JPH11316765A - System and method for data base retrieval, data base registering device, data base retrieving device, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately grasp an occurring event not through an operator and to perform proper work, operation, or prediction corresponding to it. SOLUTION: A data input means 1 which inputs waveform data changing with time, a registering means 4 which registers data inputted in arbitrary timing among respective feature quantity data extracted from respective inputted waveform data in the form of a data base 5 while giving comments etc., and a retrieving means 6 which performs retrieval from the data base 5 by using feature quantity data as a key are provided and retrieved feature data is presented to an operator together with the comment as past access given thereto. Consequently, anybody can obtain the same result through retrieval based upon feature quantity data objectively representing the state of a retrieval object and also obtain necessary information from the comment etc.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database retrieval system and method for retrieving necessary information from a database, a database registration device / retrieval device, and a recording medium, and more particularly to information that changes over time, such as waveform data. Is suitable for use in a system for registering and searching in a database.

[0002]

2. Description of the Related Art Conventionally, various operations are performed by an operator in response to various events that may occur, in accordance with the events that have occurred, and prediction of the next possible state based on the current event that has occurred. In the field. For example,
It is common practice for an operator to analyze the cause of a failure and take appropriate action when a failure occurs in equipment in a factory or a plant that manufactures products. Conversely, when a state that is a precursor to a failure occurs, some measures are taken in advance by predicting that the failure will occur.

[0003]

However, in the above example, the types of faults that occur are various, and the content of the fault that has occurred and its cause are difficult to understand immediately. In addition, the result of such determination often depends on the operator, and different determinations may be made depending on the operator.

For example, it is a common practice to record a fault that has occurred in the past as information and to make a determination with reference to the information. However, it is necessary to determine the correspondence between the record of the past fault and the current state. It is still the operator who does this, and there is room for subjectivity to enter here. Therefore, different judgments may be made depending on the operator, and in particular, when a person who is not an expert is an operator, an erroneous judgment may be made.

As described above, conventionally, regardless of the operator,
It has been difficult to accurately grasp the event that has occurred and to perform an appropriate operation, operation, prediction, or the like corresponding to the event.
Therefore, the present invention makes effective use of information on various events that have occurred in the past, regardless of the operator,
It is an object of the present invention to accurately grasp an event that has occurred and perform an appropriate operation, operation, or prediction corresponding to the event.

[0006]

According to the present invention, there is provided a database search system comprising: input means for inputting time-varying information as first numerical sequence data; and each first numerical sequence input by the input means. Registering means for assigning second numerical string data to any of the data and registering the data as a database, and search means for searching the database using the first numerical string data as a key It is characterized by.

In another aspect of the present invention, the apparatus further comprises a feature value extracting means for extracting a feature value from each of the first numerical value sequence data inputted by the input means, and the registration means comprises: The second numerical sequence data is added to an arbitrary one and registered as the database, and the search means searches the database using the feature amount data as a key. And

In another aspect of the present invention, a plurality of the input means are provided, and a synthesizing means for synthesizing the plurality of first numerical sequence data or the characteristic data thereof inputted from the plurality of input means is further provided. It is characterized by having.

In another aspect of the present invention, the search means is configured to sequentially search the database using each of the first numerical sequence data or the feature data thereof input by the input means as a key. It is characterized by.

In another aspect of the present invention, when the second numerical sequence data is added to the retrieved first numerical sequence data or its feature data on the database, It is characterized in that second numerical sequence data is also presented.

[0011] In another aspect of the present invention, the search means periodically searches the database.

Further, the database search method of the present invention
Inputting time-varying information as first numerical sequence data, providing the input first numerical sequence data with second numerical sequence data if desired, Registering the first numerical sequence data to which the first numerical sequence data or the second numerical sequence data is added as a database.

[0013] In another aspect of the present invention, inputting time-varying information as first numerical sequence data;
Registering the input first numerical sequence data as a database, and at any time, assigning the second numerical sequence data to any of the first numerical sequence data on the database. It is characterized by having a step of giving.

