JP3083657B2 - Human interface system and control method thereof - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a human interface when using computers connected by a distributed network.

[0002]

2. Description of the Related Art A conventional human interface of a computer system has been developed mainly from the viewpoint of supporting various operations in a single computer.
Recently, multiple windows have been placed on a bitmap display to handle multiple tasks in parallel, and various graphical computers have been used to create and delete documents and manage data using spreadsheets. Tools are becoming relatively easy to use. In addition, networking of computers has progressed, and various services such as electronic mail have become available via local area networks and wide area networks.

[0003]

The human interface of the conventional computer system models various tools of the computer by a concept called a so-called desktop metaphor and shows it to the user. Such a human interface makes it easy for the user to access and operate various conventional single computer tools. These are basically used by the user to learn the functions and operations of the tool. On the other hand, the number of services supported by computers has become extremely large, and when a real user performs a certain job or job, he or she searches for data in a location that the user does not know in advance and refers to it. In addition, the use of various computer services that are not frequently used is increasing. It is difficult to cope with such a request using the conventional framework of the computer human interface.

The present invention has been made to solve such a conventional problem, and an object of the present invention is to provide a human interface system capable of functionally performing an action similar to the action in the real world. To provide.

[0005]

In order to achieve the above-mentioned object, according to the present invention, a plurality of networks connected by a network are provided.
In a human interface system for using a distributed system having a computer , a plurality of agents are provided.
Holds the definition information of each state and processing operation of the
Manage, before the configured site by the agent
A site construction unit configured on each of the computers based on the definition information, and a processing operation of the agent according to an access content from a user to the constructed site.
Access management means for controlling based on the definition information ;
Request arbitration processing according to the state and processing operation of the agent
And communication path setting means for setting a communication channel between agents .

Further, the access management unit analyzes the request recognizes the available data from the user, the Agent
An automatic response function of outputting to create a response to the analysis result from the knowledge information stringent has, the automatic answering machine
Switching between response processing by Noh and interactive processing by real users
Switching function .

[0007]

SUMMARY OF] According to the present invention configured as described above, allocation, for example, office room as a network consisting of a system having a plurality of computers, various objects (desk, tools, etc.) various virtual site from therein Is constructed. Then, the access management means allows the user to access and move to the virtual site and obtain desired information.

An automatic answering machine is used as access control means.
When a user accesses another user and performs some kind of exchange, simple and routine exchanges can be automated and only when necessary. in fact users Ru can be realized features, such as corresponding. Also use this interface
This means that compared to the conventional interface to a single program, the user moves around the world inside the computer and interacts with other people. It is very easy to find and refer to data in a location unknown to the user, and to use various computer services that are not frequently used.

[0009]

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

FIG. 1 shows a computer network configuration according to an embodiment of the present invention. Reference numerals 11 to 14 denote computers such as workstations and personal computers.
They are connected by local networks 15, 17 and a wide network 16. Through these networks, each computer can perform data communication.

FIG. 2 is a block diagram of the computer according to the embodiment shown in FIG. In FIG. 1, a display unit 1 is a display device such as a display device. The input unit 2 is a keyboard
The user inputs a character input such as a mouse, a command input, a pointing of a position on a display screen, and the like. The communication unit 3 inputs and outputs data to and from an external network.
The storage unit 4 is a storage device such as a disk. The agent processing unit 5 supports agent programming. The control unit 6 controls the above elements 1 to 5. In the present embodiment, description will be made based on the framework of the agent program.

The agent definition section 7 and the stage 8 are for executing an agent program. The agent definition unit 7 is in the storage unit, and can define (program) various agents. As usual,
Here, the agent includes a model unit, a view unit, and a control unit (not shown) described below. The model part is a part that defines the internal state and behavior of the agent, and the view part contains information on the display of the agent. In the control unit, processing corresponding to an operation of a mouse as an input device is described. When the agent inputs data called a message from the outside, the agent performs a process according to the data, and if necessary, performs a process such as sending a message to a plurality of other agents. The agent processing unit 5 activates the agent by sending an agent activation message via the input unit 2 or the communication unit 3. Activating an agent means to create an instance in stage 8, which is its execution place, based on the definition of the agent. In addition, a class introduced in a normal object-oriented language and its inheritance of methods and data can be used to describe the agent. The user
A message can be sent to the agent via the input unit 2 (keyboard or mouse). Further, the agent can display information to the user via the display unit 1. An agent can be executed at a stage on a computer having the definition of the agent, or can be executed at a stage on any other computer. At this time, transmission of information is performed via the communication unit 3. In addition, the agent can be moved or copied from any stage to any stage.

FIG. 3 is a diagram showing a basic configuration of a site server and a client according to one embodiment of the present invention. In this configuration example, site DB 21 and site D
B management unit 22, visualizer 23, site handler 2
4, display data section 25, site agent definition section 2
6 and a client handler 2 on the client side
7, a display control unit 28, an input control unit 29, and a client agent definition unit 30.

The site DB 21 stores messages sent from various agents. Site DB management unit 22
Manages the virtual site, and site D
Processing such as generation, update, and deletion of the agent in B21 and processing for storing the status of the site DB in the storage unit are performed.

The visualizer 23 extracts display information of each agent from the site DB management unit 21 and constructs screen display data. The site handler 24 performs processing such as taking out the definition of the virtual site specified by the user from the site agent definition unit 26 and sending a message for generating an agent to be included in the virtual site to the site DB management unit 22. The display data section 25 displays the screen display data constructed by the visualizer 23, and the site agent definition section 26 stores the definition of the virtual site specified by the user.

The client handler 27 receives the user's view designation information via the input processing unit 29 and performs processing such as requesting the visualizer 23 for display area information having the view designation information.

The components 21 to 25 and the component 27
29 can basically be defined as an agent, and it is free to place these elements on the client or on the site server. For example, the client handler 27 is placed on the site server side. It is also possible. Further, when the processing load on the client side exceeds a predetermined value due to the load condition and the restrictions on the hardware on the client side (memory size, CPU speed, etc.), the configuration of the agent can be dynamically changed. is there. Messages are exchanged between the agents operating on the client side stage and the site server side stage via the network.

Hereinafter, the contents, operations, and the like of each of these elements will be described. For this reason, first, the virtual site in the present embodiment will be described.

