be a sentence expressing a positive feeling.

In the recommendation method of the present invention, for example,
The positive feeling may be gratitude.

The program of the present invention is a program for causing a computer to execute each step of the method of the present invention as a procedure.

The recording medium of the present invention is a computer-readable recording medium on which the program of the present invention is recorded.

In the present invention, "feelings" refers to a person's state of mind, emotions, way of thinking, mood, etc., and should be interpreted in the broadest sense, including positive and negative feelings, and is not limited in any way. Examples of the positive feelings include gratitude, happiness, fun, joy, support, praise, respect, pride, excitement, comfort, etc., and examples of the negative feelings include sadness, loneliness, pain, pain, fear, apology, worry, sorrow, etc.

Next, an embodiment of the present invention will be described with reference to the drawings. The present invention is not limited to the following embodiment. In each of the drawings, the same parts are given the same reference numerals. Furthermore, the explanations of each embodiment can be mutually incorporated unless otherwise specified, and the configurations of each embodiment can be combined unless otherwise specified.

[Embodiment 1]
FIG. 1 is a block diagram showing an example of the configuration of a recommendation device 10 of this embodiment. As shown in FIG. 1, the device 10 includes, for example, a recording unit 11, an unsent sentence acquisition unit 12, a qualification information acquisition unit 13, a fixed phrase information acquisition unit 14, a selection unit 15, an extraction unit 16, a recommended sentence generation unit 17, a recommendation unit 18, an evaluation information acquisition unit 19, and an influence update unit 20. Note that the recording unit 11, the unsent sentence acquisition unit 12, the extraction unit 16, the recommended sentence generation unit 17, the evaluation information acquisition unit 19, and the influence update unit 20 may be of any configuration and may not be included in the device 10. In addition, the extraction unit 16 may include, for example, a syntax analysis unit 161, a co-occurrence network formation unit 162, and a dictionary information acquisition unit 163.

The device 10 may be, for example, a single device including each of the above-mentioned parts, or a device to which each of the above-mentioned parts can be connected via a communication line network. The device 10 may also be connected to an external device described later via the communication line network. The communication line network is not particularly limited and may be a publicly known network, for example, wired or wireless. Examples of the communication line network include the Internet line, WWW (World Wide Web), telephone line, LAN (Local Area Network), SAN (Storage Area Network), DTN (Delay Tolerant Networking), LPWA (Low Power Wide Area), L5G (Local 5G), etc. Examples of wireless communication include Wi-Fi (registered trademark), Bluetooth (registered trademark), local 5G, and LPWA. The wireless communication may be a form in which each device communicates directly (Ad Hoc communication), infrastructure communication, indirect communication via an access point, and the like. The device 10 may be incorporated into a server as a system. The device 10 may be, for example, a personal computer (PC, for example, desktop type or notebook type), a smartphone, a tablet terminal, a wearable device (smart watch, etc.), or the like in which the program of the present invention is installed. Furthermore, the device 10 may be, for example, a form such as cloud computing or edge computing in which at least one of the parts is on a server and the other parts are on a terminal.

Figure 2 shows an example block diagram of the hardware configuration of the device 10. The device 10 includes, for example, a central processing unit (CPU, GPU, etc.) 101, a memory 102, a bus 103, a storage device 104, an input device 105, a display device 106, a communication device 107, etc. Each part of the device 10 is connected to each other via the bus 103 by its respective interface (I/F).

The central processing unit 101 is responsible for the overall control of the device 10. In the device 10, the central processing unit 101 executes, for example, the program of the present invention and other programs, and also reads and writes various information. Specifically, for example, the central processing unit 101 functions as a recording unit 11, an unsent text acquisition unit 12, a quality information acquisition unit 13, a template information acquisition unit 14, a selection unit 15, an extraction unit 16, a recommended text generation unit 17, a recommendation unit 18, an evaluation information acquisition unit 19, and an impact update unit 20.

The bus 103 can also be connected to, for example, an external device. Examples of the external device include an external storage device (external database, etc.), a printer, an external input device, an external display device, an external imaging device, etc. The present device 10 can be connected to an external network (the above-mentioned communication line network) by, for example, a communication device 107 connected to the bus 103, and can also be connected to other devices via the external network.

The memory 102 may be, for example, a main memory (primary storage device). When the central processing unit 101 performs processing, the memory 102 reads various operating programs, such as the program of the present invention, stored in the storage device 104 described below, and the central processing unit 101 receives data from the memory 102 and executes the program. The main memory may be, for example, a RAM (random access memory). The memory 102 may also be, for example, a ROM (read only memory).

