JP2006113956A - Program for creating schedule information, information processor, correction method of schedule information - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To surely carry out requested processing on a mail having contents requiring change/deletion of a schedule for improving schedule information automatic creation function. <P>SOLUTION: An analysis processing part 31 performs morphological analysis on a mail in response to storage of a new mail in mail information 15 and extracts a schedule expression word. An abstract expression analysis part 33, which is actuated when a schedule expression word necessary for creation of a schedule is not defined, transforms the undefined schedule expression word into an absolute expression. In this transformation processing, a reference source mail of a mail to be processed is extracted from a mail information file by using the analysis processing part 31 and an inter-mail relationship analysis part 34, and an expression matching the undefined schedule expression word is extracted from the extracted mail. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technology for automatically creating schedule information by using document information of an electronic mail transmitted or received by an information processing apparatus having a computer as a control entity.

  Information processing apparatuses such as cellular phones and PDAs in recent years include an application (mailer) for sending and receiving e-mails and an application (scheduler) for schedule management. In addition, in this type of information processing apparatus, an application has already been proposed in which schedule information is created by extracting information about a schedule from document information of mail sent and received by a mailer, and the created schedule information is input to the scheduler. (See Patent Document 1).

Japanese Patent Laid-Open No. 11-289568

  In the above-mentioned patent document 1, the document information of the e-mail is analyzed by morphological analysis or the like to extract information related to the schedule, and schedule information is created. Further, by setting a link between the created schedule information and the original mail, the schedule information and the corresponding document information of the mail can be displayed on the same screen.

  Patent Document 1 describes only creating schedule information from document information of a single mail, but in the real world, there are frequently cases where the schedule is changed for the convenience of the parties.

  Further, in Patent Document 1, an expression (specifically referred to as “absolute expression” in this specification) that specifically expresses the date, time, etc. is extracted and the schedule information is assembled. In many cases, an abstract word such as a pronoun is used without using an expression, or a relative expression such as “˜before” and “˜after” is used. In this specification, the expression form of the schedule in which the contents cannot be grasped specifically is called “abstract expression”.

  FIG. 10 shows an example in which the contents of the schedule are changed by the exchange of mail. Hereinafter, mails 1, 2, 3, 4, and 5 are referred to in order from the top. In each of the emails 1 to 5, words corresponding to the absolute expression of the schedule are surrounded by a frame, and the dots corresponding to the abstract expression of the schedule are given.

  The first mail 1 addressed from Mr. A to Mr. B includes absolute information “9/15”, “15:00”, and “meeting” indicating the date and purpose. The mail 2 returned from Mr. B to the mail 1 indicates that the change of time is desired by an abstract expression “2 hours ahead”. Further, the email 3 sent back from Mr. A includes an abstract expression “time change, OK” and an absolute expression “103 conference room”. At the time when the mail 3 is transmitted, three main information (date, place, purpose) of the schedule is determined. Further, in this example, in the next email 4 from Mr. B, the meeting date is changed to “16th”, and the exchange of emails for confirming the schedule information is further continued.

  As described above, since the schedule information is variously changed by the exchange of mail, it is desirable that an application for creating a schedule can correct the already set schedule information in accordance with the change. However, since the mail document information often uses expressions close to everyday conversation, there is a possibility that the information necessary for specifying the schedule information to be corrected is not sufficiently included.

  For example, the mail 4 in the example of FIG. 10 only shows that “15th” is changed to “16th”, and information such as time, place, and purpose cannot be directly specified. Of course, if the schedule set on the 15th is only a meeting, the content can be specified with reference to the schedule information on the scheduler side. However, when a plurality of schedules are set on the 15th, it is difficult to specify the change target. In addition, although the mail 2 indicates that the meeting time is to be changed two hours ago, if a meeting is set other than 15 days, the meeting to be changed cannot be identified. Become.

  Thus, just processing a single e-mail as in Patent Document 1 may not be able to acquire information necessary for correcting schedule information, and the schedule information automatic creation function cannot be sufficiently enhanced. The problem arises.

  The present invention has been made paying attention to the above-mentioned problems, and when a mail requesting to change or delete a schedule that has already been set is transmitted or received, the processing required by the mail is executed with high accuracy. Therefore, it is intended to enhance the automatic creation function of schedule information.

  A program for creating schedule information according to the present invention uses mail information managed by the mail management unit in an information processing apparatus having a mail management unit for managing transmission / reception processing of electronic mail and a schedule management unit for managing schedule information. This is for setting a function for creating schedule information to be managed by the schedule management unit. The mail manager can be set by the mailer program, and the schedule manager can be set by the scheduler program.

  The first schedule information creation program according to the present invention includes a program for giving the computer the functions of each of the following analysis means, correction target specifying means, information correction means, and correction information output means. It is out.

  The analysis means analyzes mail that is under the management of the mail management unit and extracts expressions related to the schedule. For example, after the mail document information is segmented by word by morphological analysis, a word representing the date, place, purpose, and the like can be extracted by using a specific expression extraction or the like.

When the expression indicating the schedule correction is extracted by the analyzing unit, the correction target specifying unit searches for mails transmitted / received before the processing target mail in order, and mails related to the processing target mails (Hereinafter referred to as “quoting source mail”) is extracted, and the extracted mail is analyzed by the analyzing means to identify the schedule information to be corrected.
Note that “schedule correction” here means changing or deleting the contents of a schedule that has already been set.

  The citation mail can be extracted with reference to citation-related information (such as an identifier of the citation mail) included in the header part of each mail. Further, the similarity between mails can be obtained for the subject information and text information of the mail, and a mail with a high similarity can be extracted as a citation mail. The revision target identification means extracts the words related to the schedule to be revised from the quoted email extracted in this way, or extracts the schedule information linked to the quoted email, etc. Can be specified.