FIG. 4 is a conceptual diagram visually representing the configuration of a virtual site used in the present embodiment. The site as a whole is equivalent to one section, in which Section 1
1, the second section 32 and the general manager's room 33. In the room, a desk 34, a file cabinet 35, a tool box 36, a form box 37, a job agent box 38, a clock 41 and the like are provided. The tool box 36 has various tools therein. Further, an agent 39 called a person corresponding to a human and a site agent 49 for performing a question process for the entire site also exist in the site. Each of these elements is actually defined in the site DB 21.

FIG. 5 is a diagram showing the configuration of each agent in the site DB according to this embodiment. The element indicated by O in the figure corresponds to one agent. The name given to O with "" is the name of the class representing the agent, and the directly written name is the name of the instance of the agent. The arc connecting the agents indicates the connection relationship between the agents. This means that the agent before the arc is registered in the list of agents that can send messages in each agent, and that there is a relationship indicated by the name of the arc between the two agents. It means holding. The "has" and "in" arcs in the figure indicate the spatial positional relationship of the agent's virtual site, and the agent on the original side of the arc spatially encompasses the agent before the arc. Indicates that "has" basically represents a statically located agent, and "in" represents a dynamically located agent.

Next, the outline of the agent shown in FIG. 5 will be described. First, the site space agent 51 is an agent having the following internal state, method, and view. Internal state: name, list of agents (mobile related agents, agents existing inside the site, other agents) Method: display, registration, deletion, change, information provision View: view 1 (2D model), view 2 (3D model), view 3 (image) The room agent is an agent having the following internal states, methods, and views.

Internal state: name, list of agents (agents existing inside the room, other agents) Method: display, registration, deletion, change, information provision View: View 1 (2D model), View 2 (3D model) ), View 3 (image) The “name” of the internal state is the name of the agent, and the agent list stores the ID of the agent having a relationship such as the above inclusion relation (this allows the agent instance to be uniquely identified). You. In the method part, operations for this agent, "display""register"
A definition (corresponding to a program) of a processing operation such as “delete” is described. In the “view” portion, a plurality of pieces of display information are described in order to support a plurality of display modes. Here, three types of two-dimensional display (plan view), three-dimensional display (three-dimensional view), and image display are prepared. Other definitions of the agent will be additionally described as needed. The virtual site as shown in FIG. 5 is managed by the site DB management unit 22.

These processes are started by a message sent from the site handler 24 or a message sent from various agents in the site DB 21. The site DB management unit 22 itself is generated by the site handler 24. (This is the same as a normal activation of an agent.) Generation of an agent is performed as follows. First, a message is sent to the site DB management unit 22 to generate the agent of the name of the class, which is the definition of the agent, the setting of the internal state, the ID of the agent inside the site to be located, and the generation location information. The site DB management unit 22 issues a request to the processing system to generate an instance of the specified class in the instance generation processing. The processing system generates an instance at the stage and notifies the site DB management unit 22 of the instance. The site DB management unit creates a message channel with the generated instance. Further, the generated instance ID and the generated location information are sent together with the locate message to the agent having the ID of the agent inside the site to be located. The agent that has received the message from the site DB management unit 22 creates a message channel with the agent that has received the message.

On the other hand, the deletion of the agent is performed as follows. First, dele the agent to be removed
When the te message is sent, the agent sends a delete-ready? message for all message channels. Each agent receiving the delete-ready? message performs necessary processing, and if necessary, performs some processing on the agent to be deleted. The agent to be deleted waits for delete-accept messages to arrive from all the corresponding agents, and when it arrives, sends an IAMDeleted message to all message channels and processes the command to delete itself. Put out This causes the agent to disappear from the stage. The updating of the agent is basically performed by sending a message to the agent in the same manner as the above processing.

Next, a method of displaying the contents of the site DB 21 on the display device on the client side will be described. The display processing is performed by the visualizer 23 and the display data unit 2 shown in FIG.
5. It is performed by the joint work of the client handler 27 and the display processing unit 28. When the display processing is performed, it may be necessary to change the display due to a point in time when the virtual site is generated (initial time) and a change in the state of the site DB 21. First, an initial operation will be described.

At the initial stage, a site handler 24 is generated according to an instruction from a user, and the site DB management unit 22 is generated by the site handler 24. The site handler 24 fetches the definition of the virtual site specified by the user from the site agent definition unit 26, sends a message for generating an agent to be sequentially placed in the virtual site to the site DB management unit 22, and sends the site D
Various agents are generated in B21. When all the initial agents have been generated, the site DB management unit 2
3 generates a visualizer 23 (agent). The visualizer 23 basically extracts display information of each agent from the site DB 21 and constructs screen display data in the display data section 25. In the present embodiment, a plurality of display data areas are created so as to allow a plurality of display forms for one site DB 21. In the following, taking the case of expressing in a two-dimensional plane as an example,
The operation of each element will be described.

Here, a part of the view information of the agent in the site DB 21 (described in the view area of the agent) is shown below.

“Site space”: 10 ｛Name = Part 1 2D view: Shape: rectangle (height = 80, width = 130, boundary = (solid line, 3, black), interior = (mesh1, gray 3D View:…｝ “Room”: 11 ｛Name = Lesson 1 2D view: Shape: Rectangle (L = 80, W = 40, Boundary = (dotted line 1, 1, black), Interior = (Transparent Relative position = '0,40' 3D view:…｝ “File cabinet”: 14 ｛Name = section 1 file cabinet 2D view: Shape: “file cabinet image” (“section 1”) Relative position = (0 , 20) 3D view:…｝ “Desk”: 15 ｛Name = Yamada / Desk 2D view: Shape = desk visual image (“Yamada”) Relative position = (30,15) 3D view ……｝ Site The “site space” that represents the whole The shape of the rectangle is specified as “rectangle.” The parameters of this rectangle include the vertical and horizontal lengths, the definition of the border of the rectangle (line type, width, and color), and the definition of the inside of the rectangle (fill pattern). , Color) are defined, and a relative position is defined in addition to the shape definition in the two-dimensional view of “room.” This information directly includes the area of this “room”. This defines the position of the “room” in the two-dimensional shape of the “site space”.
The shape of the file cabinet is designated as a file cabinet image ("1st section"). Although not shown in the figure, “1” is located at the center of the image data of the pattern in the shape of a file cabinet having 10 rows and 20 columns.
Represents image data in which the character "section" is overwritten. Similarly, the two-dimensional view in the “desk” also represents image data. FIG. 6 shows a display example of a file cabinet and a desk.