According to another aspect of the present invention, a step of inputting time-varying information as first numerical sequence data, and performing a search of the database by using the input first numerical sequence data as a key, Presenting the result together with the second numerical sequence data to which the result has been given.

In another aspect of the present invention, the method further comprises, after the step of inputting the first numerical sequence data, a step of extracting a characteristic amount from the input first numerical sequence data. The processing is performed using the feature data extracted from the first numerical sequence data.

Further, the database registration device of the present invention comprises:
Input means for inputting time-varying information as first numerical sequence data;
A feature value extraction means for extracting a feature value from each of the value sequence data, and a second value sequence data given to any one of the feature value data extracted by the feature value extraction means as a database. Registration means for registering.

Further, the database search device of the present invention comprises:
Input means for inputting time-varying information as first numerical sequence data;
A feature value extraction unit for extracting a feature value from each of the numerical value sequence data, and the feature value data or the second
Search means for performing a search on a database in which at least one data to which the numerical sequence data is assigned is stored, and presenting the search result together with the second numerical sequence data to which the numerical sequence data is assigned. Features.

According to the present invention, there is provided a computer-readable recording medium comprising: input means for inputting time-varying information as first numerical sequence data; and first numerical sequence data input by the input means. A program for causing a computer to function as a registering means for assigning second numerical sequence data to any one of the above and registering it as a database.

According to another aspect of the present invention, input means for inputting time-varying information as first numerical sequence data, and a feature value is extracted from each of the first numerical sequence data input by the input means. A computer for functioning as a feature value extracting means for performing the process and a registering means for adding second numerical sequence data to any one of the feature value data extracted by the feature value extracting means and registering the data as a database. Is recorded.

In another aspect of the present invention, input means for inputting information which changes over time as first numerical sequence data, and each of the first numerical sequence data input by the input means as a key, A search is performed on a database in which at least one of the first numerical sequence data or the data to which the second numerical sequence data is provided is stored, and the second numerical sequence data to which the search result is provided And a program for causing a computer to function as search means presented together.

In another aspect of the present invention, input means for inputting time-varying information as first numerical sequence data, and a feature quantity is extracted from each first numerical sequence data input by the input means. And a database storing at least one or more of the feature amount data or data to which the second numerical sequence data is added, using the feature amount data extracted by the feature amount extraction unit as a key. And a program for causing a computer to function as a search means for presenting the search results together with the second numerical string data to which the search results are assigned.

In another aspect of the present invention, a program for causing a computer to function as synthesizing means for synthesizing a plurality of first numerical sequence data input from a plurality of input means or their characteristic data is further recorded. It is characterized by the following.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a functional block diagram showing a configuration example of a database search system according to one embodiment of the present invention. In FIG. 1, reference numeral 1 denotes data input means, which periodically or irregularly inputs information that changes over time as first numerical sequence data. As an example of information that changes with time, it is assumed that waveform data is handled below. In this case, the data input means 1 is constituted by, for example, a signal detection sensor. The data input means 1 has a function of converting an input analog waveform into digital numerical sequence data.

The data input means 1 may be provided at least one in the system, but in the example of FIG.
.., are provided. By providing a plurality of data input means 1, a plurality of waveform data can be input. For example, when the database search system of the present embodiment is applied to a product manufacturing factory or plant, waveform data can be input from various steps (upper step, lower step, etc.) in the manufacturing process. In addition, by installing the data input means 1 in various facilities, it is possible to input waveform data such as temperature, humidity, pressure, etc., which change over time.

Reference numeral 2 denotes a characteristic amount extracting means for extracting characteristic amounts specific to the waveform from the first numerical sequence data (waveform data) inputted by the data input means 1. In addition,
The above-described conversion of the analog waveform into digital numerical sequence data does not necessarily need to be performed in the data input unit 1, but may be performed at least before the feature amount extraction unit 2 performs the process of extracting the feature amount. Alternatively, digital numerical sequence data may be input from the beginning.