The storage device 104 is also referred to as an auxiliary storage device, for example, in contrast to the main memory. As described above, the storage device 104 stores operating programs including the program of the present invention. The storage device 104 may be, for example, a combination of a recording medium and a drive that reads and writes from the recording medium. The recording medium is not particularly limited, and may be, for example, an internal or external type, such as a hard disk drive (HDD), CD-ROM, CD-R, CD-RW, MO, DVD, flash memory, memory card, etc. The storage device 104 may be, for example, a hard disk drive (HDD) in which the recording medium and the drive are integrated, or a solid state drive (SSD).

In the present device 10, the memory 102 and the storage device 104 can also store various information such as log information, information acquired from an external database (not shown) or an external device, information generated by the present device 10, and information used by the present device 10 when executing processing. At least a portion of the information may be stored, for example, in an external server other than the memory 102 and the storage device 104, or may be stored in a distributed manner on multiple terminals using blockchain technology or the like. The memory 102 and the storage device 104 may also store, for example, various information recorded by the recording unit 11.

The device 10 may further include, for example, an input device 105 and a display device 106. The input device 105 is, for example, a touch panel, a keyboard, a mouse, etc. The display device 106 is, for example, an LED display, a liquid crystal display, etc.

As shown in FIG. 1, the device 10 may further include a qualification information database 1, a fixed phrase information database 2, and a dictionary information database 3. The device 10 can be connected to the qualification information database 1, the fixed phrase information database 2, and the dictionary information database 3, for example, via the communication line network.

Next, an example of the recommendation method of this embodiment will be described based on the flowchart in FIG. 3. The recommendation method of this embodiment is implemented, for example, as follows using the recommendation device 10 in FIG. 1. Note that the recommendation method of this embodiment is not limited to the use of the recommendation device 10 in FIG. 1. Also, the steps in parentheses in FIG. 3 are optional steps.

The recommendation method of the present embodiment may, for example, execute at least one of the following in advance by the recording unit 11: linking the user's identification information with the qualities of the user and recording them as quality information; and linking the standard phrase, the quality information, and the influence degree for each standard phrase related to a feeling expression sentence and recording them as standard phrase information (S11, recording step). The recording unit 11 may, for example, use a known technology for identifying the qualities of a user (for example, a strengths finder, etc.) and record the qualities identified by the technology in association with the identification information of the user. The recording unit 11 may, for example, store dictionary information described later in the step (S11). A specific description of the quality information, the standard phrase information, and the dictionary information (hereinafter collectively referred to as quality information, etc.) will be described later. The recording unit 11 may, for example, build a database that stores the quality information, etc. In the present invention, the database that stores the qualification information is referred to as qualification information database 1, the database that stores the fixed phrase information is referred to as fixed phrase information database 2, and the database that stores the dictionary information is referred to as dictionary information database 3. Furthermore, the recording unit 11 may record the qualification information, etc., by storing it in an external database, or by storing the qualification information, etc. in memory 102 and storage device 104, for example.

First, for example, the unsent text acquisition unit 12 may acquire from one user an unsent emotion-expressing text in association with the identification information of the other user to whom the unsent emotion-expressing text is to be sent (S12, unsent text acquisition step). Hereinafter, an unsent emotion-expressing text is also referred to as an unsent text. The emotion-expressing text is, for example, a text expressing the emotion of the one user toward the other user. The unsent text acquisition unit 12 may acquire the text from an external device, for example, via the communication line network.

Next, the quality information acquisition unit 13 acquires quality information of the other user based on the identification information of the other user (S13, quality information acquisition step). The quality information is information indicating qualities. In other words, the qualities are, for example, nature, talent, ability, thinking, temperament, etc. Specifically, examples of the qualities include sociability, cooperation, communication skills, positive temperament, competitiveness, strategy, sense of responsibility, and introspection. The quality information acquisition unit 13 may acquire the quality information from the outside via, for example, the communication line network, or may acquire quality information recorded by the recording unit 11 (for example, quality information stored in the quality information database 1, etc.).

