KR20040054867A - Information extraction and service system for appointment scheduling - Google Patents

Abstract

PURPOSE: A system for managing a schedule and offering information using a natural language is provided to extract the appointment information from a dialog between users, easily manage the schedule without a registration procedure, and offer the living information corresponding to the schedule when the schedule is offered. CONSTITUTION: A member registration module(200) registers a member by receiving the personal information of the user. A natural language analysis module(300) analyzes the used natural language between the user and other user. An appointment information extracting module(400) extracts the appointment information from the analyzed natural language. A schedule management module(500) stores the extracted appointment information in a schedule database(140). A user information management module(600) understands/stores the personal information and the preference of the user in a user information database(120). A living information providing module(700) extracts the living information corresponding to the preference from a living information database(160) and provides it to the user.

Description

Information extraction and service system for appointment scheduling using natural language

The present invention relates to a schedule management and information providing system using natural language, and more particularly, to managing a schedule of a user and providing a schedule to the user, the schedule management using a natural language that provides living information corresponding to the schedule and An information providing system.

Modern society is a fast and complicated society, where credit is important, and modern people spend a lot of their day-by-day busy times, often causing unintentional violations of their promises. Each individual writes his or her schedule in diaries, notes, electronic notebooks, or relies on memory to maintain credit.

In such a case, the appointment may be broken because it is not remembered in a timely manner, and in order to fulfill the appointment, it is cumbersome to check the diary at any time. A reminder schedule management system is presented.

However, such a schedule management system reminds the user that there is an appointment and helps to fulfill the appointment, but it is cumbersome to input all the details of the appointment into the system, thereby wasting unnecessary time. If the schedule is not input due to user's carelessness, there is a problem of forgetting the appointment itself.

In addition to the schedule management system to help the busy life of modern people, product recommendation systems are provided to determine the user's preferences and then search for the product is provided, as the amount of information is getting more and more difficult to find the optimal product There is a problem in that a product provided in one system is limited to a specific item, and a product recommendation is made only at the request of a user.

Accordingly, the present invention is to solve such a conventional problem, the object of the present invention is to extract appointment information from the contents of the conversation between the user, and easily manages the user's schedule without a separate registration process, the schedule to the user In providing, it is intended to provide the living information corresponding to the schedule.

Other objects and features of the present invention will be more clearly understood through the preferred embodiments described below.

1 is a network configuration to which the schedule management and information providing system of the present invention is applied.

Figure 2 is a block diagram showing the configuration according to the function of the schedule management and information providing system of the present invention.

3 is a flowchart showing a natural language analysis step according to the present invention.

4A to 4C illustrate an embodiment of extracting appointment information according to the present invention.

5A to 5C illustrate another embodiment of appointment information extraction according to the present invention.

6 is a diagram illustrating an embodiment of a method of notifying appointment information to a user according to the present invention.

7A to 7C illustrate another embodiment of a method of notifying appointment information to a user according to the present invention.

According to the present invention, it is applied to a schedule management and information providing system comprising a user having a wired / wireless terminal, an operation server having a user connected via the Internet, a schedule database storing a user's schedule, and a living information database storing living information. Member registration module for receiving user's personal information and registering as a member, natural language analysis module for analyzing natural language used for communication between user and other users, appointment information for extracting appointment information from natural language analyzed from natural language analysis module Extraction module, Appointment information extraction module receives appointment information from the schedule management module for storing in the user schedule database, user information management module for identifying and storing the user's personal information and preferences in the user information database, user information management module preference A life information providing module that provides a diagram, extracts life information corresponding to a user's preference from a life information database, and provides the user with the life information corresponding to a user's schedule and controls data input / output to each database. And a schedule management and information providing system using natural language is characterized in that it comprises a control module for controlling the function of each module.

In addition, the natural language analysis module analyzes the natural language through morpheme analysis, morpheme tagging, syntax analysis, semantic analysis, and discourse analysis.

Preferably, communication may be performed in one of a server / client method or a P2P method, and the appointment information extracting module may extract each vocabulary related to time, place, human name, and behavior from natural language and analyze it according to the ground rule.

