JP2012221480A - Message processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a message processing system capable of recognizing, as visual information, not only feelings of a poster but also total and quantitative feelings that a follower is to receive by posting by analyzing a message of a short sentence posting site and adding an illustration optimum for the feeling.SOLUTION: A message (posted sentence) of twitter(R) is morphologically analyzed, combinations of respective words and feelings and combinations of the words and illustrations are turned into a table, and scores are imparted. Thus, even for posting for which only the message is inputted, the message can be displayed with the optimum illustration imparted.

Description

  The present invention provides a message on a short text posting site such as “Twitter” (trademark of US Twitter Co., Ltd.) that discloses a posting in a sentence (message) in units of hundreds of characters from a user to other users associated with the user. It relates to processing technology.

  In the short text posting site as described above, the number of characters is limited to one hundred and ten characters, so the user can follow up on personal tweets in daily life (follow-up registration in hopes of browsing messages posted by the user) However, it is difficult to grasp the emotions of the speaker at the time of posting and the total emotions of followers other than himself even if the information is disclosed to other users.

  On the other hand, if you increase the number of users you follow, the number of posts per unit time for browsing (timeline: posts displayed in chronological order) will also increase rapidly, which may make you disgusted there were. For this reason, there is a problem that the user leaves the short text posting site. When a sentence (message) is input to such a short text posting site, information of other users who sympathize with the user in real time, information of a user who has input a sentence related to the sentence input by the user, Japanese Unexamined Patent Application Publication No. 2009-193133 (Patent Document 1) presents related advertisement information, but no effort has been made to analyze and visually display the posted message itself.

  JP 2009-193133 A

  The present invention has been made in view of the above problems, and by analyzing the message of a short text posting site and adding an optimal illustration for the emotion, not only the contributor's emotion but also the follower by posting The technical challenge is to realize a message processing system that can grasp the total and quantitative emotions that will be received as visual information.

The present invention employs the following means in order to solve the above problems.
Claim 1 of the present invention is a message processing system between communication terminals via a network, the first message consisting of characters according to an application program downloaded from a server connected to the network, and the message Receiving means for accepting input of an emotion code arbitrarily associated with the user and an illustration code associated with the emotion code, morphological analysis of the first message data received in the above, and part of speech; Emotion statistics table that registers and associates words, emotion codes, and scores added by selecting combinations thereof, and non-language statistics that registers and associates parts of speech, words, illustration codes, and scores A part-of-speech priority table in which a priority order for each part of speech is registered When receiving only message data from the communication terminal or another communication terminal and receiving it as second message data, the second message data is morphologically analyzed, and the part-of-speech priority table is The part of speech that is prioritized by reference is selected, the emotional code that is searched out from the emotion statistical table for the combination having the highest score among the combinations of the part of speech, the word, and the emotional code is determined, and the priority is given. Based on the part of speech, the combination of the part of speech and the word and the illustration code having the highest score is searched from the non-language statistical table to determine an illustration code to be output, and the second message data, A message with an illustration that is combined with an emotion code determined in step 1 and an illustration code. Message processing system including a message processing server for transmission and display on the other communication terminals.

  According to the present invention, by analyzing the message of the short posting site and adding an optimal illustration for the emotion, not only the contributor's emotion but also the total and quantitative that the follower will receive by posting It is possible to realize a message processing system capable of grasping various emotions as visual information.

  It is explanatory drawing which shows the system configuration | structure of this invention.   It is a hardware block diagram of a message processing server.   It is explanatory drawing which shows the content of the emotion statistics table produced | generated when 1st message data is input, and a non-language statistics table.   It is explanatory drawing which shows the relationship between the emotion statistics table used when a 2nd message data is input, a non-language statistics table, and a part of speech priority table.