Next, the fixed phrase information acquisition unit 14 acquires fixed phrase information according to the quality information of the other user (S14, fixed phrase information acquisition process). The fixed phrase information is information linking the fixed phrase, the quality information, and the impact degree for each fixed phrase related to a feeling expression sentence. The impact degree is an index indicating the degree of psychological impact that a user having the quality indicated by the quality information will have when receiving the fixed phrase. Here, the fixed phrase may be an incomplete sentence that requests the input of at least one word, or may be a complete sentence. The incomplete sentence can become a complete sentence by inputting an important word extracted by the extraction unit 16 described later or any word input by the user into the part that requests the input of the word. There may be two or more parts that request the input of the word. The complete sentence means a sentence that is established as a sentence, that is, a sentence that makes sense without inputting a new word. The impact degree for each fixed phrase can be updated, for example, by the impact degree update unit 20 described later. The fixed phrase information acquisition unit 14 may, for example, acquire the fixed phrase information from outside via the communication line network, or may acquire fixed phrase information recorded by the recording unit 11 (for example, fixed phrase information stored in the fixed phrase information database 2).

Next, the selection unit 15 selects at least one of the plurality of fixed phrases based on the degree of influence (S15, selection step). The fixed phrase selected by the selection unit 15 may be the incomplete sentence or the complete sentence.

If the template sentence selected by the selection unit 15 is an incomplete sentence, for example, after the step (S15), the extraction unit 16 may extract at least one important word from the unsent sentence (S16, extraction step). The extraction unit 16 may use a known algorithm such as TextRank. Hereinafter, the specific process of the step (S16) will be described with an example in which the extraction unit 16 includes a syntax analysis unit 161, a co-occurrence network formation unit 162, and a dictionary information acquisition unit 163. The syntax analysis unit 161 performs syntax analysis on the unsent emotion expression sentence to extract words that are nouns and words that are adjectives (S161, syntax analysis step). Next, the co-occurrence network formation unit 162 uses the extracted words to form a co-occurrence network with each word as a vertex (S162, co-occurrence network formation step). The co-occurrence network is, for example, a network that indicates the co-occurrence relationship of each word. Next, the dictionary information acquisition unit 163 acquires dictionary information in which words related to feelings are recorded (S163, dictionary information acquisition step). The dictionary information acquisition unit 163 may acquire the dictionary information from the outside via the communication line network, for example, or may acquire dictionary information recorded by the recording unit 11 (for example, dictionary information stored in the dictionary information database 3, etc.). As the dictionary information, for example, an existing dictionary such as a Japanese emotion expression dictionary (URL: http://www.jnlp.org/SNOW/D18) can be used. Then, the extraction unit 16 extracts words that satisfy a preset condition as the important word based on the co-occurrence network and the dictionary information (S164, extraction step). The preset condition is not particularly limited and can be set arbitrarily. Specifically, the condition may be, for example, a word with a high degree, a word in the dictionary information, a word with a score calculated by a preset formula that is higher than other words, or any combination of two or more of these. The formula is not particularly limited.

Next, after the step (S16), for example, the recommended sentence generation unit 17 inputs the key word into the template sentence to generate a recommended sentence (S17, recommended sentence generation step). In other words, the recommended sentence can be said to be the template sentence that has become the complete sentence by inputting the key word.

If the steps (S16) to (S17) are not executed, the recommendation unit 18 recommends to the one user that the template sentence be used as the feeling-expressing sentence (S18, recommendation step), and ends (END). In this case, the recommendation unit 18 may recommend to the one user the template sentence that is the incomplete sentence, or may recommend to the one user the template sentence that is the complete sentence. On the other hand, if the steps (S16) to (S17) are executed, in the step (S18), the recommendation unit 18 recommends to the one user that the generated recommended sentence be used as the feeling-expressing sentence. In this case, the recommended sentence recommended to the one user is a complete sentence, as described above. The recommendation unit 18 may recommend, for example, the template sentence by displaying it on a display device (the display device 106 or an external display device), or may recommend the template sentence by transmitting it to the one user via the communication line network.

The following describes the process (steps (S19) to (S20)) when, after step (S18), the one user sends a feeling-expressing sentence to the other user using at least one of the template sentences and the recommended sentences recommended by the recommendation unit 18. Note that the sending may be performed by the device 10 or an external device, for example.

First, after the step (S18), the evaluation information acquisition unit 19 acquires evaluation information indicating the other user's evaluation of the feeling-expressing sentence (S19, evaluation information acquisition step). The evaluation is, for example, an evaluation indicating whether the feeling-expressing sentence is favorable or not, and may be a multi-level evaluation. The evaluation information acquisition unit 19 may acquire the evaluation information from outside, for example, via the communication line network.

