JP3373003B2 - Information retrieval device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for retrieving desired information from various data storage media, and more particularly to an information retrieval apparatus for interactively retrieving various information according to the purpose.

[0002]

2. Description of the Related Art A personal computer or workstation connected to a communication network can access a database in the network. In addition, these personal computers and workstations are CD-RO
It has a large-capacity storage device such as M and distributes the data stored in these storage devices via the above-mentioned network. In particular, a device such as a personal computer having a large-capacity storage device and connected to a network also has a function as a file server and is a system that stores a great amount of information as an entire network.

In such a system, a personal computer or the like constituting each terminal is more difficult to operate than a conventional medium such as a telephone or a television, which is an obstacle to restoration to a general home. Further, even if limited to text, there are people who can adapt to communication such as electronic mail and electronic news, and people who cannot adapt, which is an obstacle to sharing information using the system as described above. Therefore, there is a demand for a method that allows ordinary people to intuitively operate a personal computer to accumulate and distribute information and knowledge with little special training and learning. What makes the operation of the personal computer in such a system difficult is that the user must understand the processing and data structure in the personal computer in order to use it.

On the other hand, the inside of a telephone, television, etc. has complicated processing and structure, but the reason why they are well hidden from the user is that they have far simpler functions than a computer. This is also because the structure for accessing the information possessed by this means can only be written in a telephone directory or program table.

Since the functions of computers and the structure of information are much more complicated than those of telephones and televisions described above, it is inevitable that humans and computers interact with each other in order to obtain information and knowledge through personal computers. But,
Training and learning are indispensable in the method of using the conventional command as a means of interaction, and it is impossible for a child or an old person to use it.

Several techniques have been proposed for solving such a problem. For example, there is a so-called GUI (graphical user interface) for sequentially selecting target items by using a menu or buttons with icons and confirming the target information of the operator. This GUI
The method that uses is limited to a small number of selection items, and is not suitable for handling a large amount of complicated information.

Further, in the conventional GUI, as a method for intuitively understanding and locating the location of information, there is a method of associating an icon (graphic) in a virtual space called a desktop with information. However, it is difficult to always have an appropriate correspondence because the memory associated with a figure has a large individual difference, and since there is no action from the computer side in such a man-machine interface, what kind of system does the system have? In many cases, it was not intuitive to know if they had the ability or function.

There is also a method of using artificial intelligence as a user interface. With the help of this artificial intelligence, it is a method of solving the above-mentioned problems.

[0009]

There are many problems with the computer system described above. As a method for solving them, there are a method using GUI and a method using artificial intelligence. However, the conventional method using the GUI has a problem that it is difficult to appropriately determine what kind of ability and function the system has. Furthermore, when artificial intelligence is used as the user interface, there is a problem that the user inevitably makes requests or dialogues beyond the capabilities of the actual system because the range of artificial intelligence capabilities is unclear.

Further, since most of the knowledge information cannot be understood without actually seeing or listening, it is often difficult to describe or classify with a knowledge expression language such as a logical expression. If the information providing device depends on the knowledge representation,
It was difficult to provide the knowledge and information that the user requested. Further, there is a method of using an electronic news system, an electronic conference system, etc. by sharing knowledge information, but the information provided by such means also includes a large amount of information other than the knowledge and information that the user really needs. There were problems such as inclusion and operation, and problems such as large individual differences in adaptability to such text-based systems, and it was not usable by anyone in society.

According to the present invention, human beings of all ages, from children to old people, can naturally perform the selection of separation of knowledge and information and the interaction with a computer without special training or reference to a manual. To provide the environment.

[0012]

FIG. 1 is a block diagram showing the principle of the present invention. The present invention relates to a system for providing information by a multimedia computer.

The artificial personality section 1 is provided corresponding to the field of information to be provided, and requests target information in an interactive form with the operator in response to the operator's information request. The dialogue status management unit 2 manages the status of dialogue with the operator.

The dialogue dictionary section 3 stores information relating to dialogue. The dialogue control unit 4 searches the dialogue dictionary unit 3 according to the dialogue situation and controls the dialogue with the operator.

The input / output control section 5 is connected to the dialogue control section 4 and the artificial personality section 1 to control input / output relating to the dialogue.