After being generated, the visualizer sends a two-dimensional view collection message Send2DViews (0) to the site space agent. (0) in the message is
This is a numerical value called a display level, and is used when processing the vertical relationship on the display. Upon receipt of this message, each agent in the site DB sends its ID, the sent display level, and its own two-dimensional view to the sender of the message, and also sends a message channel in a "has" relationship from itself. Send to the agent that is connected
A message of 2DViews (display level sent +1) is sent, and a reply message (ID, display level, two-dimensional view) is further sent to the sender of the message. Next, the same processing is performed for the agent to which the message channel is connected in the relation of "in". This allows
The visualizer 23 collects two-dimensional view information and display level information of all the agents in the site DB 21. These two-dimensional view information is held as information inside the visualizer 21. From these information, there are roughly two types of methods for displaying the status of the site on the display screen of the client. One is these two
The dimensional view information is fetched on the client side, and the actual display on the screen is performed on the client side. On the other hand, the visualizer 23 creates image data for screen display in the display data section 25, This is a method of sending information to the display processing unit 28 of the client via the client handler. In the former method, the client can select a display format according to the user's request,
Further, the display level can be set independently on the client side. On the other hand, in the latter method, since one display process is performed on the visualizer side, there is an advantage that the load on the client side is reduced. Here, first, an example of image display on the visualizer side will be described.

The two-dimensional view information collected by the visualizer has a format in which display images of a plurality of agents are overlapped. The visualizer secures a display area for holding a display image of the site in the display data section. This is the same as a normal screen display buffer, but is shared by a plurality of clients via a network. The visualizer examines the collected two-dimensional view information and writes the two-dimensional view information into the display area in order from the two-dimensional view information having a lower display level. Thus, the image information of the site DB 21 in the above example is set in the display area (shown in FIG. 7).

Next, a portion for displaying the image information set in the display area on the display screen of the client will be described. Processing related to the client side is basically performed by the client handler agent. First, the client handler agent generates the visualizer agent ID from the site handler 24 and obtains the visualizer agent ID from the site handler 24.
Reserve a message channel to 3. The client handler 27 receives the user's view designation information (in this case, a two-dimensional view) via the input processing unit 29, and displays the display area information including the view in the visualizer 23.
Request to. This request is passed to the visualizer 23,
The visualizer 23 returns the address and size information of the display area for the two-dimensional view to the client handler 27. The client handler 27 sets an area value covering the entire display area in a display area definition variable indicating which part of the display area is to be displayed, and displays the display processing unit 28
, A display request is issued. Accordingly, the display processing unit 28 displays, on the display device, a display image of a range indicated by the number of display areas defined by the client handler in the display area corresponding to the two-dimensional view in the display area.

Next, operations such as pointing from the user and selection of various commands will be described. Pointing of the user is performed by moving a mouse cursor using a pointing device such as a mouse. When the user performs an operation such as pressing a mouse button at a certain position on the display of the virtual site, the input control unit detects the coordinate position, and the client handler 27
Is notified of a user event (for example, a mouse button is pressed). Client handler 27
Calculates the coordinates of the display data section from the values of the display area definition variables and the values of the mouse coordinates, and returns the values to GetAgents
The message is sent to the visualizer 23 together with the message. The visualizer 23 obtains an agent existing at the coordinates, and sends the ID and display level information to the client handler 27. Client handler 27
Selects one of these agents (usually the agent with the highest display level) and issues a request via site handler 24 to notify that agent of the user's event. The message is sent to the intended agent. Through the above processing, the event performed by the user through the input unit is transmitted to the agent in the site DB 21.
If a plurality of agents exist at the position specified by the user, the client handler 27 may handle the conflict. For example, if a phone is placed on a desk and their display levels are the same, when the user specifies the location of the phone and presses the mouse button, the user first selects one of these two on the screen The client handler 27 generates a display for performing the operation, and the user can specify the designated agent by moving the mouse to the two display areas.

Next, the movement of the agent on the display will be described. The display position of the agent is site DB2
1 is held as internal information of the agent.
The user can communicate with the agent according to the method described above, and can use this to send a position movement command to the agent. When the position information is updated, each agent issues a display change request to the visualizer 23. The visualizer 23 performs the process of moving the agent that has requested the movement, and calculates the display area of the agent that requires a new screen overwrite by the movement, and performs the rewriting process on the display portion of the display buffer. Do it.

Next, the person agent will be described. A person agent exists in a virtual site and has a characteristic of having a mapping relationship with a real person.

The person agent is one agent and has the following information as internal information.

At least one home site ID and home agent ID Current site ID, location information at the current site Various personal information (name, affiliation, address, telephone number, user identifier) View information Status information Site History information The home site is the ID of a person's initial value site, and can have a plurality of site IDs. The home agent ID is an agent owned by the person, and usually corresponds to the desk agent. In the present embodiment, this home agent is called a “personal desk”. The personal desk serves as a work environment for a person, and also has a role of maintaining a method of accessing the person. In the case of the present embodiment, the method of accessing the person is represented by the current site ID and the ID of the person agent itself. The current site ID is
The virtual site I visited at that time
D. The position information at the current site is information indicating the position of the agent in the site described above.
The various types of personal information are various types of information corresponding to the person, such as name, affiliation, address, and telephone number. View information is
It is the information of the view of the agent. The status information is various kinds of information (described later) indicating the state of the person. Also, the personal desk has a person ID, a person's current site ID, person status information, view information,
Links to various work agents are stored.

Next, a procedure when a user starts a computer will be described with respect to a portion related to a virtual site and a person. If you are not using a calculator,
The person exists in the site DB on the home side, and its status is not-login. When the user performs a normal computer startup procedure, various general processes such as input, display, and communication are generated. After this, Site Handler 2
4, a request to generate the client handler 27 is sent together with the user identifier. The site handler 24 searches for the person agent in the site DB 21 using the user identifier as a key, and notifies the client ID of the agent ID. Further, the client server performs a necessary process to perform a screen display process of the status of the site.
Further, the client handler 27 sends a message to the person agent via the site DB 22 to change the status to "login". Upon receiving this message, the person agent sends a display update message to the visualizer to change its view from the not-login display state to the login state. As a result, the display state of the person agent on the display unit changes, and it can be seen from other users that the user corresponding to this person has begun using the computer.