Next, after the step (S19), the impact update unit 20 uses the evaluation information to update the impact (S20, impact update step), and then ends (END). The details will be described later.

According to this embodiment, by taking into consideration the qualities of the other user, it is possible to recommend to the one user a feeling-expressing sentence that is effective for the other user, in other words, a feeling-expressing sentence that will easily convey the other user's feelings. In this way, this embodiment can eliminate the anxiety of the user (the sender of the feeling-expressing sentence), such as "I don't know what words to send to the other user" or "I don't know if the other person understands the feelings I sent." Furthermore, by generating the recommended sentence, it is possible to send the feeling-expressing sentence (the recommended sentence) to the other user more easily, since it does not require the one user to take the trouble.

[Embodiment 2]
The present invention will be described in detail below by taking a sentence expressing gratitude as an example of the sentiment-expressing sentence.

An example of how the recording unit 11 constructs the qualification information database 1, the fixed phrase information database 2, and the dictionary information database 3 will be described with reference to Figures 4 and 5. Figure 4 shows an example of the qualification information database 1, the fixed phrase information database 2, and the dictionary information database 3. The qualification information in Figure 4 is linked to the user's identification information (user name) and the qualities of the user. The recording unit 11 records the qualification information using publicly known technology (e.g., StrengthsFinder, etc.) as described above, and constructs the qualification information database 1 by accumulating the qualification information.

The template in the template information in FIG. 4 is an incomplete sentence, and "***" is the part that requests the input of the word. The higher the impact level, the greater the psychological impact. FIG. 5 shows an example of calculating the impact level. First, although not shown, the monitor uses the monitor's terminal (e.g., a smartphone, etc.) to identify the monitor's own qualities using a known technology such as a strengths finder. Next, the monitor uses the terminal to register the qualities in an input field of a questionnaire site ("1" in the figure). Next, a device that operates the questionnaire site (e.g., the present device 10) displays a question about the template on the questionnaire site and presents the question to the monitor ("2" in the figure). Specifically, the question is, for example, a question that can be answered by whether you like it or not, such as "Do you like the word '###'?" The "###" is, for example, a preset template itself. Next, the monitor uses the monitor's terminal to register the answer to the question in the input field of the survey site ("3" in the figure). Next, the device calculates the influence level by dividing the "number of monitors who answered that they like '###'" by the "number of monitors who answered the question" for multiple monitors who have the same quality. The recording unit 11 then records the standard phrase information by linking the standard phrase '###', the quality, and the influence level, and builds the standard phrase information database 2 by accumulating the standard phrase information.

The dictionary information in FIG. 4 links words related to "gratitude" with their readings (in katakana). The recording unit 11 builds the dictionary information database 3 by accumulating the dictionary information.

FIG. 6 shows an example of the processing of the extraction unit 16. An example of an unsent sentence acquired by the unsent sentence acquisition unit 11 is shown in the upper part of FIG. 6. The syntax analysis unit 161 performs syntax analysis on the unsent sentence, and extracts words representing nouns and words representing adjectives as shown in the figure. The co-occurrence network formation unit 162 forms the co-occurrence network shown in FIG. 6 by a process described later. Also, it is assumed that the dictionary information acquisition unit 163 has acquired the dictionary information shown in FIG. 4. For example, the extraction unit 16 extracts the nouns "approval" and "hurry up" that are close to the word "thank you" in the dictionary information in the co-occurrence network as candidates for the important word. Then, the extraction unit 16 extracts, for example, a word with a high score calculated by the following formula (1) from among the candidates as the important word. Note that the "weight" in the following formula (1) can be set arbitrarily, for example, 5, 10, 15, etc. 6, since the score of "approval" is higher than the score of "early", "approval" is extracted as the keyword by the extraction unit 16. On the other hand, the extraction unit 16 may extract "approval", which has the highest degree of "4", as the keyword without using the following formula (1).

Score = (number of degrees) + (words in adjacent dictionary information) × (weight) ... (1)

Figure 7 shows an example of processing by the co-occurrence network forming unit 162. First, as shown in the upper part of Figure 7, the co-occurrence network forming unit 162 creates an adjacency matrix with words representing nouns and words representing adjectives extracted by the syntax analysis unit 161. Here, the adjacency matrix refers to a matrix in which each column and each row corresponds to each vertex of the co-occurrence network, and parts with edges are 1. Next, as shown in the center part of Figure 7, the co-occurrence network forming unit 162 updates the adjacency matrix by setting adjacent words to "1". For example, a known technique such as a 1-gram model can be used to update the adjacency matrix. Then, the co-occurrence network forming unit 162 forms the co-occurrence network based on the adjacency matrix, as shown in the lower part of Figure 7.

