WO2009041748A1

WO2009041748A1 - Apparatus and method for playing by inferring and predicting in ubiquitous space

Info

Publication number: WO2009041748A1
Application number: PCT/KR2007/005086
Authority: WO
Inventors: Jai Hoon Kim; Dong Wook Lee; We Duke Cho
Original assignee: Ajou University Industry Cooperation Foundation
Priority date: 2007-09-28
Filing date: 2007-10-17
Publication date: 2009-04-02
Also published as: KR20090032479A

Abstract

An apparatus and method for playing an instance inferred and predicted in a ubiquitous environment is provided. An inferred/predicted instance play apparatus of the present invention includes a situation detector, an information storage, and a temporalspatial information player. The situation detector detects physical and electrical situations in a ubiquitous space. The information storage stores the detected physical and electrical situations as temporalspatial information. The temporalspatial information player obtains requested temporalspatial information in response to an instance play request and replays the temporalspatial information. The temporalspatial information player replays, if the requested temporalspatial information is retrieved in the information storage, the retrieved temporalspatial information and, otherwise, infers or predicts the requested temporalspatial information using the temporalspatial information stored in the information storage. The instance play apparatus of the present invention can infers or predicts a contextual instance at a past or future time requested by a user and plays the inferred/predicted contextual instance, whereby the user feels like existing in that spatialtemporal environment of the instance.

Description

[DESCRIPTION] [INVENTION TITLE]

APPARATUS AND METHOD FOR PLAYING BY INFERRING AND PREDICTING

IN UBIQUITOUS SPACE

[TECHNICAL FIELD]

The present invention relates a ubiquitous computing and, in particular, to an apparatus and method for playing an instance inferred and predicted in a ubiquitous environment.

[BACKGROUND ART]

With the advance of mobile computing and wireless communication technologies, ubiquitous computing era is emerging. A ubiquitous space, called spatial revolution, has been developed to a ubiquitous intelligent space defined as a human living environment of physical space embedded with computing and networking capabilities. In the ubiquitous intelligent space, all the objects are intelligently installed so as to provide various services on the basis of context recognition.

Similar to the ubiquitous intelligent space, virtual reality and augmented reality are referred. In the virtual reality, a physical space is virtually created in the form of cyber space such that a user feels like moving around in the virtual space. The augmented reality is an improved virtual reality environment created using the space connection concept of the ubiquitous space. That is, the elements of the physical space (or virtual reality) are embedded with intelligence and cooperatively process various information collected by themselves such that the virtual reality is created.

Recently, various researches on interaction and integration of the physical space and cyber space are conducted and virtual spacebased applications are widely applied for providing various services. Such researches are developing a new paradigm of information society through interspace movement and space connection. By collapsing the boundaries between the spaces and freely moving in the virtual space on the physical space, the ubiquitous space can provide various services anytime, anywhere, and in any context.

However, most of the researches are focused on how to overcome the spatial restriction but temporal restriction. That is, the abovementioned "anytime" just means any temporal instant of presence but not the movement to specific time.

In order to overcome the temporal restriction, various ideas have been proposed. However, such ideas are limited to recoding and replaying still and motion pictures. For example, TV recoding function only allows watching a program at an available time and the Aware Home project of Georgia Institute of Technology (GIT) only allows storing family photos which can reminds the family of their past in the future. That is, such proposals are limited to just replaying the data stored in specific media formats since the time travel is impossible in the physical space.

There have been researches for traveling back in time in computing (cyber space). For example, a computer is configured to store logs for debugging purpose or periodically save status of tasks for recovering, when a hardware or software error occurs, the error by rolling back to a lastly saved instant or to an instant right before, if a transaction is aborted, the transaction. A method for canceling a delivery of an email, which is transmitted already, in consideration of the transmitted email status, i.e. whether the email is read by the recipient or suspended at an email server.

That is, most of the conventional techniques for overcoming temporal restriction are limited to storing data and replaying the data later; storing status information of a task and executing, if an error occurs, the task with the stored status information again; or canceling delivery of transmitted data with reference to the status of the transmitted data. However, such conventional methods are can be only used in limited application fields and difficult to give the user a temporal traveling effect in physical space.