The present invention will be described with reference to the drawings.
As shown in FIG. 1, the present invention is a system managed by a server (SEE) connected by a network, and the network is connected to an iPhone (Apple Inc. licensed through base stations (BS1, BS2)). A mobile communication terminal such as a smartphone or an i-modo terminal (trademark of NTT DoCoMo) such as a trademark or an Android terminal (a trademark of Google Inc. in the United States) or the like is connected in a communicable state.

  As shown in FIG. 2, the server (SEE) has a central processing unit (CPU) and a main memory (MM) via a bus (BUS), a large-scale storage device (HD), a display (DISP), a keyboard (KBD), etc. And an interface (NW I / O) to the network (NW).

  A server application program (SV APL) is installed in the large-scale storage device (HD) together with an operating system (OS). The server application program (SV APL) is stored in a bus (BUS) and a main memory (MM). The functions described in the present embodiment are realized by the central processing unit (CPU) reading and executing the processing via the.

  In addition to the above, the large-scale storage device (HD) includes a user table (UTBL) for registering user personal information, various tables such as a non-language statistics table, an emotion statistics table, a part-of-speech priority table described later, The database is registered.

  These databases may be operated by a network server provided separately from the server (SEE).

  In the emotion statistics table registered in the large-scale storage device (HD) of the server (SEE), as shown in FIG. 3, the part of speech, the word, the emotion, and the score are associated with each other. Here, the combination of words and emotions is not a single one, and a plurality of emotions may be associated with one word. For example, in the same figure, two emotions of “love” and “joy” are registered for “delicious” (verb), and a score is given for each combination. The function of this score will be described later.

  In the non-language statistics table, parts of speech, words, illustration codes, and scores are associated. Here, the combination of the word and the illustration code is not a single one, and a plurality of illustration codes may be associated with one word. For example, in the figure, two illustration codes “AI001” and “YR001” are registered for “cake” (noun), and a score is assigned to each combination.

  When inputting the first message (MES1), the first user inputs message data, an emotion code, and an illustration code from the mobile communication terminal (MP1) held by the first user. Here, it is assumed that “cake was delicious” was input as message data, “affection” was selected as the emotion code, and “AI001” was selected as the illustration code.

  As described above, the message data is text, and may be input using a flick input or a software keyboard prepared as an input interface of the mobile communication terminal (MP1), or hardware linked by a communication function such as Bluetooth. You may do it from the wear keyboard.

  The emotion code is designed to prompt the user by selecting one emotion that is most suitable for the message data input by the user, and the emotion code can be input by designating an emotion icon. Further, by selecting this icon, illustrations corresponding to the emotion are sequentially displayed so that an illustration corresponding to the emotion code can be selected.

  The combination data of message data, emotion code, and illustration code input on the mobile communication terminal (MP1) is transmitted to the server (SEE) via the base station (BS1) and the network (NW).

  In the server (SEE) that has received the combination data, the central processing unit (CPU) processes the combination data in accordance with a server program (SV APL).

  Here, morphological analysis is first performed on the message data “cake was delicious”, and it is broken down into nouns and verbs “cake” and “delicious”.

  Then, the presence / absence of the combination with the emotion code selected by the first user is searched for each part of speech subjected to morphological analysis, and the score is incremented (+1). Here, since the emotion code “loving” is entered for the message data “cake was delicious”, the score of the combination of “cake” + “affection” and “delicious” + “love” Each combination score is incremented (+1).

Next, the presence / absence of a combination with the illustration code selected by the first user is searched for each part of speech that has been subjected to the morphological analysis, and the score is incremented (+1). Here, since the illustration code “AI001” is selected for the message “cake was delicious”, the combination score of “cake” + “AI001” and the combination of “delicious” + “AI001” Each score is incremented (+1).
In this way, the emotion statistics table and the non-language statistics table are constructed.

  Next, how the various tables are referred to and what kind of output display is made when the second user inputs only message data will be described with reference to FIG.

  When message data (here, “the cake that seems to be cute and melted was delicious”) is input from the mobile communication terminal (MP1) held by the second user, this message data is sent to the base station (BS1) and the network (NW). To the server (SEE).