Figure 8 shows an example of processing by the fixed phrase information acquisition unit 14, the selection unit 15, and the recommended sentence generation unit 17. The fixed phrase information acquisition unit 14 acquires fixed phrase information according to the quality "sociability" of the other user "Suzuki". In this example, as shown in Figure 8, it is assumed that fixed phrase information related to the quality "sociability" that is the same as the quality "sociability" of "Suzuki" is acquired as the fixed phrase information. As shown in the figure, the fixed phrase information acquisition unit 14 may acquire a plurality of the fixed phrase information. The selection unit 15 selects the fixed phrase with the highest degree of influence from the plurality of the fixed phrase information. As shown in Figure 8, the recommended sentence generation unit 17 generates a recommended sentence, which is a complete sentence, by inserting the key word "approval" extracted in the processing shown in Figure 6 into "***" of the fixed phrase.

In this way, the series of processes described in this embodiment can convert the unsent sentence shown in Figure 6 into the recommended sentence shown in Figure 8.

[Embodiment 3]
An example of updating the influence degree will be specifically described.

The other user, for example, inputs evaluation information indicating an evaluation of the received feeling-expressing sentence using the other user's terminal (e.g., a smartphone, etc.). In other words, the evaluation means whether or not the feeling-expressing sentence is favorable. Specifically, for example, if the other user likes the feeling-expressing sentence, the other user may evaluate the feeling-expressing sentence by tapping a button indicating favorability, such as "Like!", for the feeling-expressing sentence displayed on the display of the terminal. On the other hand, if the other user does not like the feeling-expressing sentence, for example, he or she may not tap the button indicating favorability, or may evaluate it by tapping a button indicating badness, such as "BAD."

The evaluation information acquisition unit 19 acquires the evaluation information linked to the feeling-expressing sentence. The evaluation information includes, for example, good impression evaluation information, which means that the feeling-expressing sentence is favorable, and bad impression evaluation information, which means that the feeling-expressing sentence is unfavorable. For example, when the button indicating the favorable impression is tapped, the evaluation information acquisition unit 19 acquires the good impression evaluation information as the evaluation information. On the other hand, the evaluation information acquisition unit 19 acquires the unfavorable impression evaluation information as the evaluation information, for example, when the button indicating the unfavorable impression is tapped or when the button indicating the favorable impression is not tapped within a predetermined period of time after the other user receives the feeling-expressing sentence.

For example, when the influence updating unit 20 acquires the favorable evaluation information, the influence updating unit 20 updates the influence by adding a preset weight to the influence. On the other hand, when the influence updating unit 20 acquires the unfavorable evaluation information, the influence updating unit 20 updates the influence by subtracting a preset weight from the influence. The weight is not particularly limited and can be set arbitrarily, for example, 0.03, 0.05, 1.0, etc. FIG. 9 shows an example of updating the influence when the unfavorable evaluation information is acquired. In this case, as shown in the figure, the influence updating unit 20 subtracts the weight "0.05" from the influence "0.23" of the standard phrase corresponding to the transmitted feeling expression sentence to calculate the value "0.17". Then, the influence updating unit 20 updates the influence of the standard phrase corresponding to the transmitted feeling expression sentence from "0.23" to "0.17".

In this way, by updating the degree of influence, the template phrase that is more suited to the qualities of the other user can be selected, and more effective sentences expressing feelings can be recommended to the one user.

[Embodiment 4]
The program of the present embodiment is a program for causing a computer to execute each step of the method of the present invention as a procedure. In the present invention, the "procedure" may be read as "processing." The program of the present embodiment may be recorded, for example, in a computer-readable recording medium. The recording medium is not particularly limited, and examples thereof include a read-only memory (ROM), a hard disk (HD), and an optical disk.