As an example of a calculation method for extracting a feature quantity, for example, a short-time Fourier analysis, a fast Fourier transform (F
FT), wavelet transform, cepstrum analysis, and the like. If the applied calculation method is different, the extracted feature amount is also different. Which calculation method is used to extract the feature value may be determined in accordance with the property of the target first numerical sequence data or the like.

As a specific technique for extracting the characteristic amount from the numerical sequence data as described above, the present applicant has already proposed a technique as disclosed in Japanese Patent Application No. 5-84154. . The feature extraction means 2 in the present embodiment can apply the feature extraction technique disclosed herein. Since the feature amount extraction means 2 is mainly intended to compress the search space in the database search, the extracted feature amount data may be further subjected to quantization.

When a plurality of the data input means 1 are provided, the feature quantity extracting means 2 includes the data input means 1a,
1b,..., Is subjected to feature amount extraction processing. At this time, in the present embodiment, any of the above-described various calculation methods may be used, but the same calculation method is applied to each waveform data to extract a feature amount. Feature extraction means 2
Outputs the extracted feature data together with some parameters used in the arithmetic processing.

Numeral 3 denotes a synthesizing unit, which stores the characteristic amount data extracted for each waveform data by the characteristic amount extracting unit 2 into one.
Combine into two data. There are various synthesis methods, but here, as a simple example, the calculation is performed by combining several parameters when performing the operation of extracting the feature amount by the target waveform data and forming a matrix. .

Reference numeral 4 denotes a database registration unit which adds second numerical sequence data different from the waveform data (first numerical sequence data) to the feature amount data synthesized by the synthesizing unit 3 to obtain a database 5. Register as As an example of the second numerical sequence data, text data such as a comment given to the extracted feature amount can be given. This comment can be given to data input at an arbitrary timing out of the respective feature amount data input and extracted regularly or irregularly by the data input means 1.

That is, as shown in FIG. 1, the database registration means 4 has a comment giving means 4a and a data registration means 4b. The comment assigning means 4a assigns text data such as a comment to the feature amount data in accordance with an instruction from the operator, and the data registering means 4b assigns the comment or the like to the feature data. The quantity data is registered in the database 5.

The database registration means 4 configured as described above specifically operates as follows. For example, first, the comment adding unit 4a adds a comment or the like to the feature amount data output from the synthesizing unit 3 according to an instruction from the operator. If there is no instruction from the operator, no comment or the like is given. Then, the feature amount data to which the comment or the like is added or the feature amount data to which no comment is added is registered as the database 5 by the data registration unit 4b.

The operation may be performed as follows. That is, first, the data registration unit 4b registers all the feature amount data output from the synthesis unit 3 as the database 5. Thereafter, in response to an instruction from the operator, the comment providing means 4a accesses the database 5 and adds a comment or the like to the desired feature amount data. In this case, the operator can freely perform a task of giving a comment at any time, and can add or modify the content later.

Reference numeral 6 denotes a search means. The database 5 is searched by using, as keys, waveform feature data input by the data input means 1 regularly or irregularly and generated by the feature extracting means 2 and the synthesizing means 3. I do. That is, a process of comparing the feature amount data extracted from the waveform data input at the present time with various feature amount data registered in the database 5 in the past and searching for the same or similar feature amount data from the database 5 is performed. Do.

The search means 6 comprises, for example, a waveform search engine 6a and search data 6b. At the time of data registration, the waveform search engine 6a
Is performed to convert the feature amount data stored in the search data into search data 6b for the engine. On the other hand, when searching for data,
Based on the search data 6b, a process of extracting feature amount data that is the same as or similar to the search key is performed.