  For example, when the mail 2 in the example of FIG. 10 is analyzed by the analysis means, the word “2 hours ahead” can be extracted as an expression indicating the correction of the schedule. In response to this extraction, the correction target specifying means can start the citation source mail search process and extract the mail 1. Furthermore, by analyzing this mail 1, the words “September 15”, “15:00”, and “meeting” can be extracted, and the schedule information to be corrected can be specified from these words. it can.

  The information correcting unit corrects the schedule information specified by the correction target specifying unit using an expression indicating the change of the schedule. For example, in the above example, “2 hours ahead” is replaced with the correction amount “−2 hours”, and then “15:00” extracted from the mail 1 is changed to “13:00” based on the correction amount. It can be corrected.

  If the extracted citation source mail does not include information that can specify the schedule information to be corrected, the correction target specifying means can search for a higher level citation mail. For example, when the mail 4 in FIG. 10 is a processing target, “changed on the 16th” can be extracted as an expression indicating the correction of the schedule, but the mail 3 that is the direct citation mail of the mail 4 is extracted. Does not contain the information necessary to specify the correction target. Even in such a case, if the mails 3, 2, and 1 are extracted in order and it is determined whether or not each piece of mail contains useful information, “9/15” “meeting” “15:00” of the mail 1 The schedule information to be corrected can be specified.

  The correction information output means outputs the schedule information corrected by the information correction means to the schedule management unit. Receiving this, the schedule management unit can correct the corresponding schedule information.

  Note that when the analysis means extracts words such as “change”, “before”, “after”, and “cancel” from the email to be processed, it determines that this email contains information indicating a schedule change. The schedule item to be changed and the contents of the change can be determined from the dependency state of the words before and after that.

  According to the program for creating the first schedule information, the content of the email requesting the change (including the deletion) of the already set schedule information is insufficient, and the schedule information to be changed cannot be specified only by the email. Even in this case, the schedule information to be changed can be specified with high accuracy by extracting the previous mail as the citation mail. Therefore, the user does not need to manually correct the schedule information, and convenience can be improved.

  Next, a second program for creating schedule information according to the present invention is a program for giving the function of each means of the following analysis means, schedule information creation means, conversion means, and correction information output means to a computer. Including.

  The analysis means analyzes the mail entered under the management of the mail management unit and extracts expressions corresponding to a predetermined number of schedule items. The schedule information creation means assembles schedule information in which the expression extracted by the analysis means is associated with each item. As schedule items, items managed by a general scheduler such as date, place, purpose, and promised partner can be set. The schedule information creating means can create schedule information having a standardized configuration by a method of storing expressions corresponding to the items in a template in which these items are set. For information such as date and time information that can be unified in data format, it is desirable to convert the extracted expression into a reference format and store it in the template. Further, an item for which no corresponding expression is found may be left blank.

  When the information of the predetermined schedule item is unconfirmed, the converting means sequentially traces mails transmitted / received before the mail to be processed, and extracts mail related to the mail to be processed. And a second search process for extracting an expression corresponding to the unconfirmed item from the extracted mail until a mail containing an absolute expression corresponding to the unconfirmed item is extracted. The uncertain information is converted into an absolute expression using the expression extracted in step (1).

The processing by the conversion means is preferably performed on items for which information useful for specifying the schedule, such as date and time information, has not yet been determined.
In the first search process, the citation source mail can be extracted by the same method as that executed by the change target specifying means in the first program.

  In the second search process, the citation source mail can be analyzed by the analysis means, and information corresponding to the unconfirmed item can be extracted from the analysis result. When the absolute expression is finally extracted, the absolute expression corresponding to the undetermined information can be derived by correcting the absolute expression using the expression extracted during the search. However, it is desirable to select and use expressions extracted during the search. For example, when the date is uncertain, even if abstract expressions such as “that day” and “that day” are extracted, it is not necessary to use these abstract expressions for the conversion process. If the same relative expression is quoted in the reply mail for this mail after an email to change the first set time by relative expression such as “1 hour later” It is desirable to create a correction amount from only the expression extracted from one of the emails and perform the conversion process.

  When the absolute expression corresponding to the uncertain information is finally obtained, the correction information output means outputs the schedule information including the absolute expression to the schedule management unit as the correction information for the existing schedule information.

  According to the second schedule information creation program, even if the schedule item necessary for specifying the schedule is not expressed in absolute expression in the mail requesting correction of the schedule, by tracing the citation mail Since the corresponding absolute expression can be obtained, the schedule correction information can be created correctly.

Next, in a preferred embodiment according to the second program, the conversion means extracts the citation source mail based on the citation information set in the mail management unit and the similarity of the mail document in the first search process. To do.
The cited information can be identification information written in the header part. Alternatively, when a link is set for a related transmission / reception mail in the mail management unit, this link information can be used as citation related information.

  The similarity of document information is expressed by the appearance frequency of words included in each document information, the arrangement state of words, the final edit distance of words and sentences (the number of key operations for rewriting one word to another, etc.) And the like. In addition, when processing a mail from a mailer that does not leave citation information in the header part, as described in Patent Document 2 below, a citation part in the mail body is extracted and the mail including the citation part is included. It can be determined whether or not.

Japanese Patent Laid-Open No. 10-222438

  In said aspect, the mail transmitted using the mailer's reply function can be extracted by using quotation related information. Furthermore, by using the similarity of the document information, it is possible to extract the citation mail from newly created mail without using the reply function or from the mail sent from the mailer not including the citation information in the header part. it can.