[DETAILED DESCRIPTION OF THE INVENTION]

[TECHNICAL PROBLEM]

The present invention has been made in an effort to solve the above problems, and it is an object of the present invention to provide an apparatus that is capable of inferring or predicting an instance at a requested time and playing the instance in a ubiquitous environment. It is another object of the present invention to provide a method that is capable of inferring or predicting an instance at a requested time and playing the instance in a ubiquitous environment

[TECHNICAL SOLUTION]

In accordance with an aspect of the present invention, the above and other objects are accomplished by an inferred/preferred instance play apparatus. The inferred/predicted instance play apparatus includes a situation detector for detecting physical and electrical situations in a ubiquitous space; an information storage for storing the physical and electrical situations as temporalspatial information; and a temporalspatial information player for obtaining a requested temporalspatial information in response to an instance play request and playing the temporalspatial information, wherein the temporalspatial information player plays, if the requested temporalspatial information is retrieved in the information storage, the retrieved temporalspatial information and, otherwise, infers or predicts the requested temporalspatial information using the temporalspatial information stored in the information storage.

Preferably, the situation detector includes a physical situation detector for detecting a physical situation including weather and traffic information and devices statuses collected by devices installed in the ubiquitous space; and an electronic situation detector for detecting electronic situation including CPU and memory utilization of computing system of in the ubiquitous space and Internet information.

Preferably, the information storage stores the physical and electrical situations periodically. Preferably, the information storage stores physical and electrical situations detected when a predetermined event occurs.

Preferably, the temporalspatial information player includes a reproduction controller for retrieving the requested temporalspatial information in response to a past information request, outputting, if the information is not retrieved, an inference request, and outputting a prediction request in response to a future information request; an inference/prediction unit for inferring the requested temporalspatial information using the temporalspatiai information stored before and after a time corresponding to the temporalspatial information in response to the inference request, and predicting the requested temporalspatial information in response to the prediction request; and a player for replaying the temporalspatial information retrieved, inferred, and predicted.

Preferably, the inference/prediction unit includes a pattern searcher for retrieving a historic pattern matched to the requested temporalspatial information by analyzing the requested and before and after temporalspatial information; a casual relationship analyzer for analyzing a casual relationship of instances of the temporalspatial information between before and after the request temporalspatial information; and a temporalspatial inference unit for inferring the requested temporalspatial information in response to a past information request on the basis of the historic pattern and causal relationship using a predetermined inference algorithm; and a temporalspatial prediction unit for predicting the requested temporalspatial information in response to a future information request on the basis of the historic pattern using a predetermined prediction algorithm. Preferably, the temporalspatial inference unit and the temporalspatial prediction unit store the inferred and predicted temporalspatial information into the information storage.

Preferably, the temporalspatial prediction unit compares the stored predicted temporalspatial information and occurred temporalspatial information and modifies the prediction algorithm according to the comparison result.

In accordance with another aspect of the present invention, the above and other objects are accomplished by an inferred/predicted instance player method. The inferred/predicted instance play method includes receiving a request for playing a temporalspatial information; determining whether the request is a past information request; retrieving, if the request is a past information request, the temporalspatial information from a storage; predicting, if the request is not a past information request, the temporalspatial information; and playing the temporalspatial information in a ubiquitous space. Preferably, retrieving the temporalspatial information includes inferring, if the temporalspatial information is not retrieved from the storage, the temporalspatial information.

Preferably, inferring the temporalspatial information includes obtaining a historic pattern by analyzing before and after information of the requested information; and reproducing the temporalspatial information on the basis of the historic pattern using a predetermined inference algorithm.

Preferably, inferring the temporalspatial information further includes storing the inferred temporalspatial information.

Preferably, predicting the temporalspatial information includes retrieving a historic pattern matched to the requested temporalspatial information from the storage; and reproducing the temporalspatial information on the basis of the historic pattern using a predetermined prediction algorithm.

