JP2011138343A - Electronic equipment and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve operability in a process that a user selects a pictograph arrangement place in a text. <P>SOLUTION: Electronic equipment is configured to, each time one sentence is input via a user interface, analyze the structure of an input sentence (S330), and to determine an addition candidate position as the candidate of a pictograph addition position according to the analytic result (S350), and to display a position selection screen showing the determined addition candidate position (S360). The position selection screen is configured such that the addition candidate position can be selected by a cursor operation. On the position selection screen, a cursor is moved to the addition candidate position corresponding to the arrangement pattern of the pictograph whose use probability is high (S390). Also, a pictograph selection screen configured by arraying pictograph candidates to be added to the addition candidate positions is displayed at the addition candidate position to which the cursor has been moved according to cursor movement. On the pictograph selection screen, the pictograph group is arrayed in the order of the pictograph which is closer to the past use instance. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an electronic device that can use pictograms and a program used for the device.

  2. Description of the Related Art Conventionally, as an electronic device that can use pictograms, a portable terminal device having an email sending / receiving function is known. According to this mobile terminal device, the user can select an arbitrary pictogram from the dictionary and input the pictogram by calling the dictionary in which the pictogram is registered at the place where the pictogram is to be placed when inputting the mail text. it can.

  In addition, as a conventional device, a mobile terminal device that analyzes received mail, estimates an emotion corresponding to the content of the received mail, and displays an emotion icon when an e-mail is received (for example, see Patent Document 1) An apparatus for creating an e-mail based on the input voice from the user, and corresponding to the emotion by estimating the emotion of the speaker from the input voice when creating the mail text based on the input voice from the user Devices that add pictographs to the mail text are known (see, for example, Patent Documents 2 and 3).

  In addition, as an e-mail creation device, a character string corresponding to a reading input by a user is presented as a conversion candidate in a priority order according to usage frequency, or a character string used in received mail is preferentially converted. The thing which improved the operativity which concerns on character input by showing as a candidate is known (for example, refer patent document 4).

JP 2007-148939 A JP 2007-004281 A JP 2006-277567 A JP 2006-344039 A

  By the way, as a technique for improving the operability related to pictogram input, as described above, the user's emotion is estimated and the pictogram corresponding to the emotion is automatically input at a fixed position (specifically, at the end of the mail text). There is a known level of technology that preferentially presents objects and frequently used pictograms.

That is, a technique that can improve user operability in the process of selecting a place to add pictograms in a sentence has not been known.
The present invention has been made in view of these problems, and an object of the present invention is to provide a technique capable of improving user operability in a process in which a user selects a place to add a pictogram in a sentence. .

  An electronic apparatus according to a first invention made to achieve the above object (Claim 1) includes a sentence acquisition means for acquiring a pictograph addition target sentence, a position candidate determination means, a first display control means, A second display control unit; and a pictograph adding unit.

  The position candidate determination means analyzes the sentence structure of the pictogram addition target sentence acquired by the sentence acquisition means, and determines a group of additional candidate positions that are candidates for the pictogram addition position in the pictogram addition target sentence. On the other hand, the first display control means displays a position selection screen (addition candidate position selection screen) showing a group of additional candidate positions in the pictogram addition target sentence on the display device, and the second display control means displays the position selection screen. When one additional candidate position is selected through the user interface from the group of additional candidate positions shown in, a pictogram selection screen is displayed in which pictogram groups that are pictogram candidates to be added to the selected additional candidate position are arranged. Display on the device. The pictograms here are not only those represented by image data, but also combinations of symbols such as parentheses (face characters, etc.) included in typeface data (fonts), and note marks “♪” and star marks included in typeface data. Including those expressed by a single symbol such as “★”.

  When one of the pictograms arranged on the pictogram selection screen is selected through the user interface, the pictogram adding means adds the pictogram to the selected addition candidate position in the pictogram addition target sentence, thereby adding the pictogram Edit the target sentence.

  According to this electronic device, by analyzing the sentence structure, an additional candidate position that is a candidate for a pictogram addition position in the pictogram addition target sentence is determined, and a position selection screen that can select any of the determined additional candidate positions is displayed. The user is prompted to select an additional candidate position through the user interface.

  Therefore, according to this electronic apparatus, the operability related to the selection of the pictogram addition position by the user is improved, and the convenience for adding pictograms in the text is improved. In particular, the present invention is effective in an environment using a user interface with poor operability.

  Note that the position selection screen can be a screen on which additional candidate positions can be selected by a cursor operation through a user interface, for example. In addition, the position selection screen can be a screen that displays a vocabulary for voice operation at each additional candidate position. According to this screen, the user can select any of the additional candidate positions through the microphone as the user interface by causing the user to utter the voice operation vocabulary of the additional candidate positions to be selected.

  In addition to this, the position candidate determining means is configured to determine, for example, the end of a sentence of a pictogram addition target sentence, the end of a pictogram addition target sentence, the independent word end of each phrase in the pictogram addition target sentence, and the like. (Claim 20 and Claim 21). In addition, the position candidate determination means may be configured to determine a replaceable position, which is a position where an independent word that can be replaced with a pictogram in the pictogram addition target sentence exists, as an additional candidate position (claim 22). When adding a pictograph to the replaceable position, the independent word existing at the replaceable position is deleted, and a pictograph is added instead of the independent word.

  Further, the position candidate determining means may be configured to determine a plurality of types of positions different in sentence structure as additional candidate positions (claim 2). According to this configuration, it is possible to increase the degree of freedom in selecting the added position of the pictogram, and it is possible to provide an electronic device that is highly convenient for the user. For example, the plural types of positions may include two or more positions of sentence end positions, sentence end positions, independent word end positions, and independent word positions.

  In addition, when the position candidate determination unit is configured to determine a plurality of types of positions as additional candidate positions, the electronic device has an additional candidate for each type corresponding to the operation of adding pictograms by the pictogram adding unit. It is preferable to provide a frequency recording means for recording the frequency of adding pictograms at positions on a recording medium.

  Further, the electronic device includes means for moving the cursor to a position corresponding to one of the additional candidate positions indicated on the position selection screen according to the cursor operation through the user interface, the above-described information recorded on the recording medium. Based on the addition frequency for each type, it is preferable to provide cursor position control means for initially placing the cursor at the additional candidate position of the type with the highest addition frequency.

  In this case, the second display control means can be configured to cause the display device to display the pictogram selection screen for the additional candidate position selected on the position selection screen through the cursor.

  According to this electronic device, since the cursor is initially placed at a position where the pictogram is frequently added on the position selection screen, statistically speaking, the user can perform a desired operation in the sentence with a small number of cursor operations. Emoji can be added to the position.

  Furthermore, the cursor position control means may be configured to move the cursor in order from an additional candidate position with a high frequency of addition to an additional candidate position with a low frequency of addition in accordance with a cursor operation through the user interface. (Claim 4).

  According to this configuration, since the selection cursor at the additional candidate position on the position selection screen moves in the order of the frequency of adding pictograms, it is possible to provide a highly convenient electronic device for the selection operation of pictogram addition positions. .

  By the way, many pictograms are registered in recent electronic devices (for example, mobile phone devices). For this reason, in the method of displaying all the registered pictograms on the pictogram selection screen, the operability related to the pictogram selection operation is impaired. Therefore, pictograms may be displayed on the pictogram selection screen by focusing on a part of the registered pictogram group.

  In the second invention (Claim 5), for the electronic device according to the first invention, a word in a pictogram addition target sentence corresponding to the additional candidate position is set as a keyword for each additional candidate position. A pictogram to be used as a candidate for a pictogram to be added to the additional candidate position (additional pictogram candidate) from a group of pictograms registered in advance based on the keyword setting means and the keyword set by the keyword setting means for each additional candidate position Pictograph candidate determining means for determining a group. When one of the additional candidate positions is selected through the user interface, the second display control means in the electronic device displays the pictogram group of the additional pictogram candidates for the additional candidate position selected through the user interface as the pictogram selection screen. An arranged pictogram selection screen is displayed on the display device.

  According to this electronic device, pictograms can be arranged on the pictogram selection screen by narrowing down to pictograms that are likely to be selected by the user. Therefore, it is possible to display the pictographs necessary for the user in an easy-to-see manner, and the operability related to the pictogram selection operation is improved.

  In addition, in order to arrange appropriate pictograms on the pictogram selection screen, a learning function may be provided in the electronic device. For example, the electronic device is provided with target sentence attribute specifying means for specifying the attribute of the pictograph addition target sentence. Furthermore, the identification information of the pictogram added to the pictogram addition target sentence by the pictogram adding means is expressed as “information indicating the keyword set by the keyword setting means for the additional candidate position corresponding to the position where the pictogram is added” and “ Case data registration means for registering in the pictogram example database as pictogram example data regarding use cases of pictograms is provided in association with “information representing the attributes of pictogram addition target sentences specified by the target sentence attribute specifying means”.

  Further, for the pictograph candidate determining means, for each additional candidate position, the combination of the keyword set by the keyword setting means for the additional candidate position and the attribute of the pictogram addition target sentence specified by the target sentence attribute specifying means, and the pictographic example Evaluate the degree of association with the combination of the keyword and attribute indicated by each pictogram case data registered in the database, and based on this evaluation value, preferentially add the pictogram indicated by the pictogram case data whose relevance is higher than a predetermined condition. The configuration is such that the candidate is determined (claim 6). The pictograph indicated by the pictographic case data whose relevance is higher than the predetermined condition is preferentially determined as an additional pictographic candidate with respect to the pictograph indicated by the pictographic case data whose relevance is lower than the predetermined condition.

  The pictograph candidate determination means, for example, is configured to preferentially determine the pictograph indicated by the pictograph case data whose relevance (evaluation value) is higher than a threshold value as an additional pictograph candidate, or pictographic case data with the evaluation value within the upper predetermined number Can be configured to preferentially determine the additional pictogram candidates.

  According to this electronic device, pictograms similar to past usage situations can be preferentially displayed on the pictogram selection screen, and the operability related to the pictogram selection operation is improved. Note that the target sentence attribute specifying means can be configured to specify the reader of the pictogram addition target sentence or the style of the pictogram addition target sentence as the attribute (claims 11 and 12).