Next, the movement of a person between virtual sites will be described. When moving from a virtual site A to another virtual site B, the ID of the destination virtual site B is used as a key. First, the operation of these two virtual sites will be described. The site A of the movement originates from a certain person's site movement request (mobile site ID).
), A person inquiry is made to the site server of this ID. Person inquiry refers to requesting person information necessary for a check for movement from a server at site B, obtaining the requested information from a person, sending the information to site server B, and determining whether movement is possible. Is the process of inquiring about Upon successful completion of the person inquiry,
The source server transfers the person agent to the destination server and deletes the person agent from its own site. The destination site server performs a process of registering the sent person agent in its own site DB.

For moving, the virtual site ID
In the following, several cases will be described as to how the user obtains this ID.

First, information is provided from a current site where a person is located. As shown in FIG. 8, a company organization and the like are defined in a state where a plurality of virtual sites are connected in a network. Arcs in the network
A relationship name between two virtual sites can be defined.

For example, in FIG. 8, “company” and “guard”, “acceptance”, “part 1”, “part 2”, and “library” are in an is-section-of relationship which is an organizational inclusion relationship. There is a connection.
In addition, there is a connection called “is-neigber-of” between “Part 1” and “Part 2”, “Library” and “Reception”. For these, the connection relationship name, the partner site name, and the ID are stored as other site link information in the model part of each virtual site agent. This information is called site link information, and sites in a connected relationship are called adjacent sites. With this site link information, the person can extract the information of the virtual site connected to the current site.

The second is virtual site information stored by the person himself. According to an instruction from the user, the name of the virtual site and its ID are stored in the person, and this information (called personal site information) can be extracted as required. By taking in the information inside the person who visits the site on a daily basis, operations such as tracing a plurality of virtual sites can be simplified. Also, not the virtual site, but the name of the person and the ID of the person's home agent
By storing (called personal person information), it is possible to search for a site where a certain individual is located. Further, the person holds information called site history information. This is the virtual site name and I
D is a group, and holds a history of sites (excluding home sites) visited by the person.

Third, site identification by telephone number. This means that the user enters a predetermined phone number,
In this system, a site ID is searched using the telephone number as a key. For this purpose, a telephone number agent having a function of extracting a site ID from a telephone number is prepared. The telephone number agent has a table of virtual site IDs corresponding to telephone numbers, and returns IDs in response to various agent requests. If the telephone number of the inquiry is up to the code of the area, the ID of the virtual site (referred to as area site) representing the area is returned. The area site performs tasks such as inquiring about the area and mediating movement to the virtual site. If the individual telephone number for which the inquiry was made does not exist, the telephone number agent returns the absence of the number and the ID of the area site.

The fourth is the identification of the site by the zip code / address. This means that the user inputs a predetermined postal code / address, and uses the postal code / address as a key for the site ID.
Is a method of detecting A postal code / address agent similar to the telephone number agent is prepared so that the site ID can be searched in response to an inquiry. Again,
If the site corresponding to the specified address does not exist, the fact and the ID of the higher-level area site are returned.

The fifth is the identification of a site by an inquiry in a natural language. This is realized by preparing an agent (site identification agent) for identifying a site ID by inputting a target place in a natural language and analyzing the natural language. The site identification agent has a function of detecting an appropriate site ID from a character string input by a user. The sixth is a method of moving to a partner site in communication between persons. This is done by inquiring the person's current site
It is possible to get a site ID.

The process of moving the site of the person according to the above five items will be described below. First, the user selects a command from the input unit and selects a “move” command.
This request is passed from the person agent to the mobile processing agent, and the mobile processing agent performs processing in accordance with the flows shown in FIGS. First, the mobile processing agent responds to a person who has requested a move, personal site information possessed by the person,
Retrieve personal person information, site history information, and current site information.

Next, a destination setting window as shown in FIG. 11 is displayed on the user's screen, and the “destination site ID” holding the destination ID of the destination is set to NULL.
Set to. The name of the current site is displayed at the “current” position on the display screen, and the name of the destination is displayed at the “destination” position. The "site list" displays the name of the site obtained from the personal site information and the name of the adjacent site obtained from the site link information of the current site. At this time, if the person's current site is any of the person's personal sites, the personal site is preferentially displayed, such as placing the personal site at the top, and otherwise, the adjacent site is preferentially displayed. However,
If the current site is not the home site, the home site is displayed with priority. In the position of the personal name list, the personal name obtained from the personal person information is displayed. In the “site history”, a site name is extracted from the site history information and displayed. “Destination”, “〒”, “TE”
“L” is a field for inputting a key for searching for a destination, and “Status display” is a field for displaying various messages. The mobile processing agent then waits for a user input and performs a corresponding process.

In FIG. 9S3, if it is determined that the user has selected an element of the site list by means such as a mouse, a site list element process is performed. In the site list processing (S4), the designated site is recognized as shown in FIG. 12 (S21), this site ID is taken out, and this ID is set in the "destination site" (S22). FIG. 9S
In 5, if it is determined that the user has selected an element of the personal name list by means of a mouse or the like, personal name list element processing is performed (S 6). In the personal name list process, as shown in FIG. 13, the home agent of the person corresponding to the specified personal name is taken out (S31), and the agent is requested for the state information of the other person (S3).
2). The status information is the current site information, home site information, status (not-login, talkin
g, busy, acceptable, free). not-login is a situation where the other party is not accessible from the computer.
Bu that the other party is communicating with someone
sy says that he can not communicate because he is busy
Acceptable indicates a situation where direct communication should be avoided if possible, and free indicates that direct communication is currently possible. The mobile processing agent extracts the home site ID of the partner person,
This ID is set in the “destination site” (S33, S3
4).

At S7 in FIG. 9, if it is determined that the user has selected an element of the site history by means of a mouse or the like, history site element processing is performed (S8). In historical site processing, the same processing as site list element processing is performed.
Extract this site ID and enter this I
Set D. In FIG. 9S9, if the user designates the field of "@" or "address" and it is determined whether or not the input is made by means of a keyboard or the like, postal code / address processing is performed (S10). 14 and 15
Shows a flow of postal code / address processing.