Preferably, predicting the temporalspatial information further includes storing the predicted temporalspatial information. Preferably, predicting the temporalspatial information further includes comparing the predicted temporalspatial information and correspondingly collected temporalspatial information; and correcting the prediction algorithm on the basis of the comparison result.

[ADVANTAGEOUS EFFECTS]

As described above, the instance play apparatus of the present invention can infers or predicts a contextual instance at a past or future time requested by a user and plays the inferred/predicted contextual instance, whereby the user feels like existing in a spatial temporal environment of the instance.

[BRIEF DESCRIPTION OF THE DRAWINGS]

FIG. 1 is a schematic block diagram illustrating an instance play apparatus according to an exemplary embodiment of the present invention;

FIG. 2 is a block diagram illustrating a configuration of the situation detector of FIG. 1 ;

FIG. 3 is a block diagram illustrating a configuration of the temporal-spatial reproducer of FIG. 1 ; F^IG- 4 is a block diagram illustrating a configuration of the inference/prediction unit of FIG. 3;

FIG. 5 is a flowchart illustrating an instance play method according to an exemplary embodiment of the present invention;

FIG. 6 is a flowchart illustrating a past instance processing procedure of the instance play method of FIG. 5

FIG. 7 is a flowchart illustrating an inference process of FIG. 6, and

FIG. 8 is a flowchart illustrating a prediction process of the instance play method of FIG. 5.

[BEST MODE FOR CARRYING OUT THE INVENTION!

Exemplary embodiments of the present invention are described with reference to the accompanying drawings in detail. The same reference numbers are used throughout the drawings to refer to the same or like parts. Detailed descriptions of well-known functions and structures incorporated herein may be omitted to avoid obscuring the subject matter of the present invention.

In order to provide a time traveling effect in the cyber space, the instance play apparatus of the present invention analyzes the events occurred in the real word and stores the events in the forms of classified attributes such as space, user, object, situation, and time (that are interchangeably used with temporal-spatial information) and creates a contextualized cyber space using the instance using the temporal-spatial information.

The time traveling function reproduces the temporal and spatial statuses of the events at the time when the user request to move such that the user is fantasized to be there at that time. Since the time traveling function reproduces a virtual world using the spatial-temporal information in consideration of the casual relationships of the events, it makes the user feel moved to the requested time. With this function, the present invention overcomes the temporal restriction of the virtual space and makes the user to experience the time travel as using a time machine. FlG. 1 is a schematic block diagram illustrating an instance play apparatus according to an exemplary embodiment of the present invention. As shown in FIG. 1 , the instance play apparatus 100 includes a situation detector 110, an information storage 130, and a temporal-spatial reproducer 150. Referring to FIGs. 1 to 8, the structure and operation of the instance play apparatus 100 is described hereinafter in more detail.

Before traveling to a temporal-spatial instance, i.e. reproducing the temporal-spatial information corresponding to the requested time, the instance play apparatus 100 stores the temporal-spatial information in the real world (in this embodiment, the ubiquitous space). In order to store the temporal-spatial information, the situation detector 110 detects physical situations and electrical situations in the ubiquitous space, and the information storage 130 stores the physical and electrical situations as the temporal-spatial information.

Since various events occur in the ubiquitous space continuously, the situation detector 110 should monitor the events (situations) occurring in the ubiquitous space. Although recording all events is the best way for reproduction in future, it is preferred to store preregistered interesting events for effectively managing the memory space since no event may occurs in the ubiquitous space in a long time.

An instance of the real space can be composed of various types of information. For example, a plurality computing devices installed in the ubiquitous space run for executing various tasks and the elements arranged as the part of the ubiquitous space, typically provided with electronic tags, collect information on the status and location information continuously.

The computation goals and status of the computing devices stored allow recognizing the services and tasks executed at that time, and the information collected by sensor network of the ubiquitous space allows recognizing the arrangement of the elements and objects, such that the instance can be reproduced in response to a future request.