[0016]

The signal relating to the dialogue input through the input / output control unit 5 is added to the dialogue control unit 4, and the dialogue with the operator is performed. In the dialogue, the dialogue control unit 4 uses the dialogue dictionary unit 3
, And outputs information relating to the dialogue from the required device when necessary. Then, the dialogue state at that time is managed by the dialogue state management unit 2. First, the artificial personality secretary section 1-1 peculiar to the user is selected by the dialogue in the dialogue control section 4. Subsequently, the professional-artificial-artificial-personality part 1-2 in the professional field that is shared and used by the user is also referred to, and the dialogue control part 4 makes a dialogue with respect to the professional field.

With such an artificial personality secretary section 1-1 and a specialized field artificial personality section 1-2, the artificial personality secretary section 1-1 for each user corresponds to the specialized field for each purpose. Specialized field Artificial personality department 1-2 responds and retrieves the required data. That is, the present invention utilizes a multimedia computer equipped with a function of recording and reproducing moving images and sounds to create a plurality of artificial personalities in the computer, and assigns such artificial personalities to the fields of knowledge and information. By associating, the user carries out the selection of the field as the selection of the interactive personality, and the knowledge and information in the field are searched for with the specialized personality of the field and visual information and voice such as face. It is driven by the dialogue used.

By this search, it is possible to provide information without recognizing the knowledge of a computer such as a specialty or the specialized information in the specialized field.

[0019]

The present invention will be described in detail below with reference to the drawings. FIG. 2 is a system configuration diagram of the present invention.

The input section 10 includes a microphone 11 and the microphone 11
Voice recognition unit 12 for recognizing more input voice, mouse 1
3 and a menu control & button control unit 14 for controlling the position information of the menu input by the mouse, a keyboard 15, and a command analysis unit 16.

First, the input unit 10 inputs a user's utterance or operation by voice. That is, the voice signal input from the microphone 13 is recognized by the voice recognition unit 12. Secondly, the menu control & button control unit 14 recognizes the position of the menu designated by the mouse cursor. Further recognize when the mouse button is pressed,
The menu information is output to the command analysis unit 16.
Thirdly, the key information input by operating the keys of the keyboard 15 is input to the command analysis unit 16.

That is, as described above, the input unit 10 accepts a user operation from a voice, a menu selection, a button, a keyboard and the like. These are converted into the same command format through their respective input control units and recognition units, and the command analysis unit 16 is notified of the occurrence of an event.

The command analysis unit 16 analyzes the command structure triggered by this input event, and analyzes the analysis result as the dialogue management unit 1.
9 to the dialogue control unit 20. Since the analysis of the command structure in the command analysis unit 16 changes depending on the situation of the dialogue, for example, the situation of using the voice, mouse, keyboard, the meaning of the dialogue is not analyzed here.

The dialogue management section 19 controls the dialogue with the operator and comprises a dialogue control section 20, a dialogue situation management section 21, and a dialogue dictionary 22. Dialog control unit 20
Notifies the dialogue situation management unit 21 of changes in the dialogue situation, and the dialogue dictionary 22 indicates the meaning depending on the dialogue situation of the command.
To understand and to control the output control unit 23 according to the conversation status. A speaker 26 is connected to the output controller 23 via a display device 24 and a voice synthesizer 25. If the output control unit 23 does not perform voice synthesis, the output control unit 23 directly outputs a signal of a confirmation sound generated when the mouse such as "Pi" is operated to the speaker.

The dialogue dictionary 22 manages the vocabulary of commands available to the user and judges the meaning of the command according to the situation. When the situation and command are notified from the dialogue control unit 20 described above, the dialogue situation management unit 21 is notified of the meaning, that is, the management information.

The dialogue situation management unit 21 manages dialogue situations such as which artificial personality the user is currently talking with and the necessary conditions for the response action corresponding thereto, and the response behavior depending on the personality of the artificial personality. Then, appropriate control for causing the artificial personality to take a response action is performed, and the dialogue control unit 20 is notified of the change in the situation resulting from the response action.