  In addition, the case data registration means corresponds to the identification information of the pictogram added to the pictogram addition target sentence by the pictogram addition means, "information indicating the type of position where the pictogram was added" and "position where the pictogram was added. In association with “information representing the keyword set by the keyword setting means for the additional candidate position to be performed” and “information representing the attribute of the emoji adding target sentence specified by the target sentence attribute specifying means” More preferably, it is configured to be registered in the database.

  In this case, for each additional candidate position, the pictogram candidate determining means includes a combination of the type of the additional candidate position, the keyword set by the keyword setting means for the additional candidate position, and the attribute specified by the target sentence attribute specifying means. The degree of association with the combination of “type and keyword and attribute of the position where the pictogram is added” indicated by each pictogram example data registered in the pictogram case database is evaluated. Based on this evaluation value, the degree of association is determined from a predetermined condition. The pictograph indicated by the high pictographic example data can be preferentially determined as an additional pictograph candidate (claim 7).

  If the relevance is evaluated in consideration of the pictogram position in the sentence as described above, the user's preference for pictogram input can be learned, and appropriate pictograms can be arranged on the pictogram selection screen according to the additional candidate positions.

  The second display control means may be configured to display on the display device a pictograph selection screen in which pictographs are arranged in descending order of relevance. If the pictograms are arranged in this way and the pictogram selection screen is displayed on the display device, the pictogram selection screen can be displayed on the display device by arranging the pictograms in an appropriate order suitable for the use situation of the user, Convenience related to the selection operation of pictograms is improved.

  In addition, for an electronic device that uses a response text to an externally received text as a text to be added as a pictogram, the electronic equipment may be configured to determine pictograms arranged on the pictogram selection screen based on the received text.

  Specifically, in the electronic device, for each pictogram indicated by the received pictogram identifying means for identifying the pictogram used in the received sentence corresponding to the pictogram addition target sentence, and the pictogram example data for evaluating the degree of association, the co-occurrence database The information indicating the co-occurrence frequency of the pictogram and each pictogram specified by the received pictogram specifying means is acquired, and the evaluation value of the relevance by the pictogram candidate determining means for the pictogram based on the acquired information indicating the co-occurrence frequency And a correcting means for correcting. The co-occurrence database is a database that stores information indicating the frequency of co-occurrence between pictographs for each combination of pictographs.

  Then, the pictograph candidate determining means can be configured to preferentially determine the pictograph indicated by the pictographic case data whose relevance is higher than a predetermined condition as an additional pictographic candidate based on the evaluation value corrected by the correcting means (claim). Item 9).

  According to the electronic device having such a configuration, for example, in an environment where a pictogram corresponding to the pictogram used in the received mail is used in the reply mail, the pictogram selection screen is based on the pictogram used in the received sentence. Thus, it is possible to arrange appropriate pictograms, and the convenience of the electronic device is improved.

  In addition, for each pictogram determined as an additional pictogram candidate, the correction means acquires information representing the co-occurrence frequency of the pictogram and each pictogram specified by the received pictogram specifying means from the co-occurrence database. Based on the information indicating the co-occurrence frequency, the evaluation value of the degree of association by the pictogram candidate determination unit for the pictogram is corrected, and the second display control unit corrects the pictogram determined as the additional pictogram candidate based on the corrected evaluation value May be configured to display on the display device a pictogram selection screen formed by arranging the items in the descending order of relevance.

  If the electronic device is configured in this way, the pictograms can be arranged on the pictogram selection screen in an order that is easy for the user to use based on the pictograms used in the received sentence, and the convenience of the electronic device is improved.

  In addition, it is preferable to provide a function for registering pictogram example data based on the received text in the pictogram case database in an electronic device whose response text to the received text from the outside is a text to be added. That is, in the electronic device, a received pictogram specifying means, a received sentence attribute specifying means for specifying an attribute of the received sentence, and a word corresponding to the pictogram in the received sentence are set as keywords for each pictogram used in the received sentence. Sub-keyword setting means, and for each pictogram used in the received text, identification information of the pictogram is displayed as “information indicating the attribute of the received text in which the pictogram is used” and “sub-keyword setting means for the pictogram. It is preferable to provide sub-case data registration means for registering in the pictographic case database as pictographic case data in association with the keyword information set by “(Claim 13)”.

  If the electronic device is configured in this way, it is possible to display a group of pictograms according to the usage status of pictograms in a received sentence on the pictogram selection screen by evaluating the degree of association. The received sentence attribute specifying means can be configured to specify the writer of the received sentence and the style of the received sentence as the attribute (claims 15 and 16).

  In addition, the received pictogram specifying means may be configured to specify the pictogram used in the received text and to specify the type of the position of the pictogram in the received text for each pictogram used in the received text. . Examples of the position of the pictogram include a sentence end, a sentence end, and an independent word end.

  Then, the reception case data registration means, for each pictogram specified by the received pictogram specifying means, identifies the pictogram identification information as “information indicating the type of position of the pictogram specified by the received pictogram specifying means” and “ A pictograph as pictographic example data in association with “information indicating the attribute of the received text in which the pictogram specified by the received text attribute specifying unit is used” and “information of the keyword set by the sub-keyword setting unit for the pictograph” It may be configured to be registered in the case database (claim 14).

If the electronic device is configured in this way, it is possible to display a group of pictograms according to the usage status of pictograms in the received text on the pictogram selection screen in consideration of the location of the pictograms in the received text.
In addition, the third invention (Claim 17) is provided with the target sentence attribute specifying means, the keyword setting means, and the case data registration means for the electronic device according to the first invention, and for each additional candidate position, The combination of the keyword set by the keyword setting means for the additional candidate position and the attribute of the pictogram addition target sentence specified by the target sentence attribute specifying means, and the combination of the keyword and attribute indicated by each pictogram case data registered in the pictogram case database The relevance level evaluation means for evaluating the relevance level is provided, and the second display control means is configured as follows.

  In other words, based on the evaluation result of the relevance evaluation means for the additional candidate position selected through the user interface, the pictogram selection screen is preferentially assigned to the pictograms indicated by the pictogram example data with high relevance. The second display control means is configured to assign an arrangement order to each pictogram arranged in the sequence, and to display on the display device a pictogram selection screen in which pictogram groups are arranged in order from the pictogram to which the young arrangement order is assigned. (Claim 17).

If the electronic device is configured in this way, the pictograms can be arranged in the order corresponding to the past pictogram usage state on the pictogram selection screen, and the convenience of the pictogram selection operation is improved.
The fourth invention (invention 18) is directed to the electronic device according to the first invention, on the premise that the electronic device is an electronic device whose response text to an externally received sentence is a text addition target sentence. Thus, received pictogram specifying means for specifying the pictogram used in the received text corresponding to the pictogram addition target sentence is provided, and the second display control means is configured as follows.

  That is, for each pictogram arranged on the pictogram selection screen, information representing the co-occurrence frequency of the pictogram and each pictogram identified by the received pictogram identifying means is obtained from the co-occurrence database, and information representing these co-occurrence frequencies Based on the above, a pictogram selection screen in which pictograms arranged on the pictogram selection screen are preferentially assigned a young layout order from those having the highest co-occurrence frequency, and a pictogram group is arranged in order from the pictograms assigned the young layout order The second display control means is configured to display on the display device (claim 18).

  According to this electronic apparatus, the pictogram group can be arranged and displayed on the pictogram selection screen in the order corresponding to the usage status of the pictogram in the received sentence, so that the operability related to the pictogram selection operation is improved.

  In this electronic device, information indicating the co-occurrence frequency of the pictogram added to the pictogram addition target sentence and the pictogram used in the corresponding received sentence in correspondence with the pictogram adding operation by the pictogram adding means. May be provided with a co-occurrence frequency updating means for updating the same in the co-occurrence database (claim 19).

  If the electronic device is configured in this way, it is possible to learn the usage mode of pictograms by the user, arrange pictogram groups in a more appropriate order according to the usage status of pictograms in the received sentence, and display them on the pictogram selection screen. it can.

  In addition, the functions of each means provided in the electronic device described above can be realized by a computer by a program (claim 23). A program for causing a computer to realize the functions as the above means can be provided by being stored in a recording medium (CD-ROM or DVD) or can be provided through an electric communication line.

2 is a block diagram illustrating a configuration of an electronic device 1. FIG. It is the figure which showed the kind of data which the program which the control part 10 performs, and the memory | storage part 70 memorize | store. It is a flowchart showing the reception data analysis process which the control part 10 performs. It is explanatory drawing regarding the keyword setting method corresponding to a pictogram arrangement pattern. It is a figure showing the structure of a pictogram dictionary. It is a figure showing the structure of a feature vector and pictogram example data. It is a figure showing the structure of pattern statistical data. 3 is a flowchart illustrating an input conversion process executed by a control unit 10. It is a flowchart showing the position selection cursor control process which the control part 10 performs. It is explanatory drawing regarding the movement aspect of a position selection cursor. It is a flowchart showing the pictogram selection screen display control process which the control part 10 performs. It is a figure showing the structure of a pictograph selection screen. It is a flowchart showing the relevance degree correction process which the control part 10 performs. It is a figure showing the structure of co-occurrence DB. It is a flowchart showing the co-occurrence learning process which the control part 10 performs. It is a flowchart showing the pictogram selection screen display control process of 2nd Example. It is a flowchart showing the pictogram selection screen display control process of 3rd Example. It is a flowchart showing the arrangement | sequence order determination process which the control part 10 performs.

  Embodiments of the present invention will be described below with reference to the drawings.

  As shown in FIG. 1, the electronic device 1 according to the first embodiment includes a control unit 10 including a CPU 11, a ROM 12, a RAM 13, and the like, a wireless communication unit 20 capable of wireless communication with an external network, a liquid crystal display, and the like. Display unit 30, an operation unit 40 having operation keys such as cursor keys that can be operated by the user, an audio output unit 50 including a speaker, an audio input unit 60 including a microphone, and a hard disk And a storage unit 70 for storing various data including a device and a flash memory.

  Examples of the electronic device 1 include an in-vehicle device such as a car navigation device and a mobile phone device. In the following, it is assumed that the electronic device 1 is a device capable of performing voice operation through the voice input unit 60.