In the present embodiment, since a plurality of feature data are synthesized by parameter matrix formation, the search condition is satisfied when each of the features has a characteristic. It is also possible to satisfy the search condition if it has any of the features.

An output means 7 outputs the result of the search by the search means 6. This output is performed by, for example, displaying the search results on a display device (not shown) or printing out the search results using a printer (not shown). When text data such as a comment is added to the feature amount data searched by the search means 6, the output means 7 also outputs the comment and the like.

Hereinafter, the operation of the database search system according to the present embodiment configured as described above will be described in detail for a registration phase and a search phase. First, the operation of the registration phase will be described with reference to FIG. In FIG. 2, for example, a plurality of waveform data (two waveform data in FIG. 2) input at a certain timing by a plurality of data input units 1 installed in a plant, a feature amount is extracted by a feature amount extracting unit 2. Is done. The input timing of the waveform data may be specified by the operator each time, or may be automatically input at predetermined time intervals.

The feature value data ',' of the two waveforms thus extracted are synthesized by the synthesizing means 3. The synthesized feature amount data is registered in the database 5 by the data registration unit 4b. At this time, if an instruction to add the event comment 21 is given by the operator at that time, the comment adding unit 4a adds the event to the feature amount data. In response, a comment 21 is provided. The operator determines whether to add the comment 21. The content of the comment 21 to be given can be freely described by the operator.

For example, when the database search system of the present embodiment is used as a measure against a failure in the event that a failure occurs in the plant, the operator comments on the feature data obtained when the failure occurred in the plant. Is given. The comment 21 given at that time can be a device or facility in which a failure has occurred, a method for coping with the failure, and the like.

By performing such a process each time waveform data is input from the data input unit 1, case data 22 as shown in FIG. As is apparent from the case data 22, the waveform data is automatically input here every hour. Data generated at each timing (feature data or data annotated by an instruction from the operator) are numbered 1, 2,..., 125,. Has become.

Further, the waveform search engine 6a converts the case data 22 stored in the database 5 into search data 6b for the engine for each feature amount data. By the above-described processing, the characteristic amount data of the waveform representing the state of the plant, including a normal state and an abnormal state, is somewhat different from the past characteristic amount data at the time of the abnormal state. Cases are accumulated in the database 5.

Next, the operation of the search phase will be described with reference to FIG. As shown in FIG. 3, in the search phase, similarly to the registration phase, a plurality of waveform data input at a certain timing by a plurality of data input units 1 installed in the plant are respectively extracted by a feature amount extracting unit 2. Is extracted. If the same component is included in the original waveform, the feature amount data of the parameter used in this extraction process is extracted by searching, so that it is not always necessary to use the same parameter as in the registration phase. The feature value data ',' of the two waveforms extracted in this way are combined by the combining means 3.

The synthesized feature data is provided to the waveform search engine 6a. The waveform search engine 6a uses the given feature amount data as a search key to search data 6
A search process is performed using b. In the present embodiment, when the technique of Japanese Patent Application No. 5-84154 described above is used as a technique for extracting a feature amount of a waveform and performing a search, a fuzzy search can be performed, and thus a search is performed at this time. Even if the feature amount data serving as a search key does not completely match the already registered feature amount data, the corresponding data can be extracted within a similar range to some extent.

Reference numeral 31 denotes a search result by the waveform search engine 6a. Here, the feature amount data with the identification numbers 231, 56,... Is extracted as being the same or similar to the feature amount data of the search key.
The output means 7 extracts the relevant case data from the database 5 based on the search result (identification number) 31 output from the waveform search engine 6a, and presents it to the operator as similar case data 32.

In the similar case data 32,
The data on the left has higher matching with the search key. By presenting to the operator in such a form, it is possible to inform several possible events that are currently occurring in the plant in the order of probability.