  Furthermore, in another preferable aspect of the second program, when the conversion unit extracts a relative expression related to uncertain information in the second search process, the conversion unit converts the finally extracted absolute expression from the relative expression. The correction means which corrects with the corrected amount which did is included. According to this aspect, even when the information based on the absolute expression initially set by the relative expression is changed while the mail is exchanged multiple times, the absolute expression is corrected by the correction amount based on the relative expression. , Get the right information.

  Further, when the conversion means extracts the mail linked to the predetermined schedule information as the citation mail in the first search process, the search is terminated at that time, and the schedule information linked to this mail is used. Can be set to perform the conversion process. The function of setting a link can be given to the scheduler as disclosed in Patent Document 1 described above. Alternatively, a link may be set when the system according to the program according to the present invention outputs normal schedule information other than the correction information to the scheduler.

  According to the above aspect, when mail linked to the schedule information is extracted in the process of searching for the citation source mail, conversion processing of unconfirmed information is performed using the schedule information. Can be performed quickly and accurately.

  In the first and second programs described above, when the analysis means extracts information in absolute expression for a predetermined number of schedule items, means for assembling new schedule information using these information, A program corresponding to a means for outputting the schedule information to the schedule management unit can be set.

  An information processing apparatus according to the present invention includes: a mail management unit that manages email transmission / reception processing; a schedule management unit that manages schedule information; and management of the schedule management unit using mail information managed by the mail management unit And a schedule creation unit for creating target schedule information. The mail manager can be set by a mailer program, and the schedule manager can be set by a scheduler program. The schedule creation unit can be set by the first or second schedule information creation program.

  The information processing apparatus described above is preferably an apparatus having a computer as a control entity and having a communication function, such as a personal computer, a mobile phone, and a PDA.

  Furthermore, in the schedule information creation method according to the present invention, the schedule creation unit of the information processing apparatus analyzes a mail that is under the management of the mail management unit and extracts an expression related to the schedule. And when the expression indicating the correction of the schedule is extracted in the first step, mails transmitted / received before the mail to be processed are sequentially searched to extract mails related to the mail to be processed. A second step of analyzing the extracted mail by the analyzing means to specify the schedule information to be corrected, and the schedule information specified in the second step is corrected using an expression indicating the correction of the schedule A third step and a fourth step of outputting the schedule information corrected in the third step to the schedule management unit To perform the steps.

  The above method can be executed by an analysis unit, a correction target specifying unit, an information correction unit, and a correction information output unit set in the first schedule creation program. However, apart from the program for creating a schedule, it is also possible to assemble a program for configuring these four means.

  Furthermore, in another schedule information correction method according to the present invention, a first step of analyzing an email that has been managed by the email management unit and extracting an expression corresponding to a predetermined number of schedule items; The second step of assembling the schedule information in which the expression extracted in one step is associated with each item, and when the information of the predetermined schedule item is unconfirmed, it is transmitted / received before the mail to be processed in the first step A first search process for extracting mail related to the mail to be processed, and a second search process for extracting an expression corresponding to the unconfirmed item from the extracted mail. Execute until mail containing the absolute expression corresponding to the confirmed item is extracted, and convert the unconfirmed information into an absolute expression using the expression extracted in the second search process A third step, the schedule information including the converted absolute represented in the third step, a fourth step of outputting the schedule management unit as the correction information for the existing schedule information, executing.

  The above method can be executed by each of analysis means, schedule information creation means, conversion means, and correction information output means set by the second schedule information creation program. However, apart from the program for creating a schedule, it is also possible to assemble a program for configuring these four means.

  According to the present invention, in an information processing apparatus incorporating a function for automatically creating schedule information based on information described in an e-mail, when an e-mail having contents for requesting correction of the schedule is transmitted or received, The corresponding schedule information can be corrected with high accuracy in accordance with the contents of. Therefore, the automatic creation function of schedule information can be enhanced, and the convenience of the information processing apparatus can be improved.

FIG. 1 shows functions of the information processing apparatus according to the present invention.
This information processing apparatus has a computer as a control entity, and includes a communication circuit for connecting to the Internet, a character input operation unit, a display panel (none of which are shown), and the like. A specific form is assumed to be a mobile terminal such as a mobile phone or a PDA, but is not limited to this, and can be configured as a personal computer.

  In this information processing apparatus, a mail management unit 1, a schedule management unit 2, and a schedule information creation unit 3 are incorporated. The mail management unit 1 is a function given to the computer by a so-called mailer, and the schedule management unit 2 is a function given to the computer by a so-called scheduler 2.

  The mail management unit 1 includes a communication unit 11, a mail editing unit 12, a mail information display unit 13, a mail information management unit 14, a mail information file 15, and the like. The communication unit 11 transmits and receives electronic mail using the communication circuit. The mail editing unit 12 assembles the document information of the mail to be transmitted according to the user's character input operation. The mail information display unit 13 outputs the contents of the mail transmitted / received by the communication unit 11 to the display panel.

  The mail information management unit 14 takes in electronic data of transmitted mail and received mail from the communication unit 11 and stores them in the mail information file 15. In addition, the mail information designated from the mail information file 15 can be called according to the user's mail calling operation and output to the mail editing unit 12 or the mail information display unit 13.

  The schedule management unit 2 includes a schedule display unit 21, a schedule registration unit 22, a link information management unit 23, a schedule information file 24, and the like. The schedule registration unit 22 registers new schedule information in the schedule information file 24. The schedule display unit 21 reads predetermined schedule information from the schedule information file 24 in accordance with a user's schedule read operation and outputs the read schedule information to the display panel. When the schedule registration unit 22 registers new information, the link information management unit 23 sets link information for associating both the registration information and the information source electronic mail.