  The control unit 10 in the electronic device 1 implements a function to be performed as the electronic device 1 by executing a program recorded in the ROM 12 by the CPU 11. Specifically, the ROM 12 can browse blog articles registered in an external blog server or mail software (hereinafter referred to as “app A”) that is a program capable of creating and sending / receiving e-mails. An application program such as a browser (hereinafter referred to as “application B”), which is a program that can post a comment sentence, is stored in the comment column of the article. Note that transmission / reception of electronic mail and access to the blog server are realized via the wireless communication unit 20 connected to an external network.

  The ROM 12 also has an input conversion program having a kana-kanji conversion function used when inputting characters in the apps A and B, received data for analyzing an e-mail received by the app A and a blog article received by the app B. An analysis program is stored.

  In addition, as shown in FIG. 2, the storage unit 70 co-occurs with a kana-kanji conversion dictionary, an emotion estimation dictionary, a pictographic dictionary, and a pictographic case DB (database) including pictographic case data groups representing use cases of pictographs. A co-occurrence DB (database) composed of a co-occurrence data group indicating a combination of pictograms, a user DB (database), and pattern statistical data are stored.

  Details of the emotion estimation dictionary, pictogram dictionary, pictogram example DB, co-occurrence DB, user DB, and pattern statistical data will be described as appropriate in appropriate sections. First, the control unit 10 according to the received data analysis program The received data analysis process executed by will be described. The control unit 10 repeatedly executes the received data analysis process shown in FIG. 3 during the operation of the electronic device 1.

  When the received data analysis process is started, the control unit 10 sequentially acquires text data received from the outside by the tasks of the applications A and B, which are monitoring target application programs. Specifically, a blog article in which an application A task, which is an application program to be monitored, sequentially acquires emails received from an external mail server through the wireless communication unit 20, and an application B task is received from an external blog server. Are sequentially acquired (S100). Then, every time a sentence end is detected in the acquired data (Yes in S110), one sentence up to the detected sentence end is set as a processing target sentence, and the processes after S120 are executed.

  In S120, the attribute of the processing target sentence is specified. Specifically, the style and writer of the processing target sentence are specified as attributes. With respect to the style, the processing target sentence acquisition source specifies whether the style of the processing target sentence is the “mail” style or the “blog” style. That is, when the processing target sentence is acquired from the app A, the style is specified as “mail” style, and when the processing target sentence is acquired from the application B, the style is “blog”. "Identify the style."

  In addition, the writer is identified by inquiring the writer of the processing target sentence from the task of the application A or the task of the application B that acquires the processing target sentence, and acquiring the writer's identification information from the inquiry source. The writer identification information obtained from the task of the app A is the mail address of the processing target sentence transmission source, and the writer identification information obtained from the task of the app B is the user name of the blog article poster.

  In S130 following S120, the syntax of the processing target sentence is analyzed. Specifically, the part-of-speech analysis is performed on the sentence to be processed, and the phrase and independent words in the phrase are detected. For example, in the case where the sentence to be processed is a sentence “Today is raining”, each part of speech of “today”, “ha”, “rain”, “da”, “yo” is specified, and the independent word “today” and The phrase “Today is” consisting of the appendix “ha” and the phrases “rain” that are independent words “rain” and the appendages “da” and “yo” are detected. Note that pictorial characters may be included in the processing target sentence, but the parsing is performed by ignoring the pictograms included in the processing target sentence. However, for pictograms sandwiched between attached words, this pictogram is regarded as an independent word and a phrase is detected.

  In S140 following S130, the emotion estimation of the writer of the processing target sentence is performed. Since various methods are known as emotion estimation methods, detailed description is omitted here, but in this embodiment, for example, the correspondence relationship between a keyword and an emotion word that is a word representing emotion is defined. The emotion estimation is realized by referring to the emotion estimation dictionary and specifying an emotion word having the word constituting the processing target sentence as a keyword. Examples of emotion words include “happy”, “easy”, “fun”, and “unpleasant (anger)”.

  In S145 following S140, it is determined whether or not a pictograph exists in the processing target sentence. If a pictograph exists (Yes in S145), the process proceeds to S150. If no pictograph exists (No in S145). ), The process proceeds to S220.

After shifting to S150, the control unit 10 selects one of the unselected pictograms as the processing target pictogram included in the processing target sentence as the processing target pictogram, and executes the processing after S160.
In S160, the arrangement pattern of the processing target pictogram in the processing target sentence is specified. Specifically, the processing target text of the processing target pictogram is determined by determining whether the layout pattern of the processing target pictogram is “replacement”, “word end positioning”, “sentence end positioning”, or “sentence end positioning”. The arrangement pattern in is specified.

  For example, when the processing target pictogram is arranged at the end of the sentence, it is determined that the arrangement pattern is “end of sentence arrangement”. On the other hand, when the processing target pictogram is arranged at the end of a sentence other than the end of the sentence, it is determined that the arrangement pattern is “end of sentence arrangement”. If the processing target pictogram is placed at the end of an independent word other than the end of a sentence and the end of a sentence, it is determined that the placement pattern is “end of word placement”. In addition, when the processing target pictogram is placed between the attached words, it is determined that the placement pattern is “replacement”.

  In S170 following S160, a keyword corresponding to the specified arrangement pattern is set for the processing target pictogram. For example, when the arrangement pattern is “word end arrangement”, as shown in FIG. 4A, the independent word immediately before the processing target pictogram in the processing target sentence is set as the keyword. On the other hand, when the arrangement pattern is “end of phrase arrangement”, as shown in FIG. 4B, an independent word in the phrase in which the processing target pictogram in the processing target sentence is arranged is set as a keyword.

  When the arrangement pattern is “replacement”, the representative word of the processing target pictogram is specified with reference to the pictogram dictionary, and the representative word is set as a keyword as shown in FIG. In the example of FIG. 4C, “rain” is specified as the representative word of the umbrella mark, and this is set as the keyword. As shown in FIG. 5, the pictogram dictionary is a keyword composed of image data as a pictogram body, an identification ID of the pictogram, a representative word corresponding to the pictogram, and a related word for each pictogram mounted on the electronic device 1. Memorize groups. The representative word is a word corresponding to the representative meaning of the pictogram, and the related word is a word associated with the pictogram. For example, when the pictogram is the umbrella mark shown in FIG. 4C, “rain” is registered in the pictogram dictionary as a representative word, and words associated with pictograms such as “umbrella” and “kanai” are related words. be registered.

  In addition, when the arrangement pattern is “end-of-sentence arrangement”, as shown in FIG. 4D, all independent words in the processing target sentence are set as keywords, and the emotion word estimated in S140 is set as a keyword. Set.

  Further, in S180 following S170, the features of the processing target pictograms are expressed as vectors based on the keyword group set in the processing target pictograms in S170, the style and writer specified in S120, and the processing target pictograms specified in S160. A feature vector is generated. As shown in FIG. 6, the feature vector includes an element for each predetermined keyword, an element group representing a style, an element group representing an arrangement pattern of pictograms, and an element group representing a user related to the pictograms. Is a multidimensional vector consisting of However, the value of each element is expressed in binary and takes 0 or 1.

  According to the example shown in FIG. 6, in the feature vector, the feature amount of the style is represented by three elements, the “mail” style is represented by “001”, and the “blog” style is represented by “010”. The feature quantity of the arrangement pattern is represented by four elements, “replacement” is represented by “0001”, “word end arrangement” is represented by “0010”, and “sentence end arrangement” is represented by “0100”. The “end of sentence arrangement” is expressed as “1000”.

  In addition, the user's feature quantity related to the pictogram is expressed by a 6-digit user code using six elements. For example, the user Ua is expressed as “000001”, the user Ub is expressed as “000010”, the user Uc is expressed as “000100”, the user Ud is expressed as “001000”, and the user Ue is expressed as “010000”. Many other users are represented by “100000”.

  The feature vector generation method will be described in detail. When the processing target sentence is the sentence shown in FIG. 4D, the keywords “today”, “rain”, and “easy” set in S170 in the feature vector correspond to each. Set the element value to 1 and set the other keyword elements to zero. Further, the element group representing the style in the feature vector is set to a value (for example, “001”) corresponding to the style of the sentence to be processed, and the element group representing the arrangement pattern is set to “1000” corresponding to “end sentence arrangement”. Set. Further, the 6-digit user code corresponding to the writer specified in S120 is set for the element group representing the user related to the pictogram.

  In the storage unit 70, a user DB (see FIG. 2) indicating the correspondence between the user's mail address or user name and the user code is registered. In S180, the mail address of the writer specified in S120 or When the user name is registered in the user DB, the user code assigned to the mail address or the user name is set as the corresponding element of the feature vector. On the other hand, if the mail address or user name of the writer specified in S120 is not registered in the user DB, “100000” representing many other users is set as the corresponding element of the feature vector as the user code.

Thus, in S180, a feature vector corresponding to the processing target pictogram is generated.
After the generation of the feature vector, the process proceeds to S190 to generate pictographic case data in which the generated feature vector is associated with the identification ID of the processing target pictograph (see FIG. 6), which is stored in the pictographic case DB of the storage unit 70. (S200). The pictogram case DB is composed of a group of pictogram case data having such a configuration, and is updated when the pictogram case data is registered as the electronic device 1 is used.

  After the processing of S200, all pictograms included in the processing target sentence are selected as processing target pictograms, and it is determined whether or not the processing of S150 to S200 has been executed (S210). (No in S210), the process proceeds to S150, an unprocessed pictograph is newly selected as a processing target pictograph, and the processes after S160 are executed.

  When there are a plurality of pictograms in the processing target sentence as shown in FIG. 4E by such processing, for each pictogram, an arrangement pattern specifying operation, a keyword setting operation corresponding to the arrangement pattern, a feature vector Generation operation and generation / registration operation of pictogram example data are performed. And when the process of S150-S200 is performed about all the pictograms contained in a process target sentence, an affirmation judgment is carried out in S210 and it transfers to S220.