In this embodiment, the instance play apparatus monitors and analyzes the situation of the ubiquitous space and stores the analyzed result as temporal-spatial data. Preferably, the situations are classified into physical situation and electrical situation and then stored as the temporal-spatial data. As described above, the physical situation includes physical status such as weather and traffic statuses. Typically, the elements arranged within the ubiquitous space are provided with electrical tags for collecting the status information, and the status information is part of the physical status. The electrical situation includes the information such as the CPU and memory utilizations of the computing devices of the ubiquitous system and information shared through Internet.

FIG. 2 is a block diagram illustrating a configuration of the situation detector of FIG. 1. As shown in FIG. 2, the situation detector includes a physical situation detector 111 and an electrical situation detector 113 that collect status information and transfer the collected to the information storage 130. The physical situation detector 111 collects the physical situation information such as weather and traffic status in the ubiquitous space and status information of elements of sensor network. The electrical situation detector 113 detects the electrical situation of the ubiquitous space such as CPU and memory utilizations of the computing system and information shared through Internet.

The temporal-spatial information is required to be stored in an efficiency manner rather than being stored for all the situations. In this embodiment, the physical and electrical situations are periodically collected to be stored in the information storage 130. Also, the instance play apparatus can be configured such that, if an important event occurs between the storage periods, (for example, a specific time registered by the user or an abrupt increase of the traffic), the event can be stored in the form of physical or electrical situation. As described above, the situation detector 110 detects the physical and electrical situations and stores the detected situations within the information storage 130 periodically and at a time when a predetermined event occur. The Term "instance" is interchangeably used with the temporal-spatial information. The temporal-spatial information is stored within the information storage 130 periodically such that the temporal-spatial information is used, when a replaying of a specific instance is requested, for inferring or predicting the instance and playing the inferred or predicted instance.

FIG. 3 is a block diagram illustrating a configuration of the temporal-spatial reproducer of FIG. 1, and FIG. 5 is a flowchart illustrating an instance play method according to an exemplary embodiment of the present invention. If a user inputs a request for playing an instance of a specific time point, the temporal-spatial reproducer 150 receives the request at step S510 and determines whether the time point is of the past or future at step S530. The temporal-spatial reproducer 150 produces and plays an instance according to the determination result. In more detail, the temporal-spatial reproducer 150 search for the temporal-spatial information within the information storage 130 and reproduces and plays, if the temporal-spatial information of the requested time point is retrieved (which means that the requested time point is past), the instance of the requested time point. If any temporal-spatial information of the requested time point is not retrieved in the information storage 130 (which means that the requested time point is unrecorded past or future), the temporal-spatial reproducer 150 infers the temporal-spatial information of the past time point or predicts the temporal-spatial information of the future time point and plays an instance produced by the inferred or predicted temporal-spatial information.

Referring to FIG. 3, the temporal-spatial reproducer 150 includes a reproduction controller 151 , an inference/prediction unit 153, and a player 155.

If the requested time point is past, the reproduction controller 151 retrieves the temporal-spatial information of the requested time point within the information storage 130. If the temporal-spatial information of the requested time point is retrieved, the reproduction controller 151 reproduces an instance using the temporal-spatial information. On the other hand, if the temporal-spatial information of the requested time point is not retrieved, the reproduction controller 151 determines whether the requested time point is past or future and outputs an inference or prediction request to the inference/prediction unit 153 according to the determination result.

The reproduction controller 151 contextualizes a past instance using the temporal-spatial information before and after the requested time point in response to the inference request or predicts a future instance with temporal-spatial information in response to the prediction request. The inferred or predicted instance is output to the player 155.

FIG. 4 is a block diagram illustrating a configuration of the inference/prediction unit of FIG. 3. Referring to FIG. 4, the inference/prediction unit 153 includes a pattern searcher 1531 , a casual relationship analyzer 1533, a temporal-spatial prediction unit 1537, and a temporal-spatial inference unit 1535.

In a case of past time point request, the pattern searcher 1531 analyzes the temporal-spatial information before and after the requested past time point and generates a historic pattern on the basis of the analysis result. The casual relationship analyzer 1533 analyzes the casual relationship of the temporal-spatial information before and after the request past time point.