First, in FIG. 14S41, the end of the user input is detected. This is due to the detection of the enter key on the keyboard and the execution key on the display screen. In S42, it is determined whether or not the user's input was in the "@" field.
In the case of inputting the "@" field, a search is made for a site ID using the postal code as a key in S43. This is done by receiving a zip code and sending a message to a dedicated agent that sends back the corresponding site ID and address information. This agent holds a set of postal code, site ID, and address information, and searches this table. If the search succeeds, the site ID and address information are sent, and if the search fails, a failure message is sent. In S44, the success or failure of the search is determined. If the search is unsuccessful, the fact that the search process has failed is displayed in a message field in S45 in FIG. If the search is successful, the obtained address information is displayed in the address area of the display screen in S46 in FIG. 14, the obtained site ID is set in the destination site in S47, and the process ends. Although not described here, every time the site ID is set to the destination site, the destination display agent searches for the name of this site ID (inquires of the site), and the name is displayed in the destination field of the display screen. Is displayed.

If the field input by the user is not the "@" field, it means that the input is an address field. In S48, a search process is performed using the address as a key. This is performed by sending a message to a dedicated agent that receives the address character string and the position information (optional) of the current site and returns the corresponding site ID and address character string. This agent divides the sent address string into predetermined units using an address dictionary, and
The identification of elements such as city and town is performed, and the obtained element is used as a key to access a database holding the relationship between these address keys and sites. For example, if a city name is used as a key, a plurality of data is searched. In this case, if the position information (geographical coordinates) of the current site is given, the site information (ID, address) obtained in the order of elements closer to the coordinates is arranged and sent back.

If the search fails, a failure message is sent back. In S49, the success or failure of the search process is determined. If the search is unsuccessful, the process of S45 is performed. If the search is successful, it is determined in S50 of FIG. 14 whether or not there are a plurality of search results. In FIG. 15S51, the site ID obtained for the destination site is set, and the process ends. In the case of a plurality of sites, in S52, the sites are geographically closest (that is,
The address information is displayed in the site list on the display screen in the order in which the search results are sent, a message indicating that a plurality of candidates exist in the message area of the display screen, and a message is displayed to select a candidate. . Next, in S53, it is determined whether or not the user input is a selection from the site list (for example, a selection with a mouse and a mouse button). If the user input is to select a site list element,
The ID of the site selected in 54 is set as the destination site. Further, the state of the site list is returned to the original display state, and the process ends. If the determination in S53 is no, the processing from S54 is performed.

Returning to the description of the movement processing flow in FIGS. 9 and 10, again. In FIG. 10S11, it is determined whether or not the user input is an input to the telephone number field. In that case, S
At 12, telephone number processing is performed. The telephone number process performs the same process as the postal code process using a search agent using a telephone number as a key. If the search is successful, the searched site ID is set in the destination site. Then S1
At 3, it is determined whether or not an input has been made in the destination field. In the case of input to the destination field, S14
The destination input process is performed in. In S14, the processing shown in FIG. 16 is performed. First, in FIG. 16S61, a language analysis is performed on the character string set in the destination field. This includes site name, part of speech, site type,
Normal morphological analysis using a dictionary containing site IDs
This is performed by a syntax analysis process. Thus, for example, for the input “Tokyo Station”, it is a proper noun representing the site, and the site ID is obtained from the information in the dictionary.
Further, for the input "library", information indicating that the input is a noun representing a site is obtained. Also, for the input "Library in Setagaya-ku", information is obtained that the area site "Setagaya-ku" and the site "Library" are connected by the case particle "NO".

Next, in S62, a search process of the actual destination site ID is performed based on the analysis information obtained in S61. For the input "Tokyo Station", the site ID has already been obtained because the site ID has been searched from the dictionary information. In the case of "library", it is known from language processing that the noun is a site noun, and a search for a site ID is performed based on this information. This is done by first searching for a predetermined number of sites, such as the site adjacent to the current site and its connected sites, and if there is a site corresponding to "library", move this site to the destination Treat as a site. If the relevant site does not exist as a result of the search, a request is made to the area site including the current site to search for a "library" site inside the area. ,
Treat this site as a destination site. If the input is "Library in Setagaya-ku", a request is made to the "Setagaya-ku" area site to search the internal "library" site according to the definition in the input pattern table (not shown). As a result, if the site exists,
Treat this site as a destination site. By this processing, the site ID is obtained from the character string input to the destination field.
Can be obtained. If the search for the site ID has failed, it is detected in S63, an error is displayed on the display screen in S64, and the process is terminated. If a plurality of site IDs are obtained, the candidates are presented to the user in S66 in order of proximity to the current site, and the user is allowed to select one of them. In S67, a site ID is set for the destination site, and the process ends.

Returning to the description of the movement processing flow of FIGS. 9 and 10 again. In S15 of FIG. 10, it is detected whether or not the cancel button has been selected, and in that case, the processing relating to the movement ends. In S16, it is detected whether the execution button has been selected, and if not, the processing from S3 is repeated.
If the execute button is selected, it is determined in S17 whether or not the destination site ID has already been set. If the site ID has already been set, the transfer process for the site ID is executed. If the destination site has not been set, an error message to that effect is displayed on the screen in S19, and the processing from S3 is repeated. As described above, the user can move from the current site to another site.

In the above embodiment, the movement of the site by the address has been described. In such a case, a virtual site that processes on behalf of a certain geographical area is effective.
This type of site is called an area site. Area sites are created corresponding to regional divisions such as "country", "prefecture", "city", and "town", and are connected and configured to form a hierarchy of area sites. Setagaya-ku as an example
Take the area site (called Setagaya-ku area site). The upper part of the Setagaya-ku area site is connected to the "Tokyo" area site because of the is-part-of relationship, and the lower part of the Setagaya-ku area site is the area of each town in Setagaya-ku (Nakamachi, Seta, etc.). Site is connected. FIG. 17 shows an example of a display screen when moving to the Setagaya area site. On the screen, a map of Setagaya-ku is displayed, and the map is displayed separately in an area site below Setagaya-ku. Further, the Setagaya-ku area site has a site agent for answering questions to persons within the site, and is displayed on the screen. In addition, buttons for starting search software for ward administrative services and ward information are displayed. From this "Setagaya-ku" area site, the user can access the lower "Nakamachi area"
It can be configured to receive services such as navigating to, and inquiring about the site in Setagaya and displaying the results.

Next, various functions relating to the connection processing between persons (agents) will be described. The connection process is a process of setting a communication path for communication between the persons. Here, a person A corresponding to a certain user A and a person B corresponding to a user B will be described as examples. FIG. 18 is an example of an access screen for Person B. In this example, the mouse cursor is moved over an image of Person B (Yamada) in the site DB, and an access menu for Person B is displayed by pressing a mouse button. Here, the name of the person B, the current location, and the state of the other party are displayed, and an "access" selection button for controlling whether or not to actually perform the access is displayed. As a method of accessing a person, as shown in this figure, in addition to a method of directly specifying a person in a site, it is also possible to access a person from a person's home agent. In this case, the person B stored in the home agent
Of the current site, and establishes a connection with the person B by message communication to the site.