In S220, the values Y1, Y2, Y3, Y4, and Y5 of the pattern statistical data (see FIG. 7) stored in the storage unit 70 are updated according to the arrangement pattern of pictograms existing in the processing target sentence.
As shown in FIG. 7, the pattern statistical data includes values Y1, Y2, Y3, Y4, and Y5 corresponding to the arrangement pattern classifications “replacement”, “word end placement”, “sentence end placement”, “sentence end placement”, and “no pictogram”. Have The value Y1 held in the pattern statistical data indicates a value corresponding to the appearance probability P1 of a sentence having a pictograph whose placement pattern is “replacement”, and the value Y2 held in the pattern statistical data is the placement pattern “word end placement”. The value Y3 held in the pattern statistical data indicates a value corresponding to the appearance probability P3 of a sentence having a pictographic character whose arrangement pattern is “end of paragraph arrangement”. The value Y4 held by the pattern statistical data indicates a value corresponding to the appearance probability P4 of a sentence having a pictograph whose arrangement pattern is “end of sentence arrangement”, and the value Y5 held by the pattern statistical data has no pictogram A value corresponding to the appearance probability P5 of the sentence is shown.

  In S220, among the values Y1, Y2, Y3, Y4, and Y5, the pattern statistical data is obtained by adding the fixed value X to the values Y1, Y2, Y3, Y4, and Y5 corresponding to the arrangement pattern of pictograms existing in the processing target sentence. Update. However, when a plurality of pictograms exist in different arrangement patterns in the processing target sentence, the fixed value X is divided by the number M of types of arrangement patterns existing in the processing target sentence, and the value X / M after the division is processed. The pattern statistical data is updated by adding to the values Y1, Y2, Y3, Y4, and Y5 corresponding to the arrangement pattern of pictograms existing in the sentence.

  For example, based on the processing target sentence shown in FIG. 4A, a fixed value X is added to the value Y2 of “word end arrangement”, and the other values Y1, Y3, Y4, Y5 are not changed, The pattern statistical data is updated (S220). On the other hand, if there is no pictogram in the processing target sentence, the fixed value X is added to the value Y5 of “no pictogram”, and the other values Y1, Y2, Y3, Y4 are not changed, and the pattern statistical data Update.

In addition, when the processing target sentence includes, for example, a pictograph that is “placed at the end of word” and a pictograph that is placed at the end of the sentence, the fixed value X is divided by the number of types M = 2 of the arrangement pattern, and after division Is added to the values Y2 and Y4.
Y2 ← Y2 + X / 2
Y4 ← Y2 + X / 2
On the other hand, the pattern statistical data is updated without changing the other values Y1, Y3, and Y5. Note that the value Y5 is not updated unless there is a pictogram in the text. Further, as in the example shown in FIG. 4 (e), even if there are a plurality of pictograms in the processing target sentence, when the arrangement pattern of these pictograms is the same, the number of arrangement pattern types M = 1. Handle and update pattern statistics data as described above.

  Incidentally, in the present embodiment, when a plurality of types of arrangement patterns are mixed, the pattern statistical data is updated using the value X / M, and therefore any value that M can take is divisible for the fixed value X. Set to value. For example, according to the present embodiment, X = 12 can be set.

The control unit 10 repeatedly executes the above-described processes of S100 to S220 as the received data analysis process, and updates the pictographic example DB and the pattern statistical data.
Next, the input conversion process executed by the control unit 10 according to the input conversion program will be described. The control unit 10 repeatedly performs the input conversion process shown in FIG. 8 during the operation of the applications A and B that involve character input.

  When the input conversion process is started, the control unit 10 receives character input from the user interface via the application A or application B, which is the application program that is the character input source (S300), and converts the input character to Kana-Kanji (S305). ). The character input result and the kana-kanji conversion result are displayed at the character input position on the display screen of the apps A and B in cooperation with the apps A and B.

  The user interface here is the operation unit 40 or the voice input unit 60. In the electronic device 1 of the present embodiment, a voice recognition function is built in the voice input unit 60, and when a character is input by voice, the voice input through the microphone is converted into a character (kana) by the voice recognition function. Passed to processing.

  In this input conversion process, when the end of a sentence is detected (Yes in S310), one sentence that has been input until the end of the sentence is detected is set as a process target sentence, and the processes after S320 are executed. However, in the present embodiment, when the input position of the subject and the mail body input feed provided by the application A at the time of creating the e-mail is set, and when the comment text for the blog article is input, the application B provides it. Only when the character input position is set in the comment text input feed, the process proceeds to S320, and in other cases, the Kana-Kanji conversion process is repeated in accordance with the character input in the same manner as the normal input conversion program ( S300 to S305).

  In step S320, the control unit 10 first identifies the attribute of the processing target sentence. Specifically, the style of the sentence to be processed and the reader are specified. When the character input target is the app A, it is determined that the style of the processing target sentence is “mail” style. When the character input target is the app B, the style of the processing target sentence is “blog” style. Judge.

  The reader is identified by acquiring reader identification information from the app A or app B that is the character input target. The reader identification information that can be acquired from the app A is the email address of the e-mail transmission destination, and the reader identification information that can be acquired from the app B is a user name that represents the poster of the blog article corresponding to the comment text. .

  In S330 following S320, syntax analysis of the processing target sentence is performed in the same manner as in S130. In this embodiment, since pictogram input is managed by input conversion processing, pictograms are not included in the processing target sentence at the time of execution of S330.

  In S340 subsequent to S330, the emotion estimation of the writer of the processing target sentence is performed in the same manner as in S140. However, when character input is performed by voice input, emotion estimation may be performed based on the input voice.

  When the emotion estimation is performed in S340, the control unit 10 may proceed to S350 and replace the position corresponding to the end of the sentence of the processing target sentence found by the syntax analysis, the position corresponding to the independent end of each phrase, and the pictogram. The position corresponding to the independent word and the position corresponding to the end of the sentence are set as additional candidate positions that are candidates for the pictogram addition position.

  Then, according to the setting result of the additional candidate position, a position selection screen that is a screen on which any one of the additional candidate positions can be selected by a cursor operation in the processing target sentence display area corresponding to the character input position on the display unit 30. It is displayed (S360). For example, the position selection screen is displayed with the configuration shown in FIG.

  The position selection screen shown in FIG. 8 displays a processing target sentence, and is a dotted line frame (a dotted line frame indicating that the cursor can be moved to each additional candidate position where a pictogram can be placed in the processing target sentence ( (Corresponding to the dotted line in FIG. 8) and an alphabet for selection operation by voice is displayed for each dotted line frame. When the cursor is operated for the first time, the cursor is displayed as a solid line frame in alignment with one of the dotted line frames on the position selection screen (see FIG. 10). Hereinafter, this cursor is referred to as a “position selection cursor”.

  After the display of the position selection screen, the control unit 10 waits until one of a plurality of predetermined operations is performed through the user interface (S370), and when an operation is performed, the process proceeds to S380 and later. The process corresponding to the performed operation is executed.

  For example, when the movement operation of the position selection cursor is performed (Yes in S380), the control unit 10 proceeds to S390 and executes the position selection cursor control process shown in FIG. Thus, a process of moving the position selection cursor on the position selection screen in accordance with the movement operation of the position selection cursor is performed. Thereafter, the process proceeds to S370 and waits until the next operation is performed.

  When the position selection cursor control process is started in S390, the control unit 10 determines whether or not the movement operation of the position selection cursor performed is an operation by voice (S610). Here, when one of the alphabets attached to each dotted line frame on the position selection screen is input by voice, an affirmative determination is made (Yes in S610). Then, a process of moving the position selection cursor to the dotted frame on the position selection screen corresponding to the alphabet input by voice is performed (S620). Thereafter, the process proceeds to S720.

  On the other hand, if the movement operation of the position selection cursor is not a voice operation (No in S610), it is determined whether or not the movement operation of the position selection cursor performed is the first operation after the display of the position selection screen ( S630). When it is determined that the operation is the first operation (Yes in S630), the additional candidate positions where the pictograms are arranged in the arrangement pattern having the highest appearance probability P1 to P5 with reference to the values Y1 to Y5 of the pattern statistical data. Then, the position selection cursor is displayed (initially arranged) in the dotted line frame of the additional candidate position closest to the head of the sentence to be processed (S640). Thereafter, the process proceeds to S720. However, in the position selection cursor control process, the arrangement pattern of “no pictogram” is ignored and the processes after S640 are performed.

  In addition, when the movement operation of the position selection cursor performed is the second or later operation (No in S630), whether the movement operation of the position selection cursor is a movement operation in the “positive” direction or “negative”. It is determined whether or not the movement operation is in the direction (S650).

  If it is determined that the operation is a movement operation in the “positive” direction (S650), the same position as the additional candidate position corresponding to the current position of the position selection cursor is set in the sentence end direction of the processing target sentence rather than the current position of the position selection cursor. It is determined whether there is an additional candidate position where a pictograph is arranged in the pattern (S660).

  If it is determined that the corresponding additional candidate position exists (Yes in S660), the pictograph is arranged in the same arrangement pattern as the additional candidate position corresponding to the current position of the position selection cursor. A process of moving the position selection cursor to the dotted line frame of the additional candidate position closest to the current position that is “additional candidate position existing in the sentence end direction of the target sentence” is performed (S670). Thereafter, the process proceeds to S720.

  On the other hand, if it is determined that the corresponding additional candidate position does not exist (No in S660), it is an additional candidate position corresponding to an arrangement pattern having a lower appearance probability than the arrangement pattern of pictograms corresponding to the current position of the position selection cursor. Then, a process of moving the position selection cursor to the dotted line frame of the additional candidate position existing at the position closest to the beginning of the sentence is performed (S680). Thereafter, the process proceeds to S720.

  In addition, if the control unit 10 determines that the movement operation of the position selection cursor performed is a movement operation in the “negative” direction, the control unit 10 proceeds to S690 and starts the sentence head of the processing target sentence rather than the current position of the position selection cursor. It is determined whether or not there is an additional candidate position in the direction where a pictograph is arranged in the same arrangement pattern as the additional candidate position corresponding to the current position of the position selection cursor.

  When it is determined that the corresponding additional candidate position exists (Yes in S690), the pictograph is arranged in the same arrangement pattern as the additional candidate position corresponding to the current position of the position selection cursor. A process of moving the position selection cursor to the dotted line frame of the additional candidate position closest to the current position which is the “additional candidate position existing in the head direction of the target sentence” is performed (S700). Thereafter, the process proceeds to S720.