The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

<Additional Notes>
Some or all of the above embodiments may be described as follows, but are not limited to the following:
(Appendix 1)
The method includes: acquiring qualification information; acquiring fixed phrase information; selecting; and recommending;
The qualification information acquisition unit acquires qualification information of the other user based on identification information of the other user,
The qualification information is information indicating qualifications,
The fixed phrase information acquisition unit acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection unit selects at least one of the plurality of fixed phrases based on the degree of influence,
The recommendation unit is a recommendation device that recommends to one user that the fixed phrase be used as the feeling expressing sentence.
(Appendix 2)
The system further includes an unsent sentence acquisition unit, an extraction unit, and a recommended sentence generation unit,
The unsent text acquisition unit acquires, from one user, an unsent feeling expression text and identification information of the other user to which the unsent feeling expression text is to be sent, in association with each other;
the template selected by the selection unit is an incomplete sentence that prompts for input of at least one word,
The extraction unit extracts at least one key word from the unsent sentence,
the recommended sentence generation unit inputs the key words into the template sentence to generate a recommended sentence;
The recommendation device according to claim 1, wherein the recommendation unit recommends to the one user that the generated recommended sentence be used as the feeling-expressing sentence.
(Appendix 3)
the extraction unit includes a syntax analysis unit, a co-occurrence network formation unit, and a dictionary information acquisition unit;
The parsing unit performs parsing of the unsent feeling expression sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming unit forms a co-occurrence network with each of the extracted words as a vertex,
The dictionary information acquisition unit acquires dictionary information including words related to feelings,
The recommendation device according to claim 2, wherein the extraction unit extracts words that satisfy a predetermined condition as the important word based on the co-occurrence network and the dictionary information.
(Appendix 4)
Further, the system includes an evaluation information acquisition unit and an impact update unit,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended by the recommendation unit,
The evaluation information acquisition unit acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
The recommendation device according to any one of claims 1 to 3, wherein the influence updating unit updates the influence using the evaluation information.
(Appendix 5)
Further, the recording unit is included,
The recording unit executes at least one of: linking the user's identification information with the qualities of the user and recording the linked information as quality information; and linking the standard phrase, the quality information, and an influence level for each standard phrase related to a feeling expression sentence and recording the standard phrase information;
A recommendation device described in any one of appendices 1 to 4, wherein at least one of the qualification information acquisition unit and the fixed phrase information acquisition unit acquires at least one of the recorded qualification information and the fixed phrase information.
(Appendix 6)
The recommendation device according to any one of appendices 1 to 5, wherein the feeling-expressing sentence is a sentence expressing a positive feeling.
(Appendix 7)
The recommendation device according to any one of appendixes 1 to 6, wherein the positive feeling is gratitude.
(Appendix 8)
The method includes a qualification information acquisition step, a template information acquisition step, a selection step, and a recommendation step;
The qualification information acquisition step includes acquiring qualification information of the other user based on identification information of the other user;
The qualification information is information indicating qualifications,
The fixed phrase information acquiring step acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection step includes selecting at least one of the plurality of fixed phrases based on the degree of influence,
The recommendation method includes a step of recommending to one user that the template phrase be used as the feeling expression sentence.
(Appendix 9)
The method further includes an unsent text acquisition step, an extraction step, and a recommended text generation step,
The unsent text acquisition step acquires, from one user, an unsent feeling-expressing text and identification information of the other user to which the unsent feeling-expressing text is to be sent, in association with each other;
the template selected by the selection step is an incomplete sentence that prompts input of at least one word,
The extraction step includes extracting at least one important word from the unsent text,
The recommended sentence generating step inputs the key words into the template sentence to generate a recommended sentence;
The recommendation method according to claim 8, wherein the recommendation step recommends to the one user that the generated recommended sentence be used as the feeling-expressing sentence.
(Appendix 10)
The extraction step includes a syntax analysis step, a co-occurrence network formation step, and a dictionary information acquisition step;
The parsing step includes parsing the unsent emotion-expressing sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming step uses each of the extracted words to form a co-occurrence network having each of the words as a vertex;