The temporal-spatial inference unit 1535 infers the temporal-spatial information of the requested past time point on the basis of the historic pattern and causal relationship using a predetermined inference algorithm. Preferably, the temporal-spatial inference unit 1535 stores the inferred temporal-spatial information into the information storage 130.

The temporal-spatial prediction unit 1537 predicts the temporal-spatial information of the requested time point on the basis of the historic pattern using a predetermined prediction algorithm. Preferably, the temporal-spatial prediction unit

1537 stores the predicted temporal-spatial information into the information storage

130.

In a case of future time point request, events supposed to be occurred at the requested time are arranged. Preferably, the temporal-spatial prediction unit 1537 compares the stored predicted temporal-spatial information and past temporal-spatial information and modifies the prediction algorithm according to the comparison result.

Referring to FIG. 3, the player 155 replays the instance produced using the temporal-spatial information that is retrieved by the reproduction controller 151 or inferred and predicted by the inference/prediction unit 153.

The replay method of the player is preferably provided in the form of an advanced virtual reality and augmented reality. For example, the object information stored within the information storage 130 is processed by the electronic tag system such that, in order to bring in images or behavior information, the player 155 checks the meta information of corresponding objects from an object name server (ONS) of the electronic tag system.

The player 155 also checks computing status of the requested time point using the computation information of the computing devices installed in the ubiquitous. In this manner, the player 155 brings in the images and behaviors of objects and the computing status and replays the instance with the temporal-spatial information in the ubiquitous space implemented in the form of the augmented reality and virtual space.

FIG. 6 is a flowchart illustrating a past instance processing procedure of the instance play method of FIG. 5, FIG. 7 is a flowchart illustrating an inference process of FIG. 6, and FIG. 8 is a flowchart illustrating a prediction process of the instance play method of FIG. 5.

As described above, if a user request for playing an instance of a specific time point (S510), the reproduction controller 151 determines whether the requested time point is of a past time point (S530). If the requested time point is a past time point, the player 155 plays an instance corresponding to temporal-spatial information stored at the request time point (S550).

An instance playback in response to the requested past time point is described hereinafter with reference to FIGs. 6 and 7 in more detail. Referring to FIG. 6, if it is determined that the requested time point is a past time point at step S530 of FIG. 5, the reproduction controller 151 determines whether the temporal-spatial information corresponding to the requested time point exists in the information storage 130 at step S551. If the temporal-spatial information corresponding to the requested time point exists, the reproduction controller 151 outputs the retrieved temporal-spatial information to the player 155 such that the player 155 replays the temporal-spatial information (instance) at step S553.

If the temporal-spatial information corresponding to the requested time point does not exist, the reproduction controller 151 outputs an inference request to the inference/prediction unit 153. If the inference request is received, the inference/prediction unit 153 extracts temporal-spatial information before and after the requested time point and infers the temporal-spatial information using the extracted temporal-spatial information. The inferred temporal-spatial information is output to the player 155 such that the player 155 replays the inferred temporal-spatial information at step S555.

How to replay the inferred temporal-spatial information is described with reference to FIG. 7 in more detail. The pattern searcher 1531 analyzes the temporal-spatial information at and before and after the requested time point and generates a historic pattern at step S5551. In the meantime, the casual relationship analyzer 1533 analyzes the casual relationships between the time points. The temporal-spatial inference unit 1537 infers temporal-spatial information at the requested time point using the retrieved historic pattern and/or the casual relationship and outputs the inferred temporal-spatial information to the player 155. Consequentially, the player 155 replays the temporal-spatial information output by the temporal-spatial inference unit 1537 at step S555. Next, temporal-spatial inference unit 1537 outputs the inferred temporal-spatial information to be stored into the information storage 130 at step S5555.

In this embodiment, the object information, which is collected by the electronic tags, can be inferred by comparing the instances before and after the request past time point, and the services and task status of the computing devices can be inferred by analyzing check pointing status and consistency status of the computing system.