  The schedule information creation unit 3 creates schedule information using the mail information transmitted and received by the mail management unit 1. The schedule information created here is output to the schedule registration unit 22 of the schedule management unit 2. In addition to newly registering the information in the schedule information file 24 according to the given schedule information, the schedule registration unit 22 can execute processing for changing existing schedule information and processing for deleting existing schedule information.

The schedule information creation unit 3 includes an analysis processing unit 31, an information management unit 32, an abstract expression analysis unit 33, an inter-mail relationship analysis unit 34, and the like. The analysis processing unit 31 uses the mail newly stored in the mail information file 15 as a processing target, and performs morphological analysis on information (hereinafter referred to as “text information”) and subject information constituting the text of the mail. Furthermore, processing for extracting a word related to the schedule (hereinafter referred to as “schedule expression word”) from the words decomposed by the morphological analysis is executed.
The inter-mail relationship analysis unit 34 performs a process of extracting a citation source mail of the mail to be processed for the analysis process of the abstract expression analysis unit 33. Details of this processing will be described later.

  The information management unit 32 uses the schedule expression word extracted by the analysis processing unit 31 to execute control related to creation and output of schedule information. In addition, when a schedule expression word based on an abstract expression is extracted or a schedule expression word necessary for creating schedule information is not extracted, the information management unit 32 activates the abstract expression analysis unit 33. The abstract expression analysis unit 33 converts the schedule expression word based on the abstract expression or the schedule expression word that is not directly indicated into the absolute expression schedule expression word. This conversion process involves the analysis processing unit 31 in addition to the inter-mail relationship analysis unit 34.

  FIG. 2 shows a data configuration example of schedule information created by the schedule information creation unit 3. This schedule information has a configuration in which a schedule expression word extracted from an e-mail is applied to a template in a predefined format (hereinafter referred to as “schedule template” or simply “template”). In the template of this example, an entry field is set for each item managed by a general scheduler such as date, time, person name (name of the promised partner), location, purpose, etc., and an entry field for attribute data is provided. It has been. This attribute data indicates the type of information entered in the template, and “new”, “change”, “delete” and the like can be selected. Of the information of each item, date information and time information are converted into a prescribed format before being stored in the template.

  FIG. 3 shows a flow of processing for creating the schedule information. In principle, this process is executed when a new mail is stored in the mail information file 15 and the new mail is the first processing target. However, in response to an operation for designating mail other than new mail, this procedure can be executed with the designated mail as a processing target.

  First, in the first ST1 (ST is an abbreviation of “step”. The same applies to the following), morphological analysis is performed on the text information and subject information of the noticed mail. In the next ST2, schedule expression words are extracted from the words separated by morphological analysis, and these schedule expression words are stored in the schedule template.

  Next, in ST3-5, it is checked whether the schedule expression word concerning each information of date / time, a person name, and a place is a word by an absolute expression. The date / time information is a combination of the date information and time information of the schedule template, and ST3 is “YES” only when both pieces of information are expressed in absolute expression.

  If at least one of the date information and the time information is represented by an abstract expression or there is no corresponding word, the date / time information conversion process of ST6 is executed. Similarly, with respect to the personal name information, when a word based on the abstract expression is extracted or when the corresponding word is not extracted, the personal name information conversion process of ST7 is executed. Similarly, with respect to the location information, if a word based on an abstract expression is extracted or a corresponding word is not extracted, the location information conversion process of ST8 is executed.

  Thereafter, in ST9, it is determined whether or not the noticed mail is a mail requesting to change or delete an existing schedule. This determination processing is performed based on whether or not a word indicating change or deletion of the schedule (“change”, “correction”, “cancel”, etc.) is extracted in the morphological analysis in ST1, and the conversion processing in any of ST6, 7, 8 This is done based on whether or not is executed. Note that when performing discrimination using the results of morphological analysis, instead of simply checking whether there is a word such as “change”, the dependency relationship between these words and the schedule expression word is also checked. .

If it is determined that the noticed mail is not a mail requesting a change or deletion of the schedule, ST9 becomes “NO” and the process proceeds to ST10 to set the template attribute to “new”. On the other hand, if it is determined that the noticed mail is a mail requesting a change or deletion of the schedule, the process proceeds from ST9 to ST11, and the attribute of the template is set to “change” or “delete”.
When the attributes of the schedule template are set in this way, the process proceeds to ST12 and the set schedule template is output to the schedule management unit 2.

  If the attribute of the received schedule template is “new”, the schedule management unit 2 registers information of each item based on this template in the schedule information file 24 as new information. On the other hand, when the attribute of the schedule template is “change” or “deletion”, the schedule information file 24 is searched based on the information of each item of the template, and the corresponding information is extracted. When the correspondence information is extracted, a process of rewriting the correspondence information with a template based on the attribute or a process of deleting the extracted information is executed.

  In the conversion process of ST6, 7, 8 described above, the abstract expression analysis unit 33, the analysis processing unit 31, and the inter-mail relationship analysis unit 34 extract the citation source of the attention mail from the mails stored in the mail information file 15. Is extracted, and the abstract expression or missing expression of the noticed mail is replaced with the absolute expression by using the schedule expression word included in the citation source mail.

In the personal name information conversion process in ST7 and the location information conversion process in ST8, the mail information file 15 is searched retrospectively until a citation source mail in which the corresponding information is expressed in absolute expression is found. When a schedule expression word based on an absolute expression is found, the conversion process is terminated by storing this schedule expression word in the schedule template.
If the absolute expression cannot be extracted in the personal name information conversion process, the personal name can be specified by searching the address book in the mail management unit 1 using the mail address in the header part. On the other hand, when the absolute expression is not found in the place information conversion process, the template may be blanked.