When a request for access to the person B by the user A (such as pressing a mouse button) is detected, the person A
Transmits a message of predetermined accessor information together with an access request message to person B. Upon receiving this message, Person B presents to User B that access from Person A has occurred, if possible.

Further, the person B performs a request arbitration process for requesting the information and the access permission of the person B to be presented to the partner according to the information of the partner person and his / her situation.
In this request arbitration process, a determination is made by inquiring of the user B according to the situation. This determination includes a business relationship (boss,
It determines the urgency of the other party, such as colleagues, subordinates, customers, customers, personal relationships (friends, family, etc.), and other unknown persons, and creates a message to be sent to Person A accordingly.

The person A presents the user information and the access permission / inhibition information sent from the person B to the user A. Here, when the user corresponding to the person A selects the access to the person B, the access method (the video channel,
Request to select a voice channel, a character transmission channel, etc.). Person B selects an access method, reserves the channel, and sends the access method and actual channel information to Person A. Person A performs channel connection processing based on this information,
A predetermined communication channel is established between Person and Person B. When this communication channel is determined, a processing program for communication display is started. FIG. 19 shows the above processing.

FIG. 20 is an example of a screen on the user side corresponding to person A for communicating between person A and person B, and is displayed by the communication display program. In the figure, 61, 62, and 63 indicate the name, current site, and status of the other party, respectively. Numeral 64 denotes an area for displaying the information of the other party, and a real-time moving image display, a still image prepared in advance by the user corresponding to the person B, and the like are displayed by selecting the communication means. In this area, a display indicating whether the correspondence of the person B is a real user or not is displayed. In this screen example, the display of "agent" indicates that the corresponding partner is not a real person.
Reference numeral 65 denotes an area for performing a dialogue with the other party using characters, and is divided into an area for displaying a call on the other party and an area for displaying a call on the user's side. Further, when a voice call is possible, this area can be used to display the conversation of voice conversation in characters. 66
Is an area for transferring arbitrary data to person B. Reference numeral 67 denotes an area for receiving arbitrary data from the personal computer B. Reference numeral 69 denotes an area for sending information to the partner's home agent even when direct communication with the partner is not possible. Reference numeral 70 denotes an area for instructing recording of a communication status. Reference numeral 68 denotes an area for synchronizing the movement of the opponent with the movement of the opponent or vice versa.

Next, various operations on the person and the operation and configuration of the person will be described. FIG. 21 is a block diagram showing a response process of a person. Reference numeral 71 denotes a recording unit, which has a function of recording information on an input from the partner person and an output to the partner person in a storage device (not shown). Reference numeral 72 denotes a correspondence switching unit that switches between outputting an input from the other party to the automatic processing unit 73 and outputting to the real user interface unit 74. Also,
On the contrary, switching is performed between output from the automatic processing unit and output from the real user interface. The automatic processing unit 73 analyzes requests and the like from the other party, and performs output to the other party and exchanges information with the real user via the real user interface unit according to the result.
The real user interface unit has a function of performing input / output processing for an actual user via an input / output device.

In the correspondence switching unit 72, two output switchings of K1 and k2 can be set, and the initial setting is k1
Is output to the automatic processing unit. Here, the example of the person A and the person B will be described again, and the operation of the person will be described. When there is an access request from the partner person, this is sent to the automatic processing unit 73 through the recording unit 71 and the correspondence switching unit 72. The automatic processing unit 73 performs the connection processing. In a state where the connection processing has been completed, the display screen is displayed for the user on the person A side,
The user on the person B side is notified from the automatic processing unit 73 through the real user interface 74 that the access from the person A has occurred. In this state, if the user on the person B side wants to directly communicate with the user on the person A side, the correspondence switching unit 7 via the real user interface 74.
By issuing a switching request to the user 2, the input from the person A is output to k2, and is directly sent to the user interface, and is displayed on the screen of the person B in the same manner as the user of the person A. The display screen is displayed. When the user switching request of the person B is not issued, the input from the person A is sent to the automatic processing unit as it is and processed.

FIG. 22 is a block diagram of the automatic processing unit shown in FIG. The request analysis unit 81 analyzes information transmitted from the correspondence switching unit 72 and extracts a problem to be solved. The problem solving unit 82 solves the problem extracted by the request analyzing unit 81. Knowledge search unit 83
Includes the knowledge information possessed by the person, and performs a search process on the contents in response to a request from the problem solving unit 82. Response generator 84
Has a function of determining whether the information obtained by the problem solving unit 82 can be responded to the requester. The response generation unit 85 has a function of converting the information sent from the response determination processing unit 84 into a predetermined format and sending the converted information to the correspondence switching unit 12.

The operation of the automatic processing section 73 in response to various inputs from the person A will be described below. FIG. 23 to FIG. 25 are flowcharts showing the processing of the request analysis unit 81. First, when an input from the person A is sent from the correspondence switching unit, it is determined whether or not the content is a voice input or a natural language sentence input (S71). In the case of YES here, the flow proceeds to step S72 (FIG. 24), and the fact that the processing is being performed is displayed on the display screen of the user corresponding to person A. Next, in step S73, speech recognition processing and natural language recognition processing are performed. This is performed by using a speech recognition processing and a natural language analysis processing module. The recognition result is classified into one of "greeting", "data request", "data notification", "direct dialog request", "end request", and "recognition failure". For example, if the input is "Hello" is, by the dictionary information of natural language processing module, "greeting: type1" is analyzed as of. Where type
1 is a greeting indicating a queue for starting conversation. The sentence "Where is ~?" Is analyzed as a problem of inquiring about the location of "~" as in "Data request: location (~)". Also, input such as "I want to do it directly"
"Goodbye" is analyzed as a termination request. Steps S74 to S83 and step S84 show processing for sending these analysis results to the problem solving unit. When the analysis process has failed, the problem solving unit 82 is notified together with the content determined to be analysis failure, such as “analysis failure: voice unknown”. If the input from person A is not a voice input or a natural language sentence input, step S8
In steps S86, S88, it is determined whether the request is a data notification (transfer of a document or the like through f on the display screen), an interlocking request, or an end request.
At 82, the contents are notified to the problem analysis unit, and the process is terminated.
If neither of the above is the case, a notification is sent to the problem solving unit 82 as "other request", and the process is terminated.