In this embodiment, the inference algorithm can be one of a Bayesian network algorithm which infers on the basis of probability using past history and a neural network algorithm which infers on the information comparison analysis.

Also, an intermediate stage of the computing system can be inferred by analyzing the stages of the processes, whereby most of the services and tasks can be inferred in such manners. Since the temporal-spatial information is periodically saved, there exist the casual relationships between the saved temporal-spatial data. Accordingly, a short saving-point interval facilitates to infer the temporal-spatial information.

Referring to FIG. 5 again, if the requested time point is not a past time point, the temporal-spatial reproducer 150 predicts the temporal-spatial information at the requested time point and plays the predicted temporal-spatial information in the ubiquitous space (S570). How to predict and play future temporal-spatial information is described hereinafter with reference to FIG. 8.

In a case of future time point request, the reproduction controller 151 outputs a prediction request to the inference/prediction unit 153. The inference/prediction unit 153 predicts temporal-spatial information in response to the prediction request and outputs the predicted temporal-spatial information to the player 155. In more detail, the pattern searcher 1531 searches for a historic pattern corresponding to the temporal-spatial information of the requested time within the information storage 130 at step S571. For example, the pattern searcher 1531 can retrieve a matched historic pattern by detecting whether a similar historic pattern in the basic information related to the requested time point such as date, time zone, appointment, etc.

The temporal-spatial prediction unit 1537 predicts the temporal-spatial information of the requested future time point on the basis of the retrieved historic pattern using a predetermined prediction algorithm at step S573. For example, the temporal-spatial prediction unit 1573 predicts the temporal-spatial information of the requested time point using the prediction information (such as weather, temperature, humidity, traffic, identical time zone, data, location, and schedule) and the retrieved historic pattern. Although not depicted in drawings, regarding the future prediction-based instance play, the casual relationship analyzer 1533 can analyzes the casual relationships between the time points of the retrieved historic pattern. In this case, the temporal-spatial prediction unit 1537 predicts the temporal-spatial information using the historic pattern and casual relationship. The temporal-spatial prediction unit 1537 outputs the predicted temporal-spatial information to be stored into the information storage 130 at step S575.

In the case of future time point request, some events are supposed to be occurred. The temporal-spatial information corresponding to these events is also supposed to be stored into the information storage 130, whereby the temporal-spatial prediction unit 1573 compares the stored predicted temporal-spatial information with the actually collected temporal-spatial information at step S577 and preferably modifies the prediction algorithm according to the comparison result at step S579.

Although not shown in drawings, in a case that the information for prediction is lack, it is preferable to output an announce message informing that the instance of the requested time instance cannot be reproduced.

Although the instance play apparatus and method are described in association with playing temporal-spatial information of a requested time point, the present invention is not limited thereto. For example, the instance play apparatus and method of the present invention can be applied to various virtual reality application fields such as virtual education, virtual experience, and on-line tour that strive to overcome special restriction, and unified ubiquitous space application fields such as electronic tab-based distribution, convenient and secure apartments, public facilities, and city projects.

The above described embodiments of the present invention are programmable and can be recorded in a computer-readable storage media in the form of a program read by universal digital computers. Also, the data structures of the instance play method of the above described embodiments of the present can be recorded in the computer readable storage media using various means.

The computer readable storage media includes a magnetic storage media (e.g. Read Only Memory (ROM), floppy disc, hard disc), optical digital disk (e.g. Compact Disk (CD) and Digital Video Disk (DVD)), and carrier wave (e.g. transmission on Internet).

Although exemplary embodiments of the present invention have been described in detail hereinabove, it should be clearly understood that many variations and/or modifications of the basic inventive concepts herein taught which may appear to those skilled in the present art will still fall within the spirit and scope of the present invention, as defined in the appended claims.

[INDUSTRIAL APPLICABILITY]

The instance play apparatus and method of the present invention is applicable to a ubiquitous environment.