The artificial personality unit 30 is composed of an artificial personality secretary control unit 31 unique to a user and specialized field artificial personality control units 32 and 33 shared by a plurality of users on a network. Each of the artificial personality control units 31, 32-1, 32-2,
Behavior control units 33-1 to 33-3 of each personality, moving image / sound databases 34-1 to 34-3, knowledge / information databases 35-1 to 35-3, and knowledge / information map units 36-1 to 36-3.
36-3. When performing a response action to the user, based on the control of the action control unit 33-1 to 33-3, the output to the output control unit 23 using the moving image and the audio signal of the moving image / audio database 34-1 to 34-3. Give information.
Based on this output information, the output control unit 23 outputs a voice to the speaker 2
From 6, the moving image is output from the display device. That is, the target information related to the dialogue is output from the target device.

Further, in order to allow the user to intuitively search for knowledge and information, the knowledge and information of the artificial agent are visually laid out at various levels and displayed as a knowledge / information map. . The knowledge / information map units 36-1 to 36-3 interactively scroll the display position, change the display magnification, and switch to the knowledge / information map laid out from another viewpoint by interacting with the user.

Links are defined between the knowledge / information map units 36-1 to 36-3 and the knowledge / information databases 35-1 to 35-3, and the knowledge and information decided by the user are reached by using the links. It becomes possible.

The artificial personality secretary is provided for each user. On the other hand, the specialized field artificial personality control unit 32 is different depending on the specialized field, and the specialized field artificial personality A 32-1 is the behavior control unit 33-2, the moving image / voice database 34-2, the knowledge / information base 35-2, the knowledge /
It is composed of the information map section 36-2, and it is an expert personality B.
32-1 is a behavior control unit 33-3, a video / audio database 34-3, a knowledge / information database 35-3, a knowledge /
Each field is divided as configured by the information map unit 36-3. In addition, in FIG. 2, an artificial personality secretary is provided for one user, and when there are a plurality of secretaries, a plurality of artificial personalities are provided respectively.

The operation of the system shown in FIG. 2 will be described with reference to FIG. When a dialogue is performed by the user's voice utterance (I1), conversion into a command is performed by voice recognition in step S1. In step S1, the voice recognition unit 12 and the command analysis unit 16 operate to recognize the voice. However, the recognition in step S1 is only recognition of voice, and the content of the command is not analyzed.

Further, when an input is made by a user menu operation, for example, by a mouse (I2), the menu control & button control unit 14 converts a command in step S2. When a user inputs a command by operating the keyboard (I3), the key is input.

The information converted in steps S1 and S2 and the information input from the keyboard are command analyzed in step S3. This step S3 is the command analysis unit 1.
6 control operation. And step S4
Then, the dialogue control unit 20 senses the dialogue situation and recognizes the situation. For example, the change of the situation with respect to the input command is changed in step S5. Then, in step S6, the meaning of the command is searched by searching the interactive dictionary unit 22. Finally, the selection of an artificial personality suitable for the dialogue situation is step S7.
Made in.

In the embodiment of the present invention, firstly, the dialogue action of the personal assistant personal assistant is performed in step S8-1. Then, in step S9, the synchronous control of the output of the moving image and audio data with respect to the dialogue of the artificial personality secretary is performed.

Subsequently, the same processing as described above is repeated for the display output. At the time of this repetition, the dialogue behavior S8-2 to S8-3 of the target expert's artificial personality is performed.
Is selected. That is, after the step S7, the dialogue behavior of the specialized field artificial personalities AZ is performed in the steps S8-2 and S8-3.
Selected in. Then, after the selected interactive action is performed, output synchronization control of the moving image and audio data is performed and output in step S9.

Such repetition is sequentially performed, and a display corresponding to the user and a dialogue are made depending on the purpose. For example, dialogue and behavior of artificial personality secretary is first "Hello" is output by voice when you performed in step S8-1, followed by "What kind of information the Looking" with output in voice, artificial personality secretary Correspondingly display the knowledge map. For example, the behavior control unit 33-1 of the artificial personality secretary makes a step S8-1.
When the video and audio data are output in step S9, the example of displaying the knowledge map of the artificial personality secretary as shown in FIG. 4 is "Today's news" M1, "Cooking, leisure" as shown in the figure.
M2, "Literature and Linguistics" M3, "Medical Science and Information Engineering" M4
Etc. message is displayed.