In addition, after receiving the appointment information, the schedule management module asks the user whether to register the schedule, and if the user indicates the intention to register, the schedule management module stores the information in the user information database, and generates an alarm before the corresponding time and a predetermined time of the appointment information. Notify user of appointment information.

In addition, the user information management module may be able to determine a user's preferences by extracting a high frequency of interest vocabulary from natural language, and may also determine preferences based on the user's personal information. Preference may also be determined.

In addition, the user information management module extracts other user's personal information from natural language and stores it in another person's information database.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

1 is a network configuration to which the schedule management and information providing system of the present invention is applied.

As shown, a wired terminal user, such as computer system 30, may access the operating server 10 via the Internet 20, and a wireless terminal user, such as mobile phone 34 or PDA 36, may use a mobile communication company. It is possible to connect to the operation server 10 via the Internet 20 via the server 32.

The operation server 10 refers to a concept in which hardware and software for a portal site for providing living information while managing a schedule of a user are collected. A schedule database storing a user's schedule and a living information database storing living information It is provided.

Figure 2 is a block diagram showing the configuration according to the function of the schedule management and information providing system of the present invention.

The control module 100 controls data input / output to each of the databases 120 to 160 and controls the functions of the modules 200 to 700.

In a preferred embodiment, the database controlled by the control module 100 includes a user information database 120 storing user personal information and preferences, a schedule database 140 storing a user's schedule, and providing the user to the user. It consists of a living information database 160 for storing a variety of living information for, and other people information database 180 for storing the personal information of other users, each database may be more subdivided or more extensively configured by history.

The member registration module 200 receives the user's personal information and registers as a member, stores and manages the personal information provided from the user in the user information database 120, and the personal information is determined by the age, It is desirable to include hobbies, skills, interests, and the like.

The natural language analysis module 300 exchanges a conversation between the user and another user in a natural language through an email, a chat, a short text service (SMS), etc. At this time, the conversation content between the two users is analyzed.

Natural language is a common language that humans use in their daily lives. Because of their vocabulary and many different meanings, the sentence structure of the same meaning can be expressed differently according to expression and thinking ability. As it can be changed accordingly, it is necessary to fully consider the context and situation before and after understanding natural language.

The appointment information extraction module 400 extracts appointment information from the natural language, and after the natural language analysis module 300 analyzes the natural language, extracts vocabulary related to the appointment information from the analyzed natural language.

In addition, the appointment information extraction module 400 extracts each vocabulary related to time, place, human name, and behavior from the analyzed natural language, according to the ground rules (Why, What, Whom, Who, When, Where) for the user's convenience. Analyze

Table 1 shows, for example, how the appointment information extraction module 400 extracts vocabularies corresponding to the types of appointment information.

Table 1

division Vocabulary extraction method and example Appointment time Time Vocabulary Extraction: Tomorrow, September 1, Wednesday, next week, the last… Calculate time according to context and today's time Choose the correct appointment time when multiple hours exist for one appointment Appointment Promise Behavior Vocabulary Extraction: See, See, Meeting, Attend… Promise Behavior Vocabulary Pattern: Meet / from / within Place of appointment Place Vocabulary Extraction: Seoul Station, Seoul Department Store, Place Name… Promise place extraction by core vocabulary or pattern about place Extraction of place by promise vocabulary pattern Promise Include sender and receiver by default

The schedule management module 500 receives the appointment information from the appointment information extraction module 400, stores and manages each user in the schedule database 140, and alarms a predetermined time of a date and time corresponding to a user's schedule. By generating a notification of the appointment information to the user, so that the user can fulfill the appointment.

An alarm is a general term for vibration, beep, lamp, message, and the like. For example, a user having a mobile phone 34 may transmit appointment information using a short text service, and a user who is using the computer system 30. You can inform the appointment information through a pop-up window and a beep.

In addition, the schedule management module 500, after receiving the appointment information from the appointment information extraction module 400, asks the user whether to register the schedule, and if the user indicates the intention to register and stores in the schedule database 140, the user registered If you indicate that you do not want to delete the appointment information is preferably configured.

The user information management module 600 grasps the user's personal information and preferences and stores and manages the information in the user information database 120.

Preferably, the user's preference grasp may be determined by extracting an interest vocabulary having a high frequency of use from the natural language analyzed from the natural language analysis module 300, and the member registration module 200 receives personal information (age, interest) input from the user. Field, hobbies, etc.), and can be determined by the user's life information used in the past.