  On the other hand, if it is determined that the corresponding additional candidate position does not exist (No in S690), it is an additional candidate position corresponding to an arrangement pattern having a higher appearance probability than the arrangement pattern of pictograms corresponding to the current position of the position selection cursor. Then, a process of moving the position selection cursor to the dotted line frame of the additional candidate position present at the position closest to the end of the sentence is performed (S710). Thereafter, the process proceeds to S720.

  In S720, the pictogram selection screen display control process shown in FIG. 11 is executed, and the position selection cursor is placed on the pictogram selection screen in which pictogram groups corresponding to the additional candidate positions to which the position selection cursor is placed are arranged. Displayed around the additional candidate position. Thereafter, the position selection cursor control process is terminated.

  Here, the movement mode of the position selection cursor will be described with reference to FIG. Here, the relationship between the appearance probabilities P1 to P4 of each arrangement pattern “replacement”, “word end arrangement”, “sentence end arrangement”, and “sentence end arrangement” is the occurrence probability of “sentence end arrangement” P4> appearance of “word end arrangement” A movement mode of the position selection cursor when probability P2> appearance probability P1> replacement probability P1> appearance probability P3 of “end of phrase placement” will be described. However, regarding the movement operation of the position selection cursor, the movement operation in the “forward” direction is always performed.

  In this case, each time the position selection cursor is moved, the position selection cursor moves in the order of the dotted frame F, the dotted frame B, the dotted frame E, the dotted frame A, the dotted frame D, and the dotted frame C. In the present embodiment, the position selection cursor can be moved to the desired position by moving the position selection cursor with a small number of users by moving the position selection cursor in descending order of appearance probability.

  Next, the pictogram selection screen control process executed by the control unit 10 in S720 will be described with reference to FIG. When the pictogram selection screen control process is started, the control unit 10 sets a keyword for an additional candidate position where the position selection cursor is aligned.

  The keyword setting method is the same as the processing in S170. That is, when the arrangement pattern of pictograms arranged at the additional candidate position is “word end arrangement”, the independent word immediately before the additional candidate position in the processing target sentence is set as the keyword at the additional candidate position. Further, when the arrangement pattern is “end of phrase arrangement”, an independent word in the phrase corresponding to the additional candidate position is set as a keyword at the additional candidate position.

  In addition, when the arrangement pattern is “replacement”, the independent word existing at the additional candidate position is set as a keyword, and when the arrangement pattern is “end of sentence arrangement”, all the independent words of the processing target sentence and The emotion word estimated in S340 is set as a keyword at this additional candidate position.

  After the processing in S810 is completed, the additional candidate is determined based on the keyword group set in S810, the style and reader specified in S320, and the arrangement pattern of the pictograph at the additional candidate position where the position selection cursor is aligned. A feature vector formed by expressing the feature of the position as a vector is generated. Note that the feature vector generation method is the same as the process in S180, and the configuration of the feature vector is as shown in FIG. However, in S820, a 6-digit user code corresponding to the reader specified in S320 is set for the element group representing the feature vector user.

  When the feature vector Va of the additional candidate position where the position selection cursor is aligned is generated in this way, the control unit 10 includes the feature vector Va and the feature vector Vb indicated by each pictogram case data registered in the pictogram case DB. The cosine distance (Va · Vb / | Va | · | Vb |) is calculated to evaluate the similarity between the feature vector Va and the feature vector Vb (S830).

  That is, in the present embodiment, the degree of relevance is evaluated by evaluating the degree of similarity between the feature vector Va and the feature vector Vb. At this time, the cosine distance between the feature vectors Va and Vb is calculated. Hereinafter, the evaluation value D expressed by the cosine distance is simply referred to as the degree of association D.

  In S840 following S830, the pictogram indicated by the pictogram case data having the relevance D equal to or higher than the threshold TH is displayed on the pictogram selection screen based on the relevance D calculated for each pictogram case data registered in the pictogram case DB. It is set to the target pictogram (hereinafter referred to as “display target pictogram”). The threshold value TH can be determined in advance by the designer.

  In S850 following S840, it is determined whether or not there are J or more pictograms set as display target pictograms, and if there are J or more (Yes in S850), S860 is skipped. If the number is less than J (No in S850), the process proceeds to S870 after executing the process of S860.

  In S860, a pictographic character having at least one of the keywords set at the additional candidate position where the position selection cursor is placed in the keyword group representing the representative word and the related word is extracted and extracted in the pictogram dictionary. Each pictogram is additionally set as a display target pictogram. Further, for each pictogram additionally set as a display target pictogram, a value D0 smaller than the threshold TH by a predetermined amount is assigned as the degree of association D. However, in S860, the above-mentioned pictograms extracted from the pictogram dictionary are additionally set as display target pictograms so as to avoid duplication of the same pictogram as the display target pictogram set in S840.

  In S870, the control unit 10 executes the relevance level correction process shown in FIG. 13 to correct the relevance level D calculated (or given) to the display target pictograms in S830 and S860. Thereafter, based on the relevance degree D after correction, a pictogram selection screen in which display target pictograms are arranged in descending order of the relevance degree D is displayed around the additional candidate position where the position selection cursor is aligned (S880). Thereafter, the pictogram selection screen display control process is terminated.

  FIG. 12 is an explanatory diagram showing the configuration of the pictogram selection screen, and FIG. 12A is a diagram showing the display mode of the pictogram selection screen in a state where the position selection cursor is aligned with the dotted frame D. 12B is a diagram illustrating a display mode of the pictograph selection screen in a state where the position selection cursor is aligned with the dotted frame F. FIG.

  As shown in FIG. 12, the pictograph selection screen has a configuration in which a plurality of pictographs are arranged in a line, and one of pictographs can be selected by a cursor operation. The cursor in the pictogram selection screen (hereinafter referred to as the “pictogram selection cursor”) is arranged in accordance with the pictogram displayed on the leftmost in the initial state, and in order of the pictogram arrangement direction by the cursor operation by the user. Moving. Because the pictograph selection screen is configured as such a graphical user interface, display target pictographs are arranged from left to right in descending order of relevance D on the pictograph selection screen.

  In addition, a number for selection operation by voice is assigned to each pictogram displayed on the pictogram selection screen. Pictogram selection by voice is realized by utterance of numbers assigned to pictograms.

  The pictogram selection screen can be configured to be scrollable in the pictogram arrangement direction. For pictograms that cannot be displayed at a time, the pictograms are displayed on the pictogram selection screen by a scroll operation through the user interface by the user.

  Next, the relevance correction process executed by the control unit 10 in S870 will be described with reference to FIG. When the relevance correction process is started, the control unit 10 determines whether or not the processing target sentence is a sentence constituting the reply mail (that is, a reply sentence to the received mail) (S910), and the processing target sentence is the reply mail. (Yes in S910), all the pictograms used in the received mail corresponding to the reply mail are extracted from the received mail (S920). Thereafter, the process proceeds to S950.

  On the other hand, when the processing target sentence is not a sentence constituting the reply mail (No in S910), the control unit 10 determines whether the processing target sentence is a sentence constituting a comment sentence for the blog article (S930). . If it is determined that it is not a sentence that constitutes a comment sentence (No in S930), the relevancy correction process is terminated, and if it is determined that the sentence is a sentence that constitutes a comment sentence (Yes in S930), the blog corresponding to the comment sentence All the pictograms used in the article are extracted from the corresponding blog article (S940). Thereafter, the process proceeds to S950.

  When the process proceeds to S950, the control unit 10 sets one of the pictograms set as the display target pictogram in S840 and S860 as the processing target pictogram, and co-occurs with the processing target pictogram in the pictogram group extracted in S920 or S940. All the pictograms registered in the co-occurrence DB as pictograms are extracted as reference pictograms (S960). As shown in FIG. 14, the co-occurrence DB is a co-occurrence including information representing a combination of pictograms that co-occur, information on correction coefficient k, and information on the number of co-occurrence q, for each combination of pictograms that co-occur. It has data. The correction coefficient k (≧ 1) of each co-occurrence data constituting the co-occurrence DB shows a higher value as the combination of pictograms having higher co-occurrence frequency (in other words, higher probability of co-occurrence).

  In S970 following S960, the correction coefficient k set for the combination of the reference pictogram and the processing target pictogram is read from the co-occurrence DB for each combination of the reference pictogram and the processing target pictogram. Thereafter, the degree of association D obtained in S830, which is the degree of association D between the feature vector Vb of the pictographic case data associated with the processing target pictogram and the feature vector Va of the additional candidate position, is based on the read correction coefficient k. (S980). Specifically, the degree of relevance D obtained in S830 is obtained by applying the correction coefficient k indicating the maximum value of the correction coefficients k for each read reference pictogram to the degree of relevance D corresponding to the processing target pictogram. , D ← k · D is corrected.

In S990 following S980, the reference pictogram extracted in S960 is temporarily stored in the RAM 13 as co-occurrence learning data for the processing target pictogram (S990).
Thereafter, it is determined whether or not all of the display target pictograms have been selected as processing target pictograms (S995). If not selected (No in S995), the process proceeds to S950 and new display target pictograms are processed. The target pictogram is selected and the process proceeds to S960.

  In this manner, in S950 to S990, for each display target pictogram, the degree of association D obtained in S830 for the display target pictogram is corrected, and further, a reference pictogram having a relationship that co-occurs with the display target pictogram is shared. Temporarily stored as starting learning data.

If all of the display target pictograms are selected as processing target pictograms and the processes after S960 are executed, the relevance correction process is terminated and the process proceeds to S880.
In this way, the relevance correction in S870 is performed. When the order of arrangement of pictograms on the pictogram selection screen is determined in S880 using the relevance D after correction, the pictogram selection screen displayed when creating a reply mail or creating a blog comment displays the received mail or the blog article to comment on The pictograms co-occurring with the pictograms used in are arranged at a location close to the cursor position in the initial state. Therefore, according to the present embodiment, there is an advantage that it is easy to input pictograms when creating a reply mail or comment text.

  Return the story to S390. In S390 shown in FIG. 8, the cursor movement of the position selection screen is controlled in this way according to the user's usage environment, and further, the pictogram selection screen corresponding to the cursor position of the movement destination is displayed on the display unit 30. .