Then, when an instruction to introduce medical specialization by voice is added in the response of the user, the artificial personality secretary outputs "Then, Dr. Garrigari will be introduced." The operation up to this point is the operation of the artificial personality secretary, and the action control unit of the artificial personality in the intended specialized field is activated. After this, it will be a self-introduction by an expert of artificial personality. For example, "Medical specialization Dr. Garigari. I will show you what I understand." Is output by voice and displayed as shown in the example of the medical specialization knowledge map display in Fig. 5. Here, whether the symptom-based coping method N2 is an internal medicine N3 or a surgery N4 is displayed, and further, a first aid N5, a nursing method N6, a preventive method N7, and the like are displayed. Here, when the user instructs the first aid N5 with a voice, a mouse, a keyboard, or the like, and then instructs "Tell me first aid for burns", the behavior control unit 33-2 of the artificial character of the target specialized field. , 33-3, the method for dealing with burns is output by drawing or voice. For example, as shown in a display example diagram of information provision shown in FIG. 6, a message such as “cool well with cold water” is displayed.

Further, referring to FIG. 7, the operation of the artificial personality secretary and the professional personality personality in the embodiment of the present invention will be described. When the knowledge information providing apparatus of the present invention starts to operate, the artificial personality secretary is activated in step S20. When the artificial personality secretary is activated, that is, when the activation instruction is input to the action controller 33-1 of the artificial personality secretary, the action controller 33-1 first presents the knowledge map in step S21. The presentation of the knowledge map is, for example, a display as shown in FIG. When a request (REQ) from the user is added to the presentation of this knowledge map as a command, the command is understood in step S22. In this step S22, the dialogue status management unit 21 refers to the dialogue dictionary 22 to perform state management. At this time, all of these dialogues are controlled by the dialogue control unit 20.

Since the behavior control section 33-1 of the artificial personality secretary is activated at the start of the processing, the knowledge / information map 36-1 is presented and the knowledge / information database 35-1 is provided.
All of the search for are performed in step S23 under the control of the behavior control unit 33-1 of the artificial personality secretary. Then, the associated control information is input to the dialogue status management unit 21 and the dialogue dictionary 22 and updated. Then, the behavior control units 33-2, 33-3 of the specialized field artificial personality corresponding to the instructed one
Is started in step S25. When the behavior control units 33-2 and 33-3 of the selected artificial personality expert are instructed, first, the self-introduction knowledge map presentation corresponding to the selected artificial personality is displayed to the user in step S26. A user who is an operator inputs a command corresponding to the presented knowledge map and understands the command in step S27. This is performed by referring to the dialogue status management unit 21 and the dialogue dictionary 22. These are also made interactive by the dialogue control unit 20. And further step S2
At 8, the control information of the behavior control is input to the dialogue status management unit 21 and the dialogue dictionary 22 and updated. Subsequently, in step S29, knowledge information corresponding to each purpose is provided to the user.

The above-described operation activates the artificial personality secretary, and subsequently activates the corresponding artificial personality expert to obtain the desired knowledge information. Although the present invention has been described in detail with reference to the embodiments of the present invention, the present invention is not limited to this. For example, in the embodiment, knowledge / information databases 35-2 and 35-3 are provided in the professional field personality control units 32-1 and 32-2, respectively, but this is in the file server connected to the network system. Information may be used. At this time, the target file is accessed via the network server. The knowledge / information map units 36-2 and 36-3 are also provided correspondingly. Further, the professional field personality may be provided on a CD-ROM or the like. Further, although the artificial personality secretary control unit 31 is provided in each personal computer, the artificial personality secretary control unit 31 is not limited to this and may be provided in a specific file in a centralized manner, and may be read and used via a network server when used.

In such a system, the artificial personality secretary control unit 3 for each individual responds to additions and corrections of specialized fields.
One knowledge / information map can be updated at once.