In addition, the user information management module 600 according to an embodiment of the present invention extracts other users' personal information (name, address, phone number, email, etc.) from the natural language analyzed from the natural language analysis module 300 and others The information is stored in the database 180 and managed.

The living information providing module 700 receives the user's preference from the user information management module 600, extracts the living information corresponding to the user's preference from the living information database 160, and the living information corresponding to the user's schedule. To the user.

For example, if the user's appointment information is an appointment for the purpose of purchasing a friend's birthday present, the friend's birthday is used to extract a list of products preferred by the age group to which the user belongs and schedule details such as product name, sales place, and amount. To the user together.

According to an embodiment of the present invention, the living information includes various product information, cultural information (schedules and tickets for movies, plays and performances), regional restaurant information, geographic information of the user's current location or destination, travel information, and the like. .

3 is a flowchart showing a natural language analysis step according to the present invention.

Natural language analysis module 300 of the schedule management and information providing system using the natural language according to the present invention is morpheme analysis (S10), morpheme tagging (S20), syntax analysis (S30), semantic analysis (S40), discourse analysis (S50) Analyze natural language through the steps of

Morphological analysis (S10) receives a word as an input, separates it into morphemes, and extracts parts of speech information and information necessary for analysis.

Morphological tagging (S20) is a task of attaching a tag that is additional language information. Natural language has a lot of inherent importance in terms of vocabulary, syntax, and semantics, and when the raw raw corpus is used as it is, accurate language information is extracted. Since it is impossible to do so, tagging can eliminate neutrality, and the correct words based on the context are analyzed and selected.

Phrase analysis (S30) is to analyze the syntax and syntactic structure of the sentence, semantic analysis (S40) is to analyze the meaning in the sentence and semantic relationship, discourse analysis (S50) is to the context (multiple sentences) By analyzing the meaning and analyzing the semantic relationship.

Table 2 shows the step-by-step results for processing natural language, taking the sentence "I went to Busan for company work" as an example.

Table 2

step result 1) Morphological Analysis I I (pronoun) + is (subordinate) or (verb) + is (mother) day (verb) + is (mother) Company Company (noun) + Work (noun) Company (noun) + Lee (investigation) + ㄹ (mother) 2) Morphological Tagging I I (pronoun) + is (assistant) Company Company (noun) + work (noun) 3) parsing Go (past) Subject: I (pronoun) Adverb: Busan (unique noun) + E (investigation) 4) semantic analysis Go (past) Subject: Place: Busan Reason: Company 5) Discourse Analysis Go (past) Subject: Place: Busan Reason: Company

4A to 4C illustrate an embodiment of extracting appointment information according to the present invention.

4A illustrates an example of a server / client method, which shows a message delivered to a user through an email. The message content is analyzed by the natural language analysis module 300 and the appointment information extraction module 400 analyzes appointment information from the analyzed message. Extract

The extracted appointment information is shown to the user, including contents such as date, title, alarm, etc., as shown in FIG. 4B. If the user intends to attend the meeting because he or she is willing to attend the meeting, the user clicks the register button and does not attend the meeting. If you click the Cancel button, it will not be registered in the schedule.

When the user clicks the registration button, as shown in FIG. 4C, appointment information is stored for each day to continuously manage the user's schedule, and according to the operation of the schedule management and information providing system using natural language according to the present invention. In addition to extracting appointment information, it is preferable to configure a schedule registration by a user.

5A to 5C illustrate another embodiment of appointment information extraction according to the present invention.

5A illustrates an example of a P2P method, in which a user exchanges a conversation with another user in real time through a messenger. The conversation content is analyzed by the natural language analysis module 300 and the appointment information extraction module 400 is analyzed. Extract appointment information from the conversation.

Unlike the server / client method, in the P2P method, since the operation server 10 does not recognize that the communication between the user and other users is made, in a preferred embodiment, the conversation contents are transmitted to the operation server 10 after the conversation ends. The conversation may be configured to be transmitted to the operation server 10 in the middle of the conversation.