  The ST6 date / time information conversion process is similar in that it searches the mail information file retroactively until a mail containing an absolute expression is found. However, in the abstract representation of the date / time information, an expression showing a relative value based on the set date / time, such as “2 hours ahead” (hereinafter referred to as “relative expression”). ) May be included. In order to deal with such a case, in this embodiment, the extracted relative expression is converted into a correction amount in a preset format, and the value indicated by the finally extracted absolute expression is corrected by the correction amount. Execute the process. Details will be described later.

  Next, the citation source mail extraction process that is commonly performed in the conversion processes of ST6, 7, and 8 will be described. In this embodiment, the citation source mail is extracted by using both the method of referring to the citation information included in the header information of each mail and the method of calculating the similarity of the document contents between the mails.

  A general mailer executes a process of writing the identification information of the citation mail in the header information in response to a user's reply operation or transfer operation. If the mail quoted by this identification information is positioned at the top and the quoted mail is positioned at the bottom, the citation relationship of each mail can be expressed in a tree structure as shown in FIG. When referring to the header information, the tree structure is sequentially traced from the lower level to the higher level.

  On the other hand, when using the similarity of the document, the morphological analysis is performed on the text information and the subject information, and the similarity is calculated by applying the similarity calculation method proposed in the natural language processing to the extraction result. it can. For example, the number of proper nouns common between documents, the similarity of the distribution state of meaningful words, and the like are calculated (see Non-Patent Document 1).

Satoshi Kinmei et al., "Frontier of Statistical Science (Volume 10) Language and Psychological Statistics-Analysis by Probabilistic Model of Language and Behavior" Iwanami Shoten, March 12, 2003, p91

  For mail sent without using the mailer's quoting function even though it is a response mail for a certain mail, or mail sent from a mailer that does not leave citation information in the header information, the quoting mail from the header information It becomes difficult to specify. However, in this embodiment, in addition to the method of referring to the citation information, a process of extracting the citation source mail based on the similarity of the documents between the mails is executed. Can be revealed. Therefore, as shown in FIG. 5, even when the mail information file 15 includes a mail group whose citation source is clear and a mail group whose citation source is unknown, each mail is assigned to a group (G1 in the figure). , G2, and G3)), and the upper-lower relationship can be clarified for each group.

  In this embodiment, when extracting the citation source mail, two types of pointers are set in the mail information file 15. One is a “main pointer” representing the mail to be processed. The other is a “sub-pointer” that represents the mail to be compared with the mail with the main pointer. FIG. 4 shows an example of setting the main pointer and the sub pointer. The sub pointer is set to the mail stored immediately before the mail with the main pointer, and then moves one by one in the past direction. This sub-pointer may be moved to the oldest mail in the mail information file 15. However, the present invention is not limited to this, and a range for a predetermined number of days may be set as a search range and moved within that range.

  FIG. 6 shows a basic processing procedure for extracting the citation source mail. In this example, it is assumed that the process starts from ST101, and the sub-pointer is moved within a predetermined search range.

  First, in ST101, the header information of the mail with the main pointer is analyzed to determine whether there is citation information. If the citation information is found here, ST102 becomes “YES” and the process proceeds to ST112, and the mail information file is searched based on the citation information to extract the citation mail. This is the end of the process.

  On the other hand, if the header information does not include citation information, ST102 is “NO” and the process proceeds to ST103, where a sub-pointer is set in the mail stored immediately before the main pointer.

  In the next ST104, using the function of the analysis processing unit 31, words constituting the subject information and text information of the mail with the main pointer attached are extracted. In subsequent ST105, the morphological analysis is similarly performed on the mail with the sub-pointer, and the words constituting the subject information and the text information are extracted.

  Next, in ST106, the citation relationship in the text between emails (the relationship in which one email contains most of the text of the other email and the former email contains text not found in the latter), the text Based on the date and time information included therein, the prior relationship between the creation times of the two mails is determined. Here, if it is confirmed that the information of the mail with the main pointer is created after the information of the mail with the sub-pointer, ST107 becomes “YES” and the process proceeds to ST108, where it is extracted at ST104 and 105. The degree of similarity between mails is calculated using the obtained information.

  If the mail with the main pointer quotes mail created before the mail with the sub-pointer, it may be determined that this mail was created before the mail with the sub-pointer. . In this case, ST107 becomes “NO” and ST108 is skipped.

Thereafter, in ST109 and 110, by executing ST105 to 108 while moving the subpointer forward one by one, the contents of the mail with the subpointer are analyzed at each pointer position, and the mail with the main pointer is sent. Calculate similarity. The calculated similarity is temporarily stored in the memory in a state in which the setting position of the pointer at that time is associated.
When the sub-pointer reaches the head of the search range, the process proceeds to ST111, and the mail at the position corresponding to the maximum value among the temporarily stored similarities is specified as the citation source mail.

Next, details of the date information conversion process in ST6 will be described.
As described above, the date information is a combination of the date information and time information of the schedule template. As the schedule information, it is desirable that both of these pieces of information are expressed in absolute expression. However, there are many mails that actually show only one of the date information and the time information. Therefore, in this conversion process, the date information and the time information are individually processed, and the citation source mail is extracted until both are converted into absolute expressions.

The detailed procedure of the date / time information conversion process will be described below with reference to FIGS. This process starts from ST201.
First, in ST201, the relative expression written in the date and time fields of the schedule template is changed to a correction amount. For example, “next day” can be converted to “+1 day”, and “2 hours ago” can be converted to “−2 hours”.
If only the expression that cannot derive the correction amount, such as “that day” and “that time” is written in the template, ST201 is skipped.