Also, comment 2 is added to the detected case data.
When 1 is given, the output means 7 also presents the comment 21 together. Thus, for example, if the feature amount data of the search key entered at the present time matches the feature amount data registered when a failure occurred in the past within a certain similar range, the failure occurred at the current time in the past There is a possibility that a failure similar to that described above has occurred, but a comment describing the content of the failure and a remedy is presented to the operator. As a result, the operator can take a smooth response in accordance with the presented content.

As described above, when the database search system of the present embodiment is used as a countermeasure for the occurrence of a fault, if the waveform data is automatically input at predetermined time intervals, it is determined whether a fault has occurred in the search phase. It can function as a monitor for whether or not. This time interval does not need to coincide with the time interval in the registration phase.

It should be noted that the database search system of the present embodiment described above actually uses a CPU, R (not shown).
It is constituted by a microcomputer system including OM, RAM, etc.
And is controlled by the CPU according to a work program stored in the RAM. In this case, the ROM or RAM storing the work program constitutes a recording medium of the present invention.

Also, the functions of the above-described embodiment are implemented in an apparatus connected to the various devices or in a computer in the system so that various devices are operated to realize the functions of the above-described embodiment. The present invention also includes a program which is provided by supplying program code of software for operating the above various devices in accordance with a program stored in a computer (CPU or MPU) of the system or apparatus.

When the computer executes the supplied program code, not only the functions of the above-described embodiment are realized, but also the OS (operating system) or other operating system running on the computer. Needless to say, even when the functions of the above-described embodiments are realized in cooperation with application software or the like, such program codes are included in the embodiments of the present invention.

Further, after the supplied program code is stored in a memory provided in a function expansion board of a computer or a function expansion unit connected to the computer, the function expansion board or the function expansion unit is specified based on the instruction of the program code. It is needless to say that the present invention also includes a case where the CPU or the like provided in the first embodiment performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

As a storage medium for storing the above-described work program or software, in addition to the above-described ROM and RAM, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-R
An OM, a magnetic tape, a nonvolatile memory card, or the like can be used.

In the above embodiment, a system capable of executing both the registration phase and the search phase has been described. However, a device for performing the operation of the registration phase and a device for performing the operation of the search phase are separately provided. May be configured.

As described above in detail, in the present embodiment, in the registration phase, the feature amounts of the waveform data input periodically or irregularly are extracted, and comments are added as necessary to register them in the database. Keep it. In the search phase, the feature amounts of the input waveform data are extracted on a regular or irregular basis to search the database, and the search results are presented to the operator together with comments.

As described above, according to the present embodiment, since the comment based on the past knowledge is presented by the search based on the characteristic data of the waveform that objectively represents the state of the plant, it depends on the judgment of the operator. No matter who goes, the same result can be obtained. In addition, necessary information can be easily known at that time. In other words, knowledge accumulated in the past can be shared among multiple operators, including skilled and unskilled operators, and by this, anyone can accurately grasp the events that have occurred and take appropriate measures corresponding to them. Will be able to do it.

In the above embodiment, the operator determines whether or not to add a comment to the feature amount data in the registration phase. However, this can be automatically performed. is there. For example, when the database search system according to the present embodiment is used as a countermeasure for occurrence of a failure, a means for monitoring a frequency component of the extracted feature amount data is provided, and when this suddenly changes, it is determined that a failure has occurred and prepared in advance. The comment may be automatically added. By providing such a fault monitoring means, it becomes possible to register only the feature amount data when a fault occurs in the database 5.

In the above embodiment, the case where the database search system of the present embodiment is used as a countermeasure when a failure occurs has been described as an example, but the present invention is not limited to this. . For example, in the registration phase, a predetermined comment is added to feature amount data that goes backwards from the time of occurrence of a failure and is regarded as a precursor to the failure and registers the feature amount data. Then, in the search phase, it is possible to predict the occurrence of a failure by enabling the comment to be presented when the above-mentioned precursor data is detected.