  When the pictogram selection cursor is moved while the pictogram selection screen is displayed (Yes in S400), the control unit 10 moves the pictogram selection cursor in the pictogram selection screen in accordance with this operation (S410). ). When the cursor position is moved by voice input, the cursor is moved to a pictograph corresponding to the spoken number.

If an operation for closing the pictogram selection screen is performed through the user interface (Yes in S420), processing for closing the pictogram selection screen being displayed is performed (S430).
In addition, when a confirmation operation is performed to place the pictogram at the pictogram selection cursor position on the pictogram selection screen in the processing target sentence (Yes in S440), the pictogram selection screen is closed (S450), and the pictogram selection cursor is displayed. The pictograph at the position is arranged at an additional candidate position in the processing target sentence on which the position selection cursor is placed on the position selection screen (S460). If the placement pattern of the pictogram at the additional candidate position corresponds to “replacement”, the independent word in the processing target sentence at the position where the position selection cursor is placed is deleted, and the position corresponding to the independent word The position selection screen is updated by arranging the selected pictograms. In this way, in this embodiment, the processing target sentence is edited so that pictograms are arranged in accordance with an instruction from the user.

  In addition, after the process of S460, using the pictograms arranged in the current processing target sentence and the feature vector generated in S820 for the additional candidate positions where the pictograms are arranged, the identification ID and the feature of this pictogram The pictographic case data associated with the vector is created (S470), and the created pictographic case data is registered in the pictographic case DB (S480). Thereafter, the co-occurrence learning process is executed for the pictograms arranged in the current process sentence (S490).

  When the co-occurrence learning process is started in S490, the control unit 10 first determines whether or not the co-occurrence learning data corresponding to the pictograms arranged in the processing target sentence is stored in the RAM 13 as shown in FIG. Is determined (S1010). Note that the co-occurrence learning data corresponding to the pictograms arranged in the processing target sentence this time is the pictograms arranged in the processing target sentence in the relevance correction process (see FIG. 13) as the processing target pictograms. The co-occurrence learning data created in S990 when a reference pictogram that co-occurs with the pictogram is extracted.

When the co-occurrence learning data exists, the reference pictogram corresponding to the co-occurrence learning data can be handled as a pictogram having a co-occurrence relationship with the pictogram arranged in the current processing target sentence.
Therefore, if the corresponding co-occurrence learning data exists (Yes in S1010), the process proceeds to the processing after S1020, and the reference pictograms indicated by the co-occurrence learning data and the pictograms arranged in the current processing target sentence The number of co-occurrence indicated by the co-occurrence data corresponding to the combination of is updated.

  Specifically, when the process proceeds to S1020, the control unit 10 selects one of the pictograms (reference pictograms) registered as the co-occurrence learning data corresponding to the pictograms arranged in the current processing target sentence as the co-occurrence learning target pictograms. A value obtained by adding one to the co-occurrence number q indicated in the co-occurrence data indicated by the co-occurrence data corresponding to the combination of the pictogram arranged in the current processing target sentence and the co-occurrence learning target pictogram in S1030 (Q ← q + 1).

  Then, it is determined whether or not the updated number of co-occurrence q has reached a value satisfying the update condition of the correction coefficient k (S1040). For example, when the condition for updating the correction coefficient k is set when the co-occurrence count q matches 10 or 20 or 50 or 100 or 200, the co-occurrence count q is 10 or 20 or 50 or 100 or If it matches 200, an affirmative decision is made in S1040. However, the update condition setting method presented here is merely an example.

  If it is determined that the co-occurrence number q has reached a value satisfying the update condition of the correction coefficient k (Yes in S1040), the process proceeds to S1050, and the value Δk to be added when the update condition is satisfied is The correction coefficient k stored in the co-occurrence DB is updated (k ← k + Δk). Of course, the object to be updated here is the correction coefficient k in the co-occurrence data in which the co-occurrence number q is updated in S1030. Thereafter, the process proceeds to S1060.

On the other hand, if the co-occurrence number q does not satisfy the update condition of the correction coefficient k (No in S1040), the process of S1050 is skipped and the process proceeds to S1060.
In S1060, all pictograms (reference pictograms) registered as co-occurrence learning data corresponding to the pictograms arranged in the current processing target sentence are selected as the co-occurrence learning target pictograms, and the processes after S1030 are performed. If it has not been executed (No in S1060), the process proceeds to S1020 and is not selected as a co-occurrence learning target pictogram out of the pictograms registered as co-occurrence learning data Are set as new co-occurrence learning target pictograms, and the processes after S1030 are executed.

  When the processing after S1030 is executed for all the pictograms (reference pictograms) registered as co-occurrence learning data corresponding to the pictograms arranged in the processing target sentence this time (Yes in S1060), they are stored in the RAM 13. All the co-occurrence learning data is deleted, and the co-occurrence learning process is terminated.

  When the co-occurrence learning process in S490 is completed in this way, the control unit 10 determines whether or not an operation for closing the position selection screen has been performed (S500), and an operation for closing the position selection screen has not been performed. In this case, the process proceeds to S370, the operation for moving the position selection cursor, the operation for moving the pictogram selection cursor, the operation for closing the pictogram selection screen, the operation for determining the pictogram to be arranged in the processing target sentence on the pictogram selection screen, and The process waits until one of the operations for closing the position selection screen is performed, and when it is performed, a process corresponding to the performed operation is executed.

  If the operation for closing the position selection screen is performed (Yes in S500), the position selection screen is closed and the processing target sentence display area on the display screen of the apps A and B is edited through the position selection screen. The processing target sentence is displayed, and the acceptance of the pictogram input operation of the processing target sentence is completed (S510). Finally, the pattern statistical data stored in the storage unit 70 is updated (S520), and the input conversion process is temporarily terminated. Then, the input conversion process is started again from S300.

  In S520, the values Y1, Y2, Y3, Y4, and Y5 held in the pattern statistical data are summed according to the arrangement pattern of pictograms in the processing target sentence for which the pictogram input operation has been completed, as in the processing in S220. Add a value that becomes a fixed value X.

  As described above, the electronic apparatus 1 according to the present embodiment has been described. According to the present embodiment, the sentence structure is analyzed, a position selection screen on which the arrangement position of the pictogram can be selected is displayed, and the pictogram can be easily selected by operating the cursor. Since the arrangement position can be selected, it is possible to improve the operability of the electronic device 1 in the process of selecting the location where the user adds a pictogram in the sentence. The configuration according to the electronic device 1 of the present embodiment is particularly effective when employed in a portable electronic device or an in-vehicle electronic device in which the operability of the operation unit 40 is poor.

  In addition, according to the present embodiment, the usage status of pictograms is learned by reflecting the pattern statistical data, pictogram example data, and co-occurrence DB, and the initial position of the position selection cursor on the position selection screen according to the usage status of the user In addition, since the order of movement and the arrangement order of pictograms on the pictogram selection screen are switched, it is very easy for the user to input pictograms.

  The sentence acquisition means described in “Claims” corresponds to the process of S300 in this embodiment, the position candidate determination means corresponds to the processes of S330 and S350, and the first display control means corresponds to the process of S360. Corresponding to the processing, the second display control means corresponds to the processing of S880, and the pictogram adding means corresponds to the processing of S460.

  In addition, the frequency recording means corresponds to the process of S520, and the cursor position control means corresponds to the position selection cursor control process. The keyword setting unit corresponds to the process of S810, the pictogram candidate determination unit corresponds to the process of S820 to S860, the target sentence attribute specifying unit corresponds to the process of S320, and the case data registration unit is S470. , S480.

  The received pictogram specifying means corresponds to the processes of S920, S940, S150, and S160, the correcting means corresponds to the processes of S950 to S980, and the received sentence attribute specifying means corresponds to the process of S120, and the subkeyword The setting means corresponds to the process of S170, the sub case data registration means corresponds to the processes of S180 to S200, and the relevance evaluation means corresponds to the processes of S820 and S830.

  Subsequently, a second embodiment will be described. The electronic device 1 according to the second embodiment is configured to select a pictogram to be displayed on the pictogram selection screen based on the degree of association D after correction using the correction coefficient k indicated by the co-occurrence DB. On the other hand, the contents of the pictogram selection screen display control process are different. Therefore, in the following description, the contents of the pictogram selection screen display control process executed by the electronic device 1 of the second embodiment are limited to those described with reference to FIG. 16, and the above-described implementation of the electronic device 1 of the second embodiment is included. A description of the same components as in the example is omitted. In the flowchart shown in FIG. 16, the same step numbers are assigned to the steps for performing the same processing as in the first embodiment.

  When the control unit 10 in the electronic device 1 of the second embodiment starts the pictogram selection screen display control process in S720, first, as in the above-described embodiment, a keyword is added to the additional candidate position to which the position selection cursor is aligned. Is set (S810), and the feature vector Va corresponding to the additional candidate position is generated based on the keyword group set here, the style and reader of the processing target sentence, and the arrangement pattern of the pictograms at the additional candidate position. (S820). Further, a cosine distance between this feature vector Va and the feature vector Vb indicated by each pictographic case data registered in the pictographic case DB is calculated, and the degree of association between the feature vector Va and the feature vector Vb is evaluated (S830). .

  In S1140 subsequent to S830, the pictograms indicated by the pictographic case data having a relevance level D equal to or higher than a predetermined first threshold TH1 are displayed on the pictogram selection screen based on the relevance level D calculated for each pictogram case data in S830. Is temporarily set to the display target pictogram. Since the setting operation of the display target pictogram in S1140 is a temporary setting operation, the first threshold value TH1 may be set lower than the threshold value TH in the above-described embodiment.

  In S1150 subsequent to S1140, the relevance level correction process shown in FIG. 13 is executed to correct the relevance level D calculated in S830 for each display target pictogram temporarily set in S1140. Specifically, the correction coefficient k is used as in the above-described embodiment, so that the relevance degree D of the display target pictogram related to the pictogram used in the received mail or the blog article is corrected.