[0042]

As described above, according to the present invention, a plurality of artificial personalities are provided in a computer in association with the fields of knowledge and information, and the search for knowledge and information within the field is performed in that field. By providing a means for advancing by dialogue using visual information or voice as an artificial personality of the expert, the user can carry out the selection of the field as the choice of the artificial personality with which to interact. As a result, by using a multimedia computer equipped with a recording / playback function for multimedia information such as video and audio, a communication network, and a large-capacity recording medium such as a CD-ROM, all layers from children who cannot read characters to elderly people Users can naturally interact with computers to gain knowledge and information.

[Brief description of drawings]

FIG. 1 is a principle block diagram of the present invention.

FIG. 2 is a system configuration diagram of the present invention.

FIG. 3 is an explanatory diagram of a flow of each process.

FIG. 4 is a diagram showing a display example of a knowledge map of an artificial personality secretary.

FIG. 5 is a diagram showing an example of displaying a knowledge map of medical specialization.

FIG. 6 is a diagram showing a display example of information provision.

FIG. 7 is a flowchart of an artificial personality dialogue.

[Explanation of symbols]

1 Artificial personality department 1-1 Artificial personality secretary department 1-2 Specialized Field Artificial Personality Department 2 Dialogue management department 3 Dialogue dictionary 4 Dialogue control section 5 I / O controller

Continuation of the front page (56) References JP-A-2-96874 (JP, A) JP-A-4-252375 (JP, A) Beerud Sheth, Pattie Maes, Evolving Agents For Personalized Information Filtering, Artificial Filtering, Artificial Filtering Intelligence for Applications, 1993. Proceedings. , Ninth Conference on, USA, IEEE, March 5, 1993, pp. 345-352 Yoshitaka Hiramoto, Hiroshi Doi, Mitsuru Ishizuka, Guidance system by visual software agent with voice function, 1993 Electronic information Proceedings of the Autumn Meeting of IEICE 1. Fundamentals / Boundaries, Japan, The Institute of Electronics, Information and Communication Engineers, August 15, 1993, p167 Kengo Omura, Takashi Sonoda, Takashi Matsumoto et al., Introduction of Interface Agents to Direct Manipulation Environment, The 8th Human Interface Symposium Proceedings, Japan, Society of Instrument and Control Engineers, October 21, 1992, pp. 105-110 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06F 17/30 G06F 9 / 44 JISST file (JOIS)

Claims (3)

(57) [Claims] 1. A system for providing information by a multimedia computer , which corresponds to the field of information shared and provided by a plurality of users.
Information provided by the user in an interactive manner with the user.
Get what the news is and use it to get the corresponding information
An artificial personality section that outputs to an input / output device and an initial operation of the system that is provided for each user
Sometimes, in an interactive manner with the user,
Get what the field is and respond based on it
An information retrieval device comprising: an artificial personality secretary section for activating the artificial personality section . 2. An input / output control unit for controlling input / output relating to interaction with a user in an information providing system using a multimedia computer.
And a dialog status management unit for managing the status of dialog with the user
Then, the meaning of the user command is searched to determine the meaning.
A dialogue dictionary section for notifying the dialogue status management section and a command of the user acquired from the input / output control section.
Notify the above-mentioned dialogue dictionary section, and based on the notification from other sections.
And an interactive control unit for controlling the input / output control unit, and corresponding to the field of information shared by a plurality of users and provided.
Provided by controlling the dialogue status management unit
And interactively ask what the user is looking for.
Obtained, and based on that, the input / output via the dialogue control unit
Specialized Artificial Personality Department that outputs information corresponding to the control unit
And the initial operation of the system provided for each user
Sometimes the conversation status management unit is controlled to interact with the user.
Get what the field of information the user is looking for in a formula
Then, based on that, the corresponding specialized field for artificial personality was established.
An information retrieval device comprising an artificial personality secretary unit that operates and an information retrieval device. 3. Artificial personality department and artificial personality secrecy in the specialized field
The writing section is a behavior control section that controls its own behavior and a knowledge / information database that holds the knowledge and information it possesses.
And the knowledge and information held by the knowledge / information database section.
Change the visual layout based on the interaction with the user
And knowledge / information map section for presenting performing like, retains the video and audio data, from the action control unit
Video and audio data are sent to the input / output control unit by notification.
The information retrieval apparatus according to claim 2 , further comprising a moving image / sound database unit for outputting .