As another application example, the feature amount data of the seismic waveform input in the registration phase is stored together with comments as needed, and the feature amount data that matches the search key periodically extracted in the search phase is stored. By detecting the corresponding comment when is detected, it is possible to predict the occurrence of an earthquake. In the registration phase, a medical waveform such as an electrocardiogram is input, the feature amount data is stored together with the comment, and the feature amount data searched for in the search face is presented together with the corresponding comment, so that appropriate medical care is provided. It is also possible to provide support for doing so.

In the present invention, the first numerical sequence data, which is information that changes with time, is not limited to the above-described waveform data, but may be moving image data. For example, by inputting image data such as a weather chart that changes with time and extracting feature data, and performing a search as described above, it is possible to use the data for prediction of weather information, for example. is there.

In the above embodiment, text data such as a comment is given as an example of the second numerical sequence data to be added to the feature data, but the present invention is not limited to this. For example, a comment or the like for the occurred event may be given as audio data, or image data related to the occurred event may be given.

In the above embodiment, the feature amount extracting means 2
However, as described above, this is mainly for compressing the search space in the database search. Therefore, when the capacity of the database 5 has a margin or when there is a long time interval for inputting the first numerical sequence data, the feature amount extracting means 2 is not provided, and the inputted numerical sequence data itself is not provided. May be handled in the subsequent processing.

[0063]

As described above, according to the present invention, time-varying information is inputted as first numerical sequence data, and among the inputted first numerical sequence data, data inputted at an arbitrary timing. Is added to the second numerical sequence data (for example, comment) and registered as a database, a database search is performed using the first numerical sequence data as a key, and the search result is stored in the second numerical sequence data. Since it is presented together with the data, it is possible to obtain the same result no matter who performs the search by the search based on the first numerical sequence data that objectively represents the state of the search target, and to add the comment with the necessary information. Etc. can be obtained. As a result, knowledge accumulated in the past can be shared by a plurality of operators, and anyone can accurately grasp an event that has occurred and take an appropriate action corresponding to the event.

[Brief description of the drawings]

FIG. 1 is a functional block diagram illustrating a configuration example of a database search system according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining an operation in a registration phase of the database search system according to the embodiment;

FIG. 3 is a diagram for explaining an operation in a search phase of the database search system according to the embodiment;

[Explanation of symbols]

 Reference Signs List 1 data input means 2 feature extraction means 3 synthesis means 4 database registration means 4a comment assignment means 4b data registration means 5 database 6 search means 6a waveform search engine 6b search data 7 output means 21 comments 22 case data 31 search results 32 similar Case data

Claims (21)

