JPH09245046A - Information retrieval device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform efficient retrieval aligned with the sensation of a user by utilizing the relation of the number of times of access and access date and time as an access history at the time of trying access again to information accessed in the past and performing the retrieval with the access history as a retrieval key. SOLUTION: For the respective pieces of the information managed by an information management part 1, an access history management part 2 manages the number of times of the access in the past and the access date and time as the access history and an access history feature extraction part 3 extracts the timewise change of the number of times of the access of the respective pieces of the information from the access history as an access history feature. Then, at the time of a retrieval operation, when an access period in the past for the information desired by the user and an access frequency in the period are inputted from an access history input part 4 as the retrieval key, an information retrieval part 5 retrieves the information provided with the access history feature similar to the inputted retrieval key from the information management part 1 and a retrieved result is outputted from a display device 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for retrieving desired information from a large number of pieces of information stored and managed.
When trying to access the information accessed in the past again,
By specifying the characteristics of past access history, such as information that was intensively accessed at a specific time or information that is regularly accessed, you can quickly search for the information you want without searching all the information. Regarding equipment that can.

[0002]

2. Description of the Related Art In recent years, an enormous amount of information converted into electronic data is stored in a database and managed, and it is required to efficiently retrieve desired information from such information. In order to improve the efficiency of such information retrieval, the following inventions have been conventionally proposed.

In the file system described in Japanese Patent Laid-Open No. 2-39225, header information indicating the storage position of each stored information file is arranged in the order of high frequency of use of the information file. , The header information is searched for in the most frequently used manner, so that a desired information file can be quickly searched. However, according to the present invention, although an information file that has been frequently accessed recently can be easily searched, it is possible to access information files that have been accessed a year ago, such as information that has been accessed frequently at a certain time in the past. If they tried to access the information again, they could not do a valid search.

Further, the information search system described in Japanese Patent Laid-Open No. 3-22084 discloses a menu having a tree-type data structure, in which a selection of options with a high frequency of access to data or a latest access time is selected. By creating and using the menu,
It is intended to realize quick retrieval of desired data. In the information search system described in Japanese Patent Laid-Open No. 4-10153, in the above information search system, a menu is created in which a frequently accessed option is extracted for each title extracted from the options, and the menu is created. By using it, it is intended to realize a quick search for desired data.

However, in these inventions,
Only one of the access frequency and the access time was treated as the standard for creating a menu, and there was no method that considered the relationship between the access frequency and the access time. If you try to access previously accessed information again, even if the access frequency is the same, you will access information that was accessed very often a long time ago, but is rarely accessed now, and recently accessed frequently. Information should be treated differently, and even the most recently accessed information should be considered as being divided into information that has been accessed less frequently before and information that has been accessed more frequently. In addition, even if the access frequency is low, even if there is a request to efficiently search for information that is regularly accessed or information that is concentrated and accessed at a specific time, these inventions make a request. I could not be satisfied.

Further, the method of storing and retrieving information corresponding to time-varying information disclosed in Japanese Patent Laid-Open No. 5-307569 examines the periodicity of the frequency of occurrence of current affairs information and compares it with the time period input by the user. From the frequency of information generation,
It is intended to realize to retrieve the optimum information of the time period requested by the user from the database. However, in the present invention, only the current information that periodically occurs is targeted, and although past information can be searched based on the time period, it is efficient for information that has no periodicity in past access. Search was not feasible.

[0007]

As described above, various inventions have been made in order to improve the efficiency of information retrieval, but the keys used to retrieve information in these inventions (timing , Access frequency, etc.) is different from what the user engaged in the search work actually thinks, and it was not possible to realize the search according to the user's feeling. The present invention has been made in view of the above-mentioned conventional circumstances, and when attempting to access the information accessed in the past again, the relationship between the access count and the access date and time is used as the access history, and the access history is searched as the search key. Accordingly, it is an object of the present invention to provide an information search device that realizes an efficient search according to the user's feeling.

[0008]

In order to achieve the above object, in the information retrieval apparatus according to the present invention, the access history management means accesses the individual information managed by the information management means in the past access count. Further, the access date and time is managed as an access history, and the access history characteristic extraction means extracts the time change of the number of times of access of each information as the access history characteristic from the access history. Then, in the search operation, when the user inputs the past access time to the desired information and the access frequency at that time as the search key from the access history input means, the access history feature similar to the input search key by the information search means The information managing means is searched for from the information managing means, and the searched result is output from the output means.