The dictionary information acquisition step acquires dictionary information including words related to feelings,
The recommendation method according to claim 9, wherein the extraction step extracts words that satisfy a predetermined condition as the important words based on the co-occurrence network and the dictionary information.
(Appendix 11)
Further, the method includes an evaluation information acquisition step and an impact degree update step,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended in the recommendation step,
The evaluation information acquisition step acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
The recommendation method according to any one of appendices 8 to 10, wherein the influence updating step updates the influence using the evaluation information.
(Appendix 12)
Further, a recording step is included,
The recording step includes at least one of: linking the user's identification information with the qualities of the user and recording the linked information as quality information; and linking the standard phrase, the quality information, and an influence level for each standard phrase related to a feeling expression sentence and recording the standard phrase information;
A recommendation method described in any one of Appendices 8 to 11, wherein at least one of the qualification information acquisition process and the fixed phrase information acquisition process acquires at least one of the recorded qualification information and the fixed phrase information.
(Appendix 13)
The recommendation method according to any one of appendices 8 to 12, wherein the feeling-expressing sentence is a sentence expressing a positive feeling.
(Appendix 14)
The recommendation method according to any one of appendices 8 to 13, wherein the positive feeling is gratitude.
(Appendix 15)
A program for causing a computer to execute procedures including a qualification information acquisition procedure, a template information acquisition procedure, a selection procedure, and a recommendation procedure:
The qualification information acquisition step includes acquiring qualification information of the other user based on identification information of the other user;
The qualification information is information indicating qualifications,
The step of acquiring fixed phrase information includes acquiring fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
the selecting step includes selecting at least one of the plurality of the fixed phrases based on the degree of influence;
The recommendation step recommends to one user that the template phrase be used as the feeling-expressing sentence.
(Appendix 16)
Further, the method includes an unsent text acquisition step, an extraction step, and a recommended text generation step,
The unsent text acquisition step includes acquiring, from one user, an unsent emotion expression text and identification information of the other user to which the unsent emotion expression text is to be sent, in association with each other;
the template selected by the selection step is an incomplete sentence that prompts for input of at least one word;
The extraction step includes extracting at least one important word from the untransmitted text;
the step of generating recommended sentences includes inputting the key words into the template sentences to generate recommended sentences;
The program according to claim 15, wherein the recommendation step recommends to the one user that the generated recommended sentence be used as the feeling-expressing sentence.
(Appendix 17)
The extraction step includes a syntax analysis step, a co-occurrence network formation step, and a dictionary information acquisition step;
The parsing step parsing the untransmitted sentiment expression sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming step includes forming a co-occurrence network with each of the extracted words as a vertex,
The dictionary information acquisition step acquires dictionary information including words related to feelings,
17. The program according to claim 16, wherein the extraction step extracts words that satisfy a preset condition as the important words based on the co-occurrence network and the dictionary information.
(Appendix 18)
Further, the method includes an evaluation information acquisition step and an impact degree update step,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended by the recommendation procedure,
The evaluation information acquisition step acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
18. The program according to any one of appendices 15 to 17, wherein the impact update step updates the impact using the evaluation information.
(Appendix 19)
Further, the method includes a recording procedure.
The recording step includes at least one of: linking the user's identification information with the qualities of the user and recording the linked information as quality information; and linking the standard phrase, the quality information, and an influence level for each standard phrase related to a feeling expression sentence and recording the standard phrase information;
A program described in any one of Appendices 15 to 18, wherein at least one of the qualification information acquisition procedure and the fixed phrase information acquisition procedure acquires at least one of the recorded qualification information and the fixed phrase information.
(Appendix 20)
20. The program described in any one of appendices 15 to 19, wherein the feeling-expressing sentence is a sentence expressing a positive feeling.
(Appendix 21)
21. The program of any one of claims 15 to 20, wherein the positive feeling is gratitude.
(Appendix 22)
A computer-readable recording medium having a program according to any one of appendices 15 to 21 recorded thereon.