  When the relevance correction process in S1150 ends, the control unit 10 proceeds to S1160, and the corrected relevance D among the currently temporarily set display target pictograms is greater than or equal to the second threshold TH2. Only the display target pictogram is set as a display target pictogram, and the display target pictogram with the corrected degree of association D less than the second threshold TH2 is excluded from the target of this setting. The second threshold value TH2 is set to a value larger than the first threshold value TH1. For the first threshold value TH1 and the second threshold value TH2, when the degree of association D is corrected by applying the correction coefficient k to the degree of association D equal to the first threshold value TH1, the degree of association D is The designer can arbitrarily determine it within a range that can exceed the second threshold TH2.

  In S1170 following S1160, it is determined whether or not there are J or more preset display target pictograms. If there are J or more (Yes in S1170), the process of S1180 is skipped. If the number is less than J (No in S1170), the process proceeds to S1190 after executing the process in S1180.

  In S1180, a pictographic character having at least one or more keywords set at the additional candidate position where the position selection cursor is set in the pictogram dictionary in the keyword group representing the representative word and the related word is extracted and extracted. Each pictogram is additionally set as a display target pictogram. Furthermore, for each pictogram additionally set as a display target pictogram, a value D0 smaller than the second threshold TH2 by a predetermined amount is assigned as the degree of association D.

  In S1190, a pictograph selection screen in which the display target pictograms set in this way are arranged in descending order of the degree of association D after correction is displayed around the additional candidate position where the position selection cursor is aligned ( S1190).

  Although the second embodiment has been described above, according to the second embodiment, the display target pictograms that are pictograms arranged on the pictogram selection screen are determined based on the corrected degree of association D. Pictograms corresponding to the usage status and usage status of pictograms in the blog article can be appropriately arranged as display target pictograms. Therefore, it is possible to suppress the probability of occurrence of a phenomenon that a pictograph necessary for the user is not displayed on the pictograph selection screen, and to provide the user with the highly convenient electronic device 1.

  The correcting means described in claim 9 in “Claims” corresponds to the processing of S1150 in this embodiment, and the pictogram candidate determining means corresponds to the processing of S1160 and S1180.

  Next, a third embodiment will be described with reference to FIGS. The electronic device 1 according to the third embodiment determines the arrangement order of pictograms to be displayed on the pictogram selection screen based on the co-occurrence count q indicated by the co-occurrence DB, and part of the pictogram selection screen display control process is performed. The degree is different from that of the first embodiment. Therefore, hereinafter, the pictogram selection screen display control process and the arrangement order determination process which is a subroutine of this process will be selectively described with reference to FIGS. 17 and 18. In the flowcharts shown in FIGS. 17 and 18, the same step numbers are assigned to the steps for performing the same processing as in the first embodiment.

  The pictogram selection screen display control process executed by the control unit 10 in the electronic device 1 of the third embodiment is described below with reference to S870 and S880 in the pictogram selection screen display control process of the first embodiment as shown in FIG. This is the same as the processing of S1270 and S1280.

  In S1270, the arrangement order determination process shown in FIG. 18 is executed to determine the arrangement order of pictograms to be displayed on the pictogram selection screen based on the co-occurrence number q indicated by the co-occurrence DB. In S1280, the arrangement order is determined in S1270. The pictograph selection screen is formed by arranging pictographs in the order of arrangement, and this is displayed on the display unit 30.

  Specifically, in the sequence order determination process, whether or not the processing target sentence is a sentence constituting a reply mail (S910), and whether or not the processing target sentence is a sentence constituting a comment sentence for a blog article ( S930) is determined, and if the processing target sentence is neither a sentence constituting the reply mail nor a sentence constituting the comment sentence for the blog article (No in S930), the display target pictogram set in S840 and S860 is determined. An arrangement order is assigned to a group in descending order of relevance D (S1410). More specifically, based on the degree of association D calculated in S830 and the degree of association D = D0 assigned in S860, the arrangement rank is assigned so that the higher the degree of association D is, the younger the arrangement rank is. Thereafter, the arrangement order determination process is terminated and the process of S1280 is executed, so that the pictograms are arranged in descending order of the degree of association D on the pictogram selection screen.

  On the other hand, when the processing target sentence is a sentence constituting the reply mail (Yes in S910), all the pictograms used in the received mail corresponding to the reply mail are extracted from the received mail (S920). If the processing target sentence is a sentence constituting a comment sentence for the blog article (Yes in S930), all pictograms used in the blog article corresponding to the comment sentence are extracted from the blog article (S940). .

  After the processing of S920 or S940, the control unit 10 proceeds to S950, and sets one of the pictograms set as the display target pictogram in S840 and S860 as the processing target pictogram. Then, from the pictogram group extracted in S920 or S940, all pictograms registered in the co-occurrence DB as pictograms co-occurring with the processing target pictogram are extracted as reference pictograms (S960).

  After the processing of S960, the process proceeds to S1370, and the co-occurrence number q indicated by the co-occurrence data corresponding to the combination of the reference pictogram and the processing target pictogram is obtained from the co-occurrence DB for each combination of the reference pictogram and the processing target pictogram. read out. Thereafter, the total value of the co-occurrence count q for each read combination is calculated as the array order evaluation value Q (S1380). Further, the reference pictogram extracted in S960 is temporarily stored in the RAM 13 as co-occurrence learning data for the processing target pictogram (S1390).

  When this process is completed, it is determined whether or not all the display target pictograms have been selected as the process target pictograms (S1395). If not selected (No in S1395), the process proceeds to S950 and the unselected display is displayed. The target pictogram is selected as a new processing target pictogram, and the processes after S960 are executed. By this operation, the arrangement order evaluation value Q is calculated for each display target pictogram, and the co-occurrence learning data is registered in the RAM 13.

  If all the display target pictograms are selected as the processing target pictograms (Yes in S1395), the process proceeds to S1400, and the array order is assigned to the group of display target pictograms in descending order of the array order evaluation value Q. That is, the array order is assigned such that the higher the array order evaluation value Q, the younger the array order. However, for the display target pictograms having the same arrangement order evaluation value Q, the arrangement order is assigned in descending order of the degree of association D. Thereafter, the arrangement order determination process is terminated, and the process of S1280 is executed. Thereby, pictograms are arranged on the pictogram selection screen in descending order of co-occurrence.

  As described above, the electronic apparatus 1 according to the third embodiment has been described. According to the present embodiment, since pictograms are arranged in descending order of the number of co-occurrence on the pictogram selection screen, an electronic mail reply is created. When creating a blog comment sentence, it is easy to select an emoji corresponding to the received mail or blog article.

The second display control means according to claim 18 in “Claims” corresponds to the processing of S1270 and S1280 in the present embodiment. The co-occurrence frequency update means corresponds to the operation of updating the co-occurrence count q in the co-occurrence learning process (see FIG. 15).
[Other Embodiments]
As described above, the first embodiment to the third embodiment have been described as embodiments of the present invention. However, the present invention is not limited to the above-described embodiments, and various aspects can be adopted.

  For example, when the cursor is operated for the first time on the position selection screen, it is an additional candidate position where pictogram placement is performed with the placement pattern having the highest appearance probability P1 to P5 based on the pattern statistical data, and is closest to the head of the sentence to be processed. A position selection cursor is displayed in the dotted frame at the additional candidate position (S640), and the cursor is not moved as shown in S650 to S710 in FIG. The electronic device 1 may be configured to perform a general cursor process for moving the cursor.

  In addition, in S840, a pictograph whose relevance level D is greater than or equal to the threshold value TH is set as a display target pictogram. You may do it.

  In the above embodiment, it is specified whether the processing target sentence is “mail” style or “blog” style. However, in the electronic device 1, the processing target sentence is “Do n’t / Mas” tone. It may be configured to specify the style of the processing target sentence depending on whether or not.

  Moreover, although the pictogram represented by image data was used in the said Example, a pictogram is not restricted to this. For example, the pictograms include single symbols such as musical note marks “♪” and star marks “★” included in typeface data, emoticons represented by multiple symbols such as parentheses, one or more symbols and images. A combination of pictograms represented by data can be used.

  In S830, the degree of association D is evaluated by the cosine distance between the feature vector Va and the feature vector Vb. However, the evaluation of the degree of association D is not limited to this. For example, the degree of association D can be evaluated using other machine learning methods such as a support vector machine (SVM) and a decision tree. Specifically, model learning such as a decision tree is performed using the case DB, and the evaluation value when the feature vector Va is input is set as the degree of association D in S830.

  In addition, in S180, a 6-digit user code is given to each user, but a code may be given to each user group.

DESCRIPTION OF SYMBOLS 1 ... Electronic device, 10 ... Control part, 11 ... CPU, 12 ... ROM, 13 ... RAM, 20 ... Wireless communication part, 30 ... Display part, 40 ... Operation part, 50 ... Voice output part, 60 ... Voice input part, 70: Memory unit

Claims (23)