FIGS. 26 and 27 are flowcharts showing the processing of the problem solving unit 82. When the request analysis unit 81 determines that the input from the partner person is a greeting, it determines in step S91 and selects a predetermined response to the determination. In this case, a table (not shown) stored in advance in the knowledge search unit 83 is searched, and a corresponding response is obtained. The information is notified to the response determination processing unit 84 as an answer. If the input is analyzed as a data request, a determination is made in step S94, and a knowledge search process is performed in step S95. In the knowledge search process, the knowledge search unit 83 searches for an answer using a question as a key. For example, location
In response to the question (Where is the business trip application form?), The description of “location (business trip application form) = cabinet 1” is stored in the database of the knowledge search unit 83. The answer is "cabinet 1". The same question is analyzed by the request analysis unit 81 at the level of the intention, such as “give (business trip application form)”, and the knowledge search unit 83 is given an analysis such as “give (X) = search program (X)”. If there is a description, launch the search program with the travel application form as an argument and retrieve it,
That can be the answer. With this search process,
If the answer is not obtained, the automatic processing for the person request will not be successful, and the situation is presented to the real user in a predetermined form in step S97. The success or failure of the search and the answer are notified to the response determination processing unit 84 in step S98. Those determined by the request analysis unit 81 as a data notification request, a direct interaction request, an end request, an interlocking request, and other requests are recognized in step S99, and are directly notified to the response determination processing unit 84 in step S93.
When the request analysis unit 81 fails in the request analysis, it is detected in step S100 in FIG. 27, and the process proceeds to step S101. In step S101, it is determined whether a predetermined condition such as the number of consecutive failures in the analysis or the ratio thereof is satisfied. When this condition is satisfied, it is determined that the interactive processing is defective, and the information is notified to the response determination processing unit 84 to terminate the processing. In step S103, it is determined whether or not a re-input request should be made to the other party according to the cause of the analysis failure. Causes of parsing failure include low confidence in the speech recognizer, words that are not known by natural language analysis,
Or there are various factors such as multiple interpretations of the sentence. In step S103, it is determined based on a table in which what kind of response should be taken for each of these failure factors (or failure to identify a unique interpretation). For example, when there are three probable candidates in the analysis result, it is possible to perform a process such as presenting them and having the other user select them. Step S10
If it is determined in step 3 that re-input is to be performed, re-input processing is notified to the response determination processing section 84 with a measure to be taken. If it is determined that re-input is not to be performed, it is determined in step S105 that the situation has become such that communication is difficult, and the response determination processing unit 84 is notified with this information. Step S
If it is determined in step 100 that the analysis has not failed, the response determination processing unit 84 is notified of information indicating that the problem solving process has failed in step S106, and the process ends.

FIGS. 28 to 30 are flowcharts showing the processing contents of the response determination processing unit 84. The response determination processing unit 84 determines what kind of response is to be returned according to the result of the problem solution notified from the problem solution processing unit 82. If necessary, via the real user interface unit 74,
Performs input and output for actual users. Step S111
Then, it is determined whether or not the notification from the problem solving unit 82 is a problem solving failure, a dialogue failure, a direct dialogue request, an interlocking request, or another request. In these cases, it is desirable to ask for the judgment of the real user. In step S112, it is checked whether or not the real user is available. When the real user is not accessible from the computer (when the computer is not used or away from the computer), and when the user is accessible from the computer but is performing a high priority job, the step In S113, the information that the real user cannot respond is notified to the response generation unit. If the response is accessible, the notification status from the problem solving unit 82 is displayed on the display screen of the real user in a predetermined form. Then, the process ends. If the real user is available, in step S114, the notification status from the problem solving unit is presented via the real user interface processing unit (74 in FIGS. 21 and 74), and the instruction is received. In this case, if the real user refuses the intervention, in step S116,
The response generation unit 85 is notified with the information that the real user has refused the intervention, and the process ends. When the real user instructs the automatic processing unit 73 to switch the corresponding process to the real user, a switching request is issued to the corresponding switching unit 74 in step S118. As a result, the correspondence switching unit 72 switches input / output with the partner person from the automatic processing unit 73 to the real user interface unit 74. As a result, the screen display on the real user side is connected to the partner person, and actual communication between the two persons becomes possible.
In the case where the automatic processing unit 73 is responding, this exchange situation is displayed in a reduced scale on the screen of the user on the person B side in the above example, for example, but this switching process is performed. Then, the screen changes to a screen of a predetermined size. If the designation of the real user is not the correspondence switching process in step S117, the answer to the problem solving process is input from the real user in step S119, and the process proceeds to step S120.
Then, the information is notified to the response generation unit 85 with the information, and the process is terminated. In step S121, it is determined whether or not the notification from the problem solving unit has succeeded in solving the problem.

If the problem is solved successfully, step S122
Performs a security check on the answer data obtained in. This depends on the security level of the other person, the access qualification, the security level of the data,
This is done from the access credentials. If the security check does not pass, an inquiry is made to the real user in step S123, and confirmation regarding the provision of data occurs. If the offer is rejected or the inquiry to the real user cannot be made, step S1
At 25, it is notified that it is difficult to provide, and the process ends. If the security check is passed, and if the real user is approved in step S124,
In step S126, the answer data is notified to the response generating unit 85, and the process ends. In step S127, it is determined whether the notification from the problem solving unit 82 is a data notification.

If it is a data notification, step S
At 128, information indicating that the data has been received is presented to the real user (if possible), the received data is stored at step S129, and at step S130, the fact that the data has been received is notified, and the process is terminated. Step S13
In step 1, it is determined whether the notification from the problem solving unit 82 is an end request. If the request is a termination request, the termination is notified to the response generation unit 85 in step S132, and termination processing is performed in step S133. In the termination processing, the termination of the response processing in the response generation unit 85 is confirmed, and further, it is confirmed that there is no input from the partner person within a predetermined time, and the release of the secured communication line, the deletion of the actual user screen, and the like. Do.