  In the server (SEE), the central processing unit (CPU) processes this message data in accordance with a server program (SVAPL).

  Here, the morphological analysis is first performed on the message data “the cake that seems to be cute and meltable”, and it is decomposed into the nouns “verb”, “melt”, “cake”, and “delicious”, and the verbs and adjectives.

  For each morpheme (word), the emotion statistics table is referenced to check the combination of emotion and score for the matching word. Next, referring to the part of speech priority table, the part of speech to be prioritized is selected. Here, since “adjective” is registered as the highest priority part of speech, the adjective is preferentially extracted from the word of the morphological analysis result described above. As a result, for the words “cute” and “melt”, the combination of “word” and “emotion” having the highest score is extracted. Here, the score of “cute” + “love” is 8, and the score of “melt” + “surprise” is 9, so the emotion code of “surprise” is selected.

  Here, only one emotion code is registered for each word, but even if multiple emotion codes are registered for one word, the word and emotion are registered in the same manner as above. Select an emotional code with a higher score from the combinations of codes.

  Next, the morphological analysis of the message data “The cake that seems to melt cute” was morphologically analyzed. For each word of “cute”, “melt”, “cake”, “delicious”, refer to the non-language statistics table and match Check the combination of illustration code and score for the word you want.

  Next, the part-of-speech priority table is referred to again, and the part-of-speech to be prioritized is selected. Here, as described above, “adjective” is registered as the highest priority part of speech, so the adjective is preferentially extracted from the word of the morphological analysis result. As a result, a combination of “word” and “illustration code” having the highest score is extracted for the words “cute” and “melt”. Here, since the score of “cute” + “KW002” is 16 and the score of “melt” + “BX005” is 17, an illustration code of “BX005” having a higher score is selected.

  As described above, the combination data (message with illustration) of the original message data, the selected emotion code “surprise”, and the illustration code “BX005” is generated, and the communication terminal (SEE) MP1) or other communication terminal (MP2).

  Thus, even when the second message (MES2) consisting only of sentences is posted, the emotion and the most matching illustration at that time are displayed by the emotion statistics table, part-of-speech priority table and non-language statistics table of the server (SEE). You can select and display.

  The present invention can be used for a short text posting site and a short text distribution system via a network called Twitter (trademark of Twitter company), mixi voice (trademark of mixi company), Facebook (trademark of Facebook company, USA), messenger, etc. .

SEE server NW network BS1, BS2 base station MP1, MP2 portable communication terminal CPU central processing unit MM main memory BUS bus DISP display device KBD keyboard HD large-scale storage device OS operating system SV APL server application program MES1 first message MES2 second Message YR001, KN002, AI001, BX005 Illustration code

Claims (1)

A message processing system between communication terminals via a network,
In accordance with an application program downloaded from a server connected to the network, a first message composed of characters, an emotion code arbitrarily associated with the message by the user, and an illustration code associated with the emotion code Accepting means for accepting input,
A morphological analysis of the first message data received in the above, and an emotion statistics table for registering in association with a part of speech, a word, an emotion code, and a score added by selecting a combination thereof,
A non-linguistic statistics table that registers part of speech, words, illustration codes, and scores in association with each other;
It consists of a part-of-speech priority table in which the priority order for each part of speech is registered,
When receiving only message data from the communication terminal or another communication terminal and receiving it as second message data,
The second message data is subjected to morphological analysis, the part-of-speech priority table is selected with reference to the part-of-speech priority table, and the combination having the highest score among the combination of part-of-speech, word, and emotion code is selected as the emotion. Determine the emotion code to be retrieved from the statistical table and output,
Based on the priority part of speech, determine the illustration code to be output by searching the part of speech and the combination having the highest score among the combinations of the word and the illustration code from the non-language statistical table;
A message processing system comprising: the second message data, the emotion code determined as described above, and a message processing server that transmits and displays a message with an illustration combined with an illustration code to the communication terminal or another communication terminal.

Images