Sentence acquisition means for acquiring an emoji addition target sentence that is a sentence to which an emoji is added;
Position candidate determination means for analyzing a sentence structure of the pictogram addition target sentence acquired by the sentence acquisition means and determining a group of additional candidate positions that are candidates for pictogram addition positions in the pictogram addition target sentence;
First display control means for causing a display device to display a position selection screen showing a group of the additional candidate positions in the pictogram addition target sentence;
When one additional candidate position is selected through a user interface from the group of the additional candidate positions displayed on the position selection screen, a pictographic group that is a candidate for a pictogram to be added to the selected additional candidate position is arranged. Second display control means for causing the display device to display a pictogram selection screen comprising:
When one of the pictograms arranged on the pictogram selection screen is selected through the user interface, the pictogram addition is performed by adding the selected pictogram to the selected addition candidate position in the pictogram addition target sentence. A pictogram adding means for editing the target sentence;
An electronic device comprising:   2. The electronic apparatus according to claim 1, wherein the position candidate determining unit determines a plurality of types of positions different in a predetermined sentence structure as the additional candidate positions. The electronic device is
Corresponding to the pictogram adding operation by the pictogram adding means, the frequency recording means for recording the frequency of adding pictograms to the additional candidate position for each type on a recording medium;
Means for moving the cursor to a position corresponding to any one of the additional candidate positions shown on the position selection screen according to a cursor operation through the user interface, and adding each type recorded on the recording medium A cursor position control means for initially placing a cursor at the additional candidate position of the type having the highest addition frequency based on the frequency;
With
The electronic apparatus according to claim 2, wherein the second display control unit causes the display device to display the pictogram selection screen for the additional candidate position selected on the position selection screen through the cursor. The cursor position control means moves the cursor in order from the additional candidate position of the type with the higher addition frequency to the additional candidate position of the type with the lower addition frequency in accordance with a cursor operation through the user interface. The electronic device according to claim 3. The electronic device is
For each of the additional candidate positions determined by the position candidate determining means, a keyword setting means for setting a word in the pictogram addition target sentence corresponding to the additional candidate position as a keyword;
For each additional candidate position, a pictogram group as an additional pictogram candidate that is a candidate for a pictogram to be added to the additional candidate position is determined from pre-registered pictogram groups based on the keyword set by the keyword setting unit. An emoji candidate determination means to
With
When the one additional candidate position is selected from the group of the additional candidate positions through a user interface, the second display control unit arranges the pictogram group determined as the additional pictogram candidates at the selected additional candidate position. The electronic device according to any one of claims 1 to 4, wherein the pictogram selection screen is displayed on the display device. A target sentence attribute specifying means for specifying an attribute of the pictogram addition target sentence;
The identification information of the pictogram added to the pictogram addition target sentence by the pictogram adding means, the information representing the keyword set by the keyword setting means for the additional candidate position corresponding to the position where the pictogram is added, and the Example data registration means for registering in the pictogram case database as pictogram example data related to the use case of pictograms, associating with information representing the attribute of the pictogram addition target sentence specified by the target sentence attribute specifying means,
With
The pictograph candidate determination means, for each additional candidate position,
The keyword set by the keyword setting means for the additional candidate position and the attribute combination specified by the target sentence attribute specifying means, and the keyword and the attribute indicated by each pictogram case data registered in the pictogram case database 6. The degree of relevance with the combination is evaluated, and on the basis of the evaluation value, the pictogram indicated by the pictogram example data having a relevance higher than a predetermined condition is preferentially determined as the additional pictogram candidate. Electronic equipment. The case data registration unit corresponds to the identification information of the pictogram added to the pictogram addition target sentence by the pictogram addition unit, the information indicating the type of the position where the pictogram is added, and the position where the pictogram is added. It associates with the information indicating the keyword set by the keyword setting means for the additional candidate position and the information indicating the attribute of the pictogram addition target sentence specified by the target sentence attribute specifying means, and the pictogram as the pictogram example data Register in the case database
The pictograph candidate determination means, for each additional candidate position,
A combination of the type of the additional candidate position, the keyword set by the keyword setting unit with respect to the additional candidate position, and the attribute specified by the target sentence attribute specifying unit, and each pictogram case registered in the pictogram case database Evaluate the type of position where the pictogram indicated by the data is added and the degree of association with the combination of the keyword and the attribute. Based on this evaluation value, priority is given to the pictogram indicated by the pictogram example data having a degree of association higher than a predetermined condition. The electronic apparatus according to claim 6, wherein the additional pictograph candidate is determined. The second display control means causes the display device to display the pictogram selection screen in which pictograms determined as the additional pictogram candidates are arranged in order of the relevance. 7. The electronic device according to 7. The sentence acquisition means acquires a response sentence to a received sentence from the outside as the pictogram addition target sentence,
The electronic device is
A received pictogram specifying means for specifying a pictogram used in a received sentence corresponding to the pictogram addition target sentence;
For each pictogram indicated by the pictogram example data, the pictogram candidate determination means evaluates the relevance from a co-occurrence database, which is a database that stores information representing the co-occurrence frequency between the pictograms for each combination of pictograms. And information indicating the co-occurrence frequency of each pictogram specified by the received pictogram specifying means, and information indicating the acquired co-occurrence frequency for each pictogram indicated by the pictogram example data for evaluating the degree of association Based on the correction means for correcting the evaluation value of the degree of association by the pictogram candidate determination means for the pictogram,
With
The pictograph candidate determining unit preferentially determines a pictograph indicated by the pictographic example data having a higher degree of association than a predetermined condition as the additional pictographic candidate based on the evaluation value corrected by the correcting unit. The electronic device as described in any one of Claims 6-8. The sentence acquisition means acquires a response sentence to a received sentence from the outside as the pictogram addition target sentence,
The electronic device is
A received pictogram specifying means for specifying a pictogram used in a received sentence corresponding to the pictogram addition target sentence;
For each pictogram determined as the candidate for the additional pictogram, the pictogram and the received pictogram specifying means are identified from the co-occurrence database, which is a database that stores information representing the co-occurrence frequency between the pictograms for each combination of pictograms. The information representing the co-occurrence frequency with each pictogram is acquired, and for each pictogram determined as the additional pictogram candidate, the pictogram candidate determination means for the pictogram based on the acquired information representing the co-occurrence frequency Correcting means for correcting the evaluation value of the relevance by
With
The second display control means displays, on the display device, the pictogram selection screen in which pictograms determined as the additional pictogram candidates are arranged in descending order of the relevance based on the evaluation value corrected by the correction means. The electronic device according to claim 8, wherein:   The electronic device according to any one of claims 6 to 10, wherein the target sentence attribute specifying unit specifies a reader of the pictogram addition target sentence as the attribute.   The electronic device according to any one of claims 6 to 11, wherein the target sentence attribute specifying unit specifies the style of the pictogram addition target sentence as the attribute. The sentence acquisition means acquires a response sentence to a received sentence from the outside as the pictogram addition target sentence,
The electronic device is
A received pictogram specifying means for specifying a pictogram used in the received sentence;
A received sentence attribute specifying means for specifying an attribute of the received sentence;
Sub-keyword setting means for setting, as a keyword, a word corresponding to the pictogram in the received sentence in which the pictogram is used for each pictogram identified by the received pictogram identifying means;
For each pictogram specified by the received pictogram specifying means, the identification information of the pictogram is used for the information indicating the attribute of the received text and the pictogram in which the pictogram specified by the received text attribute specifying means is used. Sub case data registration means for registering in the pictogram case database as the pictogram case data in association with the keyword information set by the sub keyword setting means;
The electronic apparatus according to claim 6, further comprising: The received pictogram specifying means specifies the pictogram used in the received text, and for each pictogram used in the received text, specifies the position type of the pictogram in the received text,
The reception case data registration means includes, for each pictogram specified by the received pictogram specifying means, identification information of the pictogram and information indicating the type of position of the pictogram in the received sentence specified by the received pictogram specifying means. And the pictogram example data associated with the information indicating the attribute of the received text in which the pictogram specified by the received text attribute specifying means is used and the keyword information set by the sub-keyword setting means for the pictogram. The electronic device according to claim 13, wherein the electronic device is registered in the pictograph case database as follows. The electronic device according to claim 13 or 14, wherein the received sentence attribute specifying unit specifies a writer of the received sentence as the attribute. The electronic device according to any one of claims 13 to 15, wherein the received sentence attribute specifying unit specifies a sentence style of the received sentence as the attribute. A target sentence attribute specifying means for specifying an attribute of the pictogram addition target sentence;
A keyword setting means for setting a word in the pictogram addition target sentence corresponding to the additional candidate position as a keyword for each additional candidate position;
The identification information of the pictogram added to the pictogram addition target sentence by the pictogram adding means, the information representing the keyword set by the keyword setting means for the additional candidate position corresponding to the position where the pictogram is added, and the Example data registration means for registering in the pictogram case database as pictogram example data related to the use case of pictograms, associating with information representing the attribute of the pictogram addition target sentence specified by the target sentence attribute specifying means,
For each additional candidate position, the combination of the keyword set by the keyword setting unit with respect to the additional candidate position and the attribute specified by the target sentence attribute specifying unit, and each pictogram case data registered in the pictogram case database Relevance evaluation means for evaluating relevance with the combination of the keyword and the attribute indicated by
With
The second display control unit preferentially assigns a young arrangement order to the pictograms indicated by the pictorial example data having the high relevance level based on the evaluation result of the relevance level evaluation unit for the selected additional candidate position. In this way, an arrangement order is assigned to each of the pictograms arranged on the pictogram selection screen corresponding to the selected additional candidate position, and a pictogram selection is made by arranging pictogram groups in order from the pictogram to which the young arrangement order is assigned The electronic device according to any one of claims 1 to 3, wherein a screen is displayed on the display device. The sentence acquisition means acquires a response sentence to a received sentence from the outside as the pictogram addition target sentence,
The electronic device includes a received pictogram identifying unit that identifies a pictogram used in a received sentence corresponding to the pictogram addition target sentence,
The second display control means, for each pictogram arranged on the pictogram selection screen, from the co-occurrence database that is a database that stores information indicating the co-occurrence frequency between the pictograms for each combination of pictograms Information indicating the co-occurrence frequency with each pictogram specified by the pictogram specifying means is acquired, and the co-occurrence frequency is high for the pictograms arranged on the pictogram selection screen based on the information indicating the co-occurrence frequency. A young arrangement order is assigned preferentially from a thing, and a pictogram selection screen in which the pictogram group is arranged in order from a pictogram to which the young arrangement order is assigned is displayed on the display device. The electronic device according to claim 3. Corresponding to the pictogram addition operation by the pictogram adding means, information indicating the co-occurrence frequency of the pictogram added to the pictogram addition target sentence and the pictogram used in the corresponding received sentence is the co-occurrence frequency. The electronic device according to claim 18, further comprising a co-occurrence frequency updating unit that updates the occurrence database. The electronic device according to any one of claims 1 to 19, wherein the position candidate determination unit determines each sentence end in the pictogram addition target sentence as the additional candidate position. The electronic device according to any one of claims 1 to 20, wherein the position candidate determination unit determines an independent word end of each phrase in the pictogram addition target sentence as the additional candidate position. The position candidate determination means determines a replaceable position that is a position where an independent word that can be replaced with a pictogram in the pictogram addition target sentence is present as the additional candidate position,
When adding the selected pictogram to the replaceable position, the pictogram adding means deletes the independent word existing at the replaceable position and adds the selected pictogram instead of the independent word The electronic device according to any one of claims 1 to 21, wherein the electronic device is characterized in that   The program for making a computer implement | achieve the function as each means with which the electronic device of Claim 1 is provided.