In step S134, it is determined whether or not the notification from the problem solving section 82 has been re-input. In the case of re-input, a message to be presented to the partner person is selected based on the cause of failure by the problem solving unit 82 and is notified to the response determination processing unit 84. This is sent to the response generating unit 85 in step S135. Notify and end the process. Step S
In 136, it is determined whether or not the notification from the problem solving unit 82 has difficulty communicating, and if so, step S
In step 139, a response difficult message is selected, and step S11 is performed.
Proceed to 2. In a step S137, a call difficult message is selected.
And the process ends.

The response generation unit 85 includes a response determination processing unit 84
A response message to the partner person is created based on the information notified from the server, and is sent to the partner person via the correspondence switching unit 72. Table 1 shows an example of an output sentence generated by the response generator for each piece of notification information.

[0072]

[Table 1] These output sentences are converted by means of communication with the partner person. For example, when the voice communication channel is open, information converted to voice data according to the voice parameters of the real user is also transmitted. Also, in the case where data to be sent corresponds to data having a certain unit (a so-called object) such as a document, information that can identify what the transmitted data is so that it is not a conversational area of the other user but an area of the receiving port. Is added. Thus, the file can be controlled to be displayed on the display screen of the other user in the form of one icon. With the above processing, when the person A accesses the person B, the user on the person A side is as if the user of the person B is responding, even if the user corresponding to the person B does not directly correspond. The communication can be performed, and the process can be changed to the actual communication between the users when a predetermined trouble occurs in the communication.

With the above-described configuration, simple question answering processing can be performed by a person's automatic answering function without being performed by an actual user. Further, when a failure occurs in the automatic response, it is possible to switch with the actual user, so that the dialogue can be continued without any delay.

[0074]

As described above, according to the present invention, various virtual sites are constructed on a computer on a network, communication between users accessing the virtual sites becomes possible, and movement of the virtual sites is achieved. Thus, access to information in various places can be naturally and easily realized. In addition, by inserting an agent with an automatic response function called a person for user access, functions such as simple exchanges and routine exchanges are automated, and real users can respond only when necessary. it can. Also, by providing this interface, you can move around the world inside the computer and talk with other people, etc., so that when the actual user performs a certain job or work, the place where the user does not know in advance Find the data at
It is very easy to refer to and use various computer services that are not frequently used.

[Brief description of the drawings]

FIG. 1 is a diagram showing a computer network configuration according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a computer according to an embodiment of the present invention.

FIG. 3 is a diagram showing a basic configuration of a site server and a client.

FIG. 4 is a conceptual diagram visually representing the configuration of a virtual site.

FIG. 5 is a diagram showing a configuration of each agent in a site DB.

FIG. 6 is a diagram showing a display example of a file cabinet and a desk.

FIG. 7 is a diagram showing a display image example of a virtual site. .

FIG. 8 is a diagram illustrating a connection configuration example between virtual sites.

FIG. 9 is a first flowchart illustrating a person moving process;
FIG.

FIG. 10 is a second partial diagram of the flowchart showing the person moving process.

FIG. 11 is a diagram showing a display example for specifying a destination.

FIG. 12 is a flowchart showing a site list process.

FIG. 13 is a flowchart illustrating a personal name list process.

FIG. 14 is a first partial diagram of a flowchart showing postal code / address processing.

FIG. 15 is a second partial diagram of a flowchart showing postal code / address processing.

FIG. 16 is a flowchart showing a mobile site designation process based on language input.

FIG. 17 is a diagram showing an example of an area site screen.

FIG. 18 is a diagram illustrating an example of an access screen for a person.

FIG. 19 is a diagram for explaining a process of establishing a communication channel between persons.

FIG. 20 is a diagram illustrating a screen display example for interperson communication.

FIG. 21 is a block diagram related to a response processing operation of a person.

FIG. 22 is a block diagram of an automatic processing unit.

FIG. 23 is a first branch diagram of the flowchart showing the processing of the request analysis unit.

FIG. 24 is a second branch diagram of the flowchart showing the processing of the request analysis unit.

FIG. 25 is a third subsection of the flowchart showing the processing of the request analysis unit.

FIG. 26 is a first branch diagram of the flowchart showing the processing of the problem solving unit.

FIG. 27 is a second branch diagram of the flowchart showing the processing of the problem solving unit.

FIG. 28 is a first flowchart of the flowchart showing the processing of the response determination processing unit.

FIG. 29 is a second partial diagram of the flowchart showing the processing of the response determination processing section.

FIG. 30 is a third partial diagram of the flowchart showing the processing of the response determination processing section.

[Explanation of symbols]

 5 Agent processing unit 7 Agent definition unit 8 Stage 21 Site DB 22 Site DB management unit 23 Visualizer 24 Site handler 25 Display data unit 26 Site agent definition unit 27 Client handler 28 Display control unit 29 Input control unit 30 Client agent definition unit 71 Recording unit 72 Correspondence switching unit 73 Automatic processing unit 74 Real user interface unit 81 Request analysis unit 82 Problem solving unit 83 Knowledge search unit 84 Response determination processing unit 85 Response generation unit

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Etsushi Ito 1 Komukai Toshiba-cho, Sachi-ku, Kawasaki-shi, Kanagawa Prefecture Toshiba Research Institute, Inc. (56) References JP-A-5-113941 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) G06F 13/00 G06F 3/14-3/153 G06F 15/00 G06T 15/00 H04N 7/15

Claims (3)

(57) [Claims] 1. A plurality of meters connected by a network
Definition of the state and processing behavior of each agent in a human interface system for using a distributed system with a computer
Holds or manages information and is configured by the agent
Site to be configured on each computer based on the definition information
And site construction unit that built, the access from the user with respect to the sites built
The processing operation of the agent is defined in the
Access control means for controlling based on the information and request arbitration processing according to the state and processing operation of the agent
And a communication channel setting means for setting a communication channel between agents . 2. The access management means recognizes data obtained from a user and analyzes the request.,
From the knowledge information of the agentIn this analysis result
Automatic response function that creates and outputs responses to responsesWhen, Response processing by this automatic response function and dialogue by real users
A switching function for switching between processing and  2. The human body according to claim 1, wherein
Interface system. 3. A plurality of meters connected by a network.
In a control method of a human interface system for using a distributed system having a computer, a definition of a state and a processing operation of each of a plurality of agents.
Holds or manages information and is configured by the agent
Site to be configured on each computer based on the definition information
Within the access from the user to the site built and built
The processing operation of the agent is defined in the
Controlled on the basis of the distribution, request arbitration processing according to the state and the processing operation of the agent
A method for controlling a human interface system, comprising: setting a communication channel between agents ;