  In ST202, a main pointer is set to the noticed mail (the mail from which the template is extracted), and in the next ST203, the above-described citation mail extraction process is executed. If the citation mail is extracted here, ST204 becomes “YES” and the process proceeds to ST205 to check whether there is schedule information linked to the citation mail.

  Here, when link information is set in the citation source mail, ST205 becomes “YES” and the process proceeds to ST206. In ST206, schedule information corresponding to the citation mail is specified based on the link information, and date information and time information of the specified schedule information are read from the schedule information file 24.

  In the next ST207, it is checked whether a correction amount is stored in the date column or time column of the template. If this determination is “YES”, the process proceeds to ST208, and the information with the correction amount set in the information extracted in ST206 is corrected with the correction amount. Further, in ST209, the template information is rewritten with the date information after correction processing (including both date and time).

  On the other hand, if the correction amount is not stored in either the date column or the time column, ST207 is “NO”. In this case, in ST210 to 213, whether or not absolute expression information is stored in the date column and time column of the template is individually checked. If not stored, the information extracted in ST206 is stored in that column. Execute the process. In addition, when the information by an absolute expression is stored in the date column or the time column, the absolute expression is maintained.

  Next, when link information is not set in the citation source mail extracted in ST203, ST205 becomes “NO” and the process proceeds to ST215. In ST215, morphological analysis of the text information and subject information is executed for the citation source mail, and a schedule expression word related to the date and time is extracted.

  In the following ST216 to 225, processing for date information is executed. First, in ST216, it is determined whether or not an absolute expression is stored in the date column of the template. If an absolute expression is stored here, the process from ST217 onward is skipped and the process proceeds to ST226.

  If the date field of the template is empty, or if abstract expression information or the correction amount is stored, ST216 becomes “NO” and the process proceeds to ST217. In ST217, it is determined whether or not date information in absolute expression is extracted in ST215. If absolute expression date information is extracted, the process proceeds from ST217 to ST218, and it is determined whether or not a correction amount is stored in the date column of the template. If the correction amount is stored, the process proceeds to ST219, and the date information in the absolute expression extracted from the citation source mail is corrected with the correction amount. Further, in the next ST220, the date information of the template (that is, the correction amount) is rewritten with the corrected absolute expression.

  On the other hand, if the correction amount is not stored in the date column of the template, the process proceeds from ST218 to ST221, and the date information of the template is rewritten to an absolute expression extracted from the citation source mail.

  Next, in ST215, when date information in absolute expression is not extracted and date information in relative expression is extracted, ST216 and ST217 are “NO”, ST222 is “YES”, and the process proceeds to ST223. Then, it is determined whether or not the correction amount is stored in the date column of the template. If the correction amount is not stored, in the next ST224, 225, the relative expression extracted from the citation source mail is converted into the correction amount, and the date information of the template is rewritten with the converted correction amount.

  On the other hand, when the correction amount is already stored in the date column of the template, ST223 becomes “YES”, and ST224 and 225 are skipped. Also, in ST215, when neither the date information of absolute expression nor the date information of relative expression is extracted (when there is no corresponding word / or an abstract expression that cannot be replaced with the correction amount), ST217, 222 Are both “NO”, and the process proceeds to the next step without performing the template rewriting process.

  In ST226 to 235, the same processing as ST215 to 225 is executed for the time information. That is, when the absolute expression is not stored in the time column of the template and the time information of the absolute expression is included in the citation mail, information based on the absolute expression is stored in the time column. Further, when the correction amount is not stored in the time column of the template and the time information of the relative expression is included in the citation source mail, the correction amount converted from the relative expression is stored in the time column. .

  According to the processing of ST215 to 235, if at least one of date information and time information includes absolute expression information in the citation mail, the information in the corresponding column of the template is determined to be absolute expression. be able to. Further, when information on relative expression is included in the mail of interest or the predetermined citation mail, a correction amount for correcting the absolute expression can be obtained based on the relative expression.

  After executing ST215 to 235, in ST236, it is determined whether or not absolute expression information (excluding the correction amount) is stored in the date column and time column of the template. If this determination is “YES”, the process is terminated at this point, but if the determination is “NO”, the process proceeds to ST237, and the main pointer is moved to the citation source mail extracted in ST203. Thereafter, the process returns to ST203, and the citation source mail for the mail with the new main pointer (ie, the previous citation source mail) is extracted.

  Hereinafter, by repeating the above-described flow, while extracting the citation-related mails in the reverse order of the time flow, the template is rewritten using the extraction result and the process of extracting schedule information regarding the date and time for each extracted mail Execute the process. If a citation mail that is linked to the schedule information is extracted at a predetermined time, ST205 becomes “YES”, and the date and time are determined as absolute expression information by the processing of ST206 to 213 described above. .

  Further, if it becomes impossible to extract the citation source mail even after searching all the mails in the mail information file 15, the process proceeds from ST204 to ST214. In ST214, the mail transmission time or reception time to which the main pointer is attached is extracted as date information. Thereafter, the process proceeds to ST207, and thereafter, the same processing as when date information is extracted from the schedule information is executed.

  FIG. 9 shows a specific example in which schedule information creation processing is performed using seven mails M1 to M7 exchanged between friends A and B. In this example, for the sake of simplicity, only the creation of date / time information will be described, with a white frame for words corresponding to the absolute expression of date or time, and a halftone dot for words corresponding to the abstract expression. Frames and double lines are added to the words related to other schedules. Also, only the date column and time column are shown for the template (the upper column is the date column and the lower column is the time column).

  In this example, it is assumed that none of the mails M1 to M7 is linked to the schedule information. The first mail M1 includes absolute expressions “September 10” and “7 pm” regarding the date and time when the meal is scheduled. When creating new schedule information from the mail M1, “September 10” is extracted as date information and “7:00 pm” is extracted as time information, and stored in a template. The time information is converted to “19:00” prior to the template storage based on a preset format.