According to the present invention, it is possible to recommend sentences that express one's feelings and that are effective for the other person. Therefore, the present invention is useful, for example, when sending feelings such as gratitude in a sentence.

1 Qualification information database 2 Fixed phrase information database 3 Dictionary information database 10 Recommendation device 11 Recording unit 12 Unsent sentence acquisition unit 13 Qualification information acquisition unit 14 Fixed phrase information acquisition unit 15 Selection unit 16 Extraction unit 161 Syntax analysis unit 162 Co-occurrence network formation unit 163 Dictionary information acquisition unit 17 Recommendation information generation unit 18 Recommendation unit 19 Evaluation information acquisition unit 20 Impact analysis unit 101 Central processing unit 102 Memory 103 Bus 104 Storage device 105 Input device 106 Display device 107 Communication device

Claims (11)

The method includes an unsent text acquisition unit, a qualification information acquisition unit, a template information acquisition unit, a selection unit, and a recommendation unit,
The unsent text acquisition unit acquires, from one user, an unsent feeling expression text and identification information of the other user to which the unsent feeling expression text is to be sent, in association with each other;
The qualification information acquisition unit acquires qualification information of the other user based on identification information of the other user,
The qualification information is information indicating qualifications,
The fixed phrase information acquisition unit acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection unit selects at least one of the plurality of fixed phrases based on the degree of influence,
The recommendation unit is a recommendation device that recommends to one user that the fixed phrase be used as the feeling expressing sentence. Further, the present invention includes an extraction unit and a recommended sentence generation unit,
the template selected by the selection unit is an incomplete sentence that prompts for input of at least one word,
The extraction unit extracts at least one key word from the unsent feeling-expressing sentence,
the recommended sentence generation unit inputs the key words into the template sentence to generate a recommended sentence;
The recommendation device according to claim 1 , wherein the recommendation unit recommends to the one user that the generated recommended sentence be used as the feeling expressing sentence. the extraction unit includes a syntax analysis unit, a co-occurrence network formation unit, and a dictionary information acquisition unit;
The parsing unit performs parsing of the unsent feeling expression sentence to extract words that are nouns and words that are adjectives;
the co-occurrence network forming unit forms a co-occurrence network with each of the extracted words as a vertex,
The dictionary information acquisition unit acquires dictionary information including words related to feelings,
The recommendation device according to claim 2 , wherein the extraction unit extracts, as the important word, a word that satisfies a preset condition based on the co-occurrence network and the dictionary information. Further, the system includes an evaluation information acquisition unit and an impact update unit,
When the one user transmits a feeling expressing sentence to the other user using at least one of the template sentence and the recommended sentence recommended by the recommendation unit,
The evaluation information acquisition unit acquires evaluation information indicating an evaluation of the other user with respect to the feeling expressing sentence,
The recommendation device according to claim 2 , wherein the influence updating unit updates the influence using the evaluation information. Further, the recording unit is included.
The recording unit links the user's identification information with the qualifications of the user and records them as qualification information;
The recommendation device according to claim 1 , wherein the qualification information acquisition unit acquires the recorded qualification information. Further, the recording unit is included.
The recording unit links the standard phrase, the quality information, and an influence degree for each standard phrase related to a feeling expression sentence, and records the standard phrase as standard phrase information;
The recommendation device according to claim 1 , wherein the fixed phrase information acquisition unit acquires the recorded fixed phrase information . The recommendation device according to any one of claims 1 to 6, wherein the feeling-expressing sentence is a sentence expressing a positive feeling. The recommendation device according to claim 7, wherein the positive emotion is gratitude. The method includes an unsent text acquisition step, a qualification information acquisition step, a template information acquisition step, a selection step, and a recommendation step;
The unsent text acquisition step acquires, from one user, an unsent feeling-expressing text and identification information of the other user to which the unsent feeling-expressing text is to be sent, in association with each other;
The qualification information acquisition step includes acquiring qualification information of the other user based on identification information of the other user,
The qualification information is information indicating qualifications,
The fixed phrase information acquiring step acquires fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
The selection step includes selecting at least one of the plurality of fixed phrases based on the degree of influence,
The recommendation method includes a step of recommending to one user that the template phrase be used as the feeling expression sentence. Further, the method includes an extraction step and a recommended sentence generation step,
the template selected by the selection step is an incomplete sentence that prompts input of at least one word,
The extraction step includes extracting at least one important word from the unsent feeling-expressing sentence,
The recommended sentence generating step inputs the key words into the template sentence to generate a recommended sentence;
The recommendation method according to claim 9 , wherein the recommending step recommends to the one user that the generated recommended sentence be used as the feeling expressing sentence. A program for causing a computer to execute procedures including an unsent text acquisition procedure, a qualification information acquisition procedure, a template information acquisition procedure, a selection procedure, and a recommendation procedure:
The unsent text acquisition step includes acquiring, from one user, an unsent emotion expression text and identification information of the other user to which the unsent emotion expression text is to be sent, in association with each other;
The qualification information acquisition step includes acquiring qualification information of the other user based on identification information of the other user;
The qualification information is information indicating qualifications,
The step of acquiring fixed phrase information includes acquiring fixed phrase information corresponding to qualification information of the other user,
The template information is information linking the template, the quality information, and an influence level for each template related to a feeling expression sentence,
the degree of influence is an index indicating the degree of psychological influence a user having the quality indicated by the quality information experiences when receiving the template message,
the selecting step includes selecting at least one of the plurality of the fixed phrases based on the degree of influence;
The recommendation step recommends to one user that the template phrase be used as the feeling-expressing sentence.