Claims

[CLAIMS]

1. An inferred/predicted instance play apparatus comprising: a situation detector for detecting physical and electrical situations in a ubiquitous space; an information storage for storing the physical and electrical situations as temporalspatial information; and a temporalspatial information player for obtaining a requested temporalspatial information in response to an instance play request and playing the temporalspatial information, wherein the temporalspatial information player plays, if the requested temporalspatial information is retrieved in the information storage, the retrieved temporalspatial information and, otherwise, infers or predicts the requested temporalspatial information using the temporalspatial information stored in the information storage.

2. The inferred/predicted instance play apparatus of claim 1 , wherein the situation detector comprises: a physical situation detector for detecting a physical situation including weather and traffic information and devices statuses collected by devices installed in the ubiquitous space; and an electronic situation detector for detecting electronic situation including CPU and memory utilization of computing system of in the ubiquitous space and Internet information.

3. The inferred/predicted instance play apparatus of claim 1 , wherein the information storage stores the physical and electrical situations periodically.

4. The inferred/predicted instance play apparatus of claim 3, wherein the information storage stores physical and electrical situations detected when a predetermined event occur.

5. The inferred/predicted instance play apparatus of claim 1 , wherein the temporalspatial information player comprises: a reproduction controller for retrieving the requested temporalspatial information in response to a past information request, outputting, if the information is not retrieved, an inference request, and outputting a prediction request in response to a future information request; an inference/prediction unit for inferring the requested temporalspatial information using the temporalspatial information stored before and after a time corresponding to the temporalspatial information in response to the inference request, and predicting the requested temporalspatial information in response to the prediction request; and a player for replaying the temporalspatial information retrieved, inferred, and predicted.

6. The inferred/predicted instance play apparatus of claim 5, wherein the inference/prediction unit comprises: a pattern searcher for retrieving a historic pattern matched to the requested temporalspatial information by analyzing the requested and before and after temporalspatial information; a casual relationship analyzer for analyzing a casual relationship of instances of the temporalspatial information between before and after the request temporalspatial information; and a temporalspatial inference unit for inferring the requested temporalspatial information in response to a past information request on the basis of the historic pattern and causal relationship using a predetermined inference algorithm; and a temporalspatial prediction unit for predicting the requested temporalspatial information in response to a future information request on the basis of the historic pattern using a predetermined prediction algorithm.

7. The inferred/predicted instance play apparatus of claim 6, wherein the temporalspatial inference unit and the temporalspatial prediction unit store the inferred and predicted temporalspatial information into the information storage.

8. The inferred/predicted instance play apparatus of claim 7, wherein the temporalspatial prediction unit compares the stored predicted temporalspatial information and occurred temporalspatial information and modifies the prediction algorithm according to the comparison result.

9. An inferred/predicted instance play method comprising: receiving a request for playing a temporalspatial information; determining whether the request is a past information request; retrieving, if the request is a past information request, the temporalspatial information from a storage; predicting, if the request is not a past information request, the temporalspatial information; and playing the temporalspatial information in a ubiquitous space.

10. The inferred/predicted instance play method of claim 9, wherein retrieving the temporalspatial information comprises inferring, if the temporalspatial information is not retrieved from the storage, the temporalspatial information.

11. The inferred/predicted instance play method of claim 10, wherein inferring the temporalspatial information comprises: obtaining a historic pattern by analyzing before and after information of the requested information; and reproducing the temporalspatial information on the basis of the historic pattern using a predetermined inference algorithm.

12. The inferred/predicted instance play method of claim 11 , wherein inferring the temporalspatial information further comprises storing the inferred temporalspatial information.

13. The inferred/predicted instance play method of claim 9, wherein predicting the temporalspatial information comprises: retrieving a historic pattern matched to the requested temporalspatial information from the storage; and reproducing the temporalspatial information on the basis of the historic pattern using a predetermined prediction algorithm.

14. The inferred/predicted instance play method of claim 13, wherein predicting the temporalspatial information further comprises storing the predicted temporalspatial information.

15. The inferred/predicted instance play method of claim 14, wherein predicting the temporalspatial information further comprises: comparing the predicted temporalspatial information and correspondingly collected temporalspatial information; and correcting the prediction algorithm on the basis of the comparison result.