  The attribute data of the template created from the mail M1 is “new”. The schedule registration unit 22 newly registers the template in the schedule information file 24 based on the attribute data. Hereinafter, this registration information is referred to as “original data”.

Next, the flow of processing when creating the schedule information for correcting the original data from the mail with the latest mail M7 as a processing target will be described with reference to FIGS.
Since the mail M7 only includes time information of the relative expression “one hour ago”, it is necessary to obtain absolute date and time information by the date / time information conversion process. In the date / time information conversion process in this case, first, in ST201 of FIG. 7, “1 hour ago” extracted from the mail M7 is replaced with the correction amount “−1.0 hour” and written in the time field of the template. (See (a) in the figure.)

  Thereafter, each mail is extracted retrospectively by the extraction process of the citation source mail. The mail M6 extracted first includes time information of relative expression “1 hour ago”, but does not include date information. Therefore, after the mail M6 is extracted as the citation mail, ST216, 217, 222 in FIG. 8 are all “NO” and the process proceeds to ST226, where ST226 is “NO”, ST227 is “NO”, ST232 becomes “YES” and the process proceeds to ST233. In ST233, the time column of the template is checked, but since the correction amount is already stored in the time column at this time, the determination is “YES”. For this reason, the template M is not rewritten by the mail M6, and the process proceeds to the next citation mail extraction process.

  Next, since the extracted mail M5 does not include any expressions related to the date and time, after each step of ST216, 217, 222 becomes “NO” for the date, ST226, 227, Each step 232 is “NO”. Therefore, the rewriting of the template by the mail M5 is not executed, and the process proceeds to the next citation mail extraction process.

Next, the extracted mail M4 includes the absolute expression “7:30” for the time. Therefore, after ST216, 217, 222 are all “NO” and the process proceeds to ST226, ST226 is “NO”, ST227 is “YES”, and the process proceeds to ST228.
In ST228, the time column of the template is checked, but the correction value of “−1.0 hour” is already included in the time column at this point. As a result, ST228 becomes “YES”, the process proceeds to ST229, and the extracted absolute expression “7:30” is corrected by the correction amount “−1.0 hour”. This correction is performed after converting the absolute expression into a standard format (in this example, 19:30). In ST230, as shown in (b) of the figure, the time column of the template is rewritten with the corrected absolute expression “18:30”.

Next, the extracted mail M3 includes only the abstract expression “time”, so the template is not updated.
Next, the extracted mail M2 includes a relative expression “next day” for the date. As a result, ST216 and ST217 are “NO”, and ST222 is “YES”. Further, since the date field of the template is blank, ST223 becomes “NO” and the process proceeds to ST224, and the relative expression is converted into the correction amount “+1 day”. Further, in the next ST225, this correction amount is stored in the date column of the template. As a result, the template is updated to the contents as shown in (c) of the figure.

  Thereafter, when the mail M1 is extracted as the last citation mail, the date information “September 10” and “7 pm” in absolute expression are extracted. As a result, after ST216 becomes “NO” and ST217 becomes “YES” for the date, the correction amount “+1 day” is stored in the date column of the template, so ST218 becomes “YES”. Therefore, proceeding to ST219, the absolute expression is corrected by the correction amount, and in the next ST220, the corrected date information “September 11” is stored in the date column of the template.

  On the other hand, for the time, since the absolute expression “18:30” is already stored in the time column of the template, ST226 becomes “YES”, and the processing using the above “7 pm” is not performed. Therefore, the template is updated to the contents as shown in (d) in the figure.

  When the template update by the mail M1 is completed, ST236 becomes “YES”, and a series of processes related to the conversion of the date / time information is completed. Since the mail M7 includes the word “change”, the attribute data of the template is set to “change”. Therefore, the schedule management unit 2 that receives this template can rewrite the date column and time column of the original data based on the attribute data.

  Further, in the above example, the sender (A or B) of the received mail can be set for the personal name information. As for location information, although a specific expression cannot be extracted, the word “rice” in the mail M1 can be adopted as an absolute expression. Further, when there is a citation mail indicating a specific place in the mail 1, the location information can be specified from the citation mail when creating the original data.

  In the embodiment described above, the date / time information conversion process is executed when date information in absolute expression is not obtained regardless of the content of the mail being noticed. It may be determined whether or not there is an expression for changing or deleting the date / time in the mail by analysis, and the date / time information conversion process is executed when it is determined that the expression exists.

It is a functional block diagram of the information processing apparatus concerning one Example of this invention. It is explanatory drawing which shows the data structure of a schedule template. It is a flowchart which shows the creation processing procedure of schedule information. It is explanatory drawing which showed the quotation relation of the transmission / reception mail by the tree structure. It is explanatory drawing which shows the example which grouped the mail in a mail information file based on citation relation. It is a flowchart which shows the extraction process procedure of a quotation origin mail. It is a flowchart which shows the procedure of a date information conversion process. It is a flowchart which shows the procedure of a date information conversion process. It is explanatory drawing which shows the specific example which produces schedule information using several mail which has a quotation relation. It is explanatory drawing which shows the example by which a schedule is adjusted by exchange of mail.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Mail management part 2 Schedule management part 3 Schedule information creation part 31 Analysis processing part 32 Information management part 33 Abstract expression analysis part 34 Inter-mail relation analysis part

Claims (8)