[Claims] An input means for inputting time-varying information as first numerical sequence data, and a second numerical data for an arbitrary one of the first numerical sequence data input by the input means. A database search system comprising: registration means for assigning numerical string data and registering the database as a database; and search means for searching the database using the first numerical string data as a key. 2. The registering means according to claim 1, wherein said registering means assigns second numerical string data to any of the first numerical string data input by said input means, 2. The database retrieval system according to claim 1, further comprising data registration means for sequentially registering the input numerical string data or the numerical string data generated by said adding means as a database. 3. The data registering means for sequentially registering each of the first numerical sequence data inputted by the input means as a database, and an arbitrary one of the first numerical sequence data on the database. 2. A database search system according to claim 1, further comprising: an assigning unit that assigns second numerical sequence data to the data. 4. The apparatus according to claim 1, further comprising: a feature amount extracting unit configured to extract a feature amount from each of the first numerical sequence data input by the input unit. 2. The method according to claim 1, wherein the second numerical sequence data is added and registered as the database, and the search means searches the database using the feature amount data as a key. Database search system described. 5. The apparatus according to claim 1, wherein a plurality of said input means are provided, and a synthesizing means for synthesizing a plurality of first numerical sequence data input from said plurality of input means or characteristic amount data thereof is further provided. The database search system according to claim 1. 6. The search means according to claim 1, wherein said search means sequentially searches said database using each of the first numerical sequence data or its characteristic amount data inputted by said input means as a key. The database search system according to any one of claims 1 to 5. 7. When the second numerical sequence data is added to the retrieved first numerical sequence data or the feature data thereof on the database, the retrieval means may execute the second numerical sequence data. The database search system according to any one of claims 1 to 6, wherein the database search system is also presented. 8. The database search system according to claim 6, wherein said search means periodically searches said database. 9. The database search system according to claim 1, wherein the first numerical sequence data is waveform data. 10. The data processing apparatus according to claim 1, wherein the second numerical sequence data is text data such as a comment.
The database search system according to any one of the above items. 11. A step of inputting time-varying information as first numerical sequence data, and a step of adding second numerical sequence data to the input first numerical sequence data if desired. And registering the input first numerical sequence data or the first numerical sequence data to which the second numerical sequence data is added as a database. 12. A step of inputting time-varying information as first numerical sequence data, and a step of registering the input first numerical sequence data as a database. A database search method comprising a step of assigning second numerical sequence data to any of the above first numerical sequence data. 13. A step of inputting time-varying information as first numerical sequence data, performing a search of the database using the input first numerical sequence data as a key, and providing a search result. 13. The method of presenting a database according to claim 11, wherein the step of presenting together with the second numerical sequence data is performed. 14. The method according to claim 1, further comprising: after the step of inputting the first numerical sequence data, extracting a feature amount from the input first numerical sequence data. The database search method according to any one of claims 11 to 13, wherein processing is performed using feature amount data extracted from the data. 15. An input means for inputting time-varying information as first numerical sequence data, and a characteristic amount extracting means for extracting a characteristic amount from each of the first numerical sequence data input by the input means. Registering means for adding second numerical sequence data to any one of the feature amount data extracted by the feature amount extracting means and registering the data as a database. apparatus. 16. Input means for inputting time-varying information as first numerical sequence data, and characteristic amount extracting means for extracting characteristic amounts from each of the first numerical sequence data input by the input means. Searching a database in which at least one or more of the feature amount data or data to which the second numerical sequence data is added is stored using the feature amount data extracted by the feature amount extraction unit as a key. And a search means for presenting a search result together with the second numerical sequence data to which the search result has been added. 17. An input means for inputting time-varying information as first numerical data, and a second numerical data for any of the first numerical data inputted by the input means. A computer-readable recording medium on which a program for causing a computer to function as a registering means for adding numerical string data and registering it as a database is recorded. 18. Input means for inputting time-varying information as first numerical sequence data, feature amount extracting means for extracting a characteristic amount from each of the first numerical sequence data input by the input means, And a program for causing a computer to function as registration means for adding second numerical sequence data to any one of the feature quantity data extracted by the feature quantity extraction means and registering the data as a database. A computer-readable recording medium characterized by the above-mentioned. 19. Input means for inputting time-varying information as first numerical sequence data, and said first numerical sequence data using each of the first numerical sequence data input by said input means as a key. Alternatively, as a search means for performing a search on a database in which at least one data to which the second numerical sequence data is added is stored and presenting the search results together with the second numerical sequence data to which the search result is added A computer-readable recording medium on which a program for causing a computer to function is recorded. 20. Input means for inputting time-varying information as first numerical sequence data, feature amount extracting means for extracting a characteristic amount from each of the first numerical sequence data input by the input means, Using the feature amount data extracted by the feature amount extraction means as a key, a search is performed on a database in which at least one of the feature amount data or data to which the second numerical sequence data is added is stored. A computer-readable recording medium having recorded thereon a program for causing a computer to function as search means for presenting a search result together with the second numerical sequence data to which the search result is assigned. 21. A program for causing a computer to function as a synthesizing means for synthesizing a plurality of first numerical sequence data input from a plurality of input means or their characteristic amount data. Item 17
21. The computer-readable recording medium according to any one of claims 20 to 20.