The user makes an appointment to meet with another user to buy a gift, and the extracted appointment information is shown to the user, including contents such as date, time, place, destination, and alarm, as shown in FIG. 5B. It is registered in the schedule as shown in 5c.

6 is a diagram illustrating an embodiment of a method of notifying appointment information to a user according to the present invention.

Referring to the accompanying drawings, it shows an example of the screen of the mobile phone 34 that informs the user by using the short text service transmitted to the mobile phone 34, the appointment information extracted according to the dialogue content of Figure 5a, By informing the time, place, and object, and providing the living information that the user deems necessary by the menu, the service can be used by the user's needs.

7A to 7C illustrate another embodiment of a method of notifying appointment information to a user according to the present invention.

Referring to FIG. 7A, when a user accesses an operation server using the computer system 30, the schedule management module 500 searches for a schedule of the corresponding user, searches whether appointment information exists on a corresponding date, and exists appointment information. If the pop-up window informs the user that the appointment information.

When the user wants to know the details of the appointment information, as shown in FIG. 7B, the appointment information is divided by history and displayed to the user. When the user clicks the recommendation information button at the bottom, various products corresponding to the purpose of the appointment information are displayed. Provided as shown in FIG. 7C.

Referring to the accompanying drawings, since the main purpose of the appointment information is to purchase a gift, a list of various products appear by the preferences grasped from the user information management module 600, and will be accompanied by a meal for ancillary purposes In addition, the restaurant near the appointment place is provided to the user as shown in FIG. 7D.

In a preferred embodiment of the present invention, when the user is carrying the mobile phone 34 or PDA 36, the user's location is determined and various living information available at the current location (restaurants, lodging facilities, gas stations) Location, etc.), and provide the location and the shortest distance of the place the user wants to move.

Although the above has been described with reference to the preferred embodiment of the present invention, various modifications and changes are possible without departing from the scope of the present invention. Therefore, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification but should be defined by the claims of the patent.

As described above, the schedule management and information providing system using natural language according to the present invention automatically extracts and registers appointment information from daily conversation contents even if a user does not directly input a schedule, so as to manage his schedule. This saves effort, hassle and time.

In addition, by providing the living information corresponding to the user's situation and preferences, the user can obtain the optimal living information, it is possible to save the effort and time required to find a variety of information.

Claims (10)

Applied to a schedule management and information providing system comprising a user having a wired / wireless terminal, an operation server having a schedule database storing the user's schedule and a living information database storing living information, wherein the user is connected through the Internet, Member registration module for receiving a user's personal information to register as a member; A natural language analysis module analyzing natural language used for communication between the user and another user; An appointment information extraction module for extracting appointment information from the natural language analyzed from the natural language analysis module; A schedule management module receiving the appointment information from the appointment information extraction module and storing the appointment information in the schedule database; A user information management module which grasps the user's personal information and preferences and stores them in a user information database; Life information for receiving the user's preference from the user information management module, extracting the living information corresponding to the user's preference from the living information database, and providing the user with the living information corresponding to the user's schedule A providing module; And a control module for controlling data input and output to each of the databases and for controlling the functions of the respective modules. The system of claim 1, wherein the communication is performed in one of a server / client method and a P2P method. The system of claim 1, wherein the natural language analysis module analyzes the natural language through morpheme analysis, morpheme tagging, syntax analysis, semantic analysis, and discourse analysis. The system of claim 1, wherein the appointment information extracting module extracts each vocabulary related to time, place, human name, and behavior from the natural language and analyzes the vocabulary based on land rules. The schedule management module of claim 1, wherein the schedule management module asks the user to register a schedule after receiving the appointment information, and stores the schedule information in a user information database when the user indicates a registration intention. And informational systems. The schedule management module of claim 1, wherein the schedule management module generates an alarm in advance of a corresponding time and a predetermined time corresponding to the appointment information to inform the appointment information to the user. system. The system of claim 1, wherein the user information management module extracts an interest vocabulary having a high frequency of use from the natural language to determine a preference of the user. The system of claim 1, wherein the user information management module determines a preference based on the user's personal information. The system of claim 1, wherein the user information management module determines a preference based on the living information used by the user in the past. The system of claim 1, wherein the user information management module extracts the personal information of the other user from the natural language and stores the personal information in another person information database.