Schedule information to be managed by the schedule management unit using the mail information managed by the mail management unit in an information processing apparatus having a mail management unit for managing transmission / reception processing of electronic mail and a schedule management unit for managing schedule information A program for setting the function to create
Analyzing means for analyzing the mail under the control of the mail management unit and extracting expressions related to the schedule;
When an expression indicating schedule correction is extracted by the analyzing unit, mails transmitted / received before the mail to be processed are sequentially searched and mails related to the mail to be processed are extracted and extracted. Correction target specifying means for analyzing the received mail by the analysis means and specifying the schedule information to be corrected;
Information correcting means for correcting the schedule information specified by the correction target specifying means using an expression indicating correction of the schedule;
A program for creating schedule information, which includes a program for giving the function of each means of the schedule information modified by the information modification means to a schedule management unit. Schedule information to be managed by the schedule management unit using the mail information managed by the mail management unit in an information processing apparatus having a mail management unit for managing transmission / reception processing of electronic mail and a schedule management unit for managing schedule information A program for setting the function to create
Analyzing means for analyzing the mail under the control of the mail management section and extracting expressions corresponding to a predetermined number of schedule items;
Schedule information creation means for assembling schedule information in which the expression extracted by the analysis means is associated with each item;
When the information of the predetermined schedule item is unconfirmed, a first search process for extracting mail related to the processing target mail by tracing back and forth the mail transmitted / received before the processing target mail and extracted The second search process for extracting the expression corresponding to the unconfirmed item from the mail is executed until the mail including the absolute expression corresponding to the unconfirmed item is extracted, and is extracted by the second search process. Conversion means for converting the uncertain information into an absolute expression using an expression;
Correction information output means for outputting the schedule information including the absolute expression converted by the conversion means to the schedule management unit as correction information for the existing schedule information;
A program for creating schedule information including a program for giving the function of each means to a computer.   3. The conversion unit according to claim 2, wherein, in the first search process, the conversion unit extracts a mail related to the processing target mail based on the citation information set by the mail management unit and the similarity of the mail document. A program for creating schedule information.   The conversion means includes correction means for correcting the finally extracted absolute expression with a correction amount converted from the relative expression when a relative expression related to uncertain information is extracted in the second search process. Program for creating schedule information described in 2 or 3. A mail management unit that manages email transmission / reception processing, a schedule management unit that manages schedule information, and a schedule that creates schedule information to be managed by the schedule management unit using mail information managed by the mail management unit An information processing apparatus comprising a creation unit,
The schedule creation unit
Analyzing means for analyzing the mail under the control of the mail management unit and extracting expressions related to the schedule;
When an expression indicating schedule correction is extracted by the analyzing unit, mails transmitted / received before the mail to be processed are sequentially searched and mails related to the mail to be processed are extracted and extracted. Correction target specifying means for analyzing the received mail by the analysis means and specifying the schedule information to be corrected;
Information correcting means for correcting the schedule information specified by the correction target specifying means using an expression indicating correction of the schedule;
An information processing apparatus comprising each means of: correction information output means for outputting schedule information corrected by the information correction means to a schedule management unit. A mail management unit that manages email transmission / reception processing, a schedule management unit that manages schedule information, and a schedule that creates schedule information to be managed by the schedule management unit using mail information managed by the mail management unit An information processing apparatus comprising a creation unit,
The schedule creation unit
Analyzing means for analyzing the mail under the control of the mail management section and extracting expressions corresponding to a predetermined number of schedule items;
Schedule information creation means for assembling schedule information in which the expression extracted by the analysis means is associated with each item;
When the information of the predetermined schedule item is unconfirmed, a first search process for extracting mail related to the processing target mail by tracing back and forth the mail transmitted / received before the processing target mail and extracted The second search process for extracting the expression corresponding to the unconfirmed item from the mail is executed until the mail including the absolute expression corresponding to the unconfirmed item is extracted, and is extracted by the second search process. Conversion means for converting the uncertain information into an absolute expression using an expression;
Correction information output means for outputting the schedule information including the absolute expression converted by the conversion means to the schedule management unit as correction information for the existing schedule information;
An information processing apparatus comprising the means. A mail management unit that manages email transmission / reception processing, a schedule management unit that manages schedule information, and a schedule that creates schedule information to be managed by the schedule management unit using mail information managed by the mail management unit A method executed by the schedule creation unit of the information processing apparatus including the creation unit,
A first step of analyzing an email that is managed by the email management unit and extracting an expression related to the schedule;
When an expression indicating schedule correction is extracted in the first step, mails sent and received before the mail to be processed are searched in order to extract mails related to the mail to be processed and extracted. A second step of analyzing the received mail by the analyzing means and identifying the schedule information to be corrected;
A third step of correcting the schedule information identified in the second step using an expression indicating the correction of the schedule;
A schedule information correction method, comprising: executing a fourth step of outputting the schedule information corrected in the third step to a schedule management unit. A mail management unit that manages email transmission / reception processing, a schedule management unit that manages schedule information, and a schedule that creates schedule information to be managed by the schedule management unit using mail information managed by the mail management unit A method executed by the schedule creation unit of the information processing apparatus including the creation unit,
A first step of analyzing an email entered under the management of the email management unit and extracting an expression corresponding to a predetermined number of schedule items;
A second step of assembling schedule information in which the expression extracted in the first step is associated with each item;
First search processing for extracting mail related to the processing target mail by tracing back the mails transmitted / received before the processing target mail in the first step when the information of the predetermined schedule item is unconfirmed And a second search process for extracting an expression corresponding to the unconfirmed item from the extracted mail until a mail containing an absolute expression corresponding to the unconfirmed item is extracted. A third step of converting the uncertain information into an absolute expression using the expression extracted in
A schedule information correction method comprising: executing a fourth step of outputting schedule information including the absolute expression converted in the third step to the schedule management unit as correction information for the existing schedule information.

Images