That is, in this information retrieval apparatus, for each piece of information managed by the information management means, the number of times the user has accessed that information in the past and the date and time when the access was made are accessed. The access history management means manages the history. The access history characteristic extraction means arranges the access history for each piece of information according to the passage of time, and extracts the characteristics of the access history regarding the access timing and the access frequency. When a user accesses the information in the past and searches for information that he / she wants to access again, when the information was accessed in the past,
The access frequency, such as frequent access or frequent access, is specified by the access frequency history input means as a search key. The information retrieval means compares the access history of the input time and frequency with the characteristics of the access history of each information extracted by the access information characteristic extraction means, retrieves information having a similar access history, and performs this retrieval. The result is displayed and output by output means such as a display device.

Further, in the information retrieving apparatus according to the present invention, the access history managing means manages the past number of accesses and the access date and time as an access history for each piece of information managed by the information managing means. The access history feature extraction means extracts the time change of the access count of each information from the history as the access history feature. Then, when performing a search operation, when the user inputs a temporal transition of the past access frequency to desired information in the form of a sentence from the access frequency transition input means, the access frequency transition analysis means inputs the temporal access frequency of the past access frequency. Analyze the transition and extract the access count and access date and time,
The access transition feature extraction unit extracts the time change of the number of times of access of information from the extracted access count and access date and time as the access transition feature, and the information search unit has an access history feature similar to the extracted access transition feature. The information is retrieved from the information management means, and the retrieved result is output from the output means.

In other words, in this information retrieval device, when the past access time or the number of accesses is unclear, the user may use a sentence or a sentence to say "I used to access very often, but in the last few months I have not accessed it at all. It is input from the access transition input means by using a temporal transition of the past access frequency such as "not available" as a search key. The input temporal transition of the access frequency is analyzed by the access frequency transition analysis means, and the access transition characteristic extraction means extracts the analyzed characteristic of the access transition. Then, the information search means compares the access transition feature extracted by the access transition feature extraction means with the access history feature extracted by the access information feature extraction means, and has a similar access history. The existing information is searched, and the search result is output by the output means.

Further, in the information retrieving apparatus according to the present invention, the access history managing means manages the past access count and access date and time as an access history for each piece of information managed by the information managing means. The access history feature extraction means extracts the time change of the access count of each information from the history as the access history feature. When the user inputs the time distribution of the past access frequency to the desired information in the form of a graph from the access frequency distribution input means during the search operation, the access frequency distribution feature extraction means inputs the temporal access frequency of the past access frequency. The distribution is analyzed to extract a characteristic, the information retrieval means retrieves information having an access history characteristic similar to the extracted access frequency distribution characteristic from the information management means, and outputs the retrieved result from the output means.

That is, in this information retrieval apparatus, when it is difficult to express the access time or the access frequency in a sentence or a sentence, or when a more detailed frequency is designated, the user has a temporal distribution of the access frequency. Is input from the access distribution input means using the graph representing the as a search key.
The access frequency distribution characteristic extraction means analyzes the temporal distribution of the past access frequencies input and extracts the characteristics, and the information retrieval means extracts the characteristics of the extracted access distribution by the access information characteristic extraction means. Information having a corresponding access history is searched for in comparison with the characteristics of the access history, and the search result is output by the output means.

[0014]

DETAILED DESCRIPTION OF THE INVENTION An information retrieval apparatus according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of an information search device according to the first embodiment of the present invention.
The information retrieval apparatus of the present embodiment includes an information management unit 1 that manages individual information stored as a database, a past access count and access date and time for individual information managed by the information management unit 1. An access history management unit 2 that manages as an access history, an access history feature extraction unit 3 that extracts, as an access history feature, a change in the number of times of access to each piece of information from the access history managed by the access history management unit 2. An access history input unit 4 for inputting, as a search key, a time when a user accesses the information and an access frequency at that time in a search operation for information that has been accessed in the past;
The feature of the input search key is the access history feature extraction unit 3
An information retrieval unit 5 that retrieves information having similar characteristics as compared with the characteristics of the access history extracted in 1. and a display device 6 as an output unit that displays and outputs the retrieval result are provided.

The information management unit 1 manages a large number of information (documents, books, etc.) stored in a database (not shown) by their names (document names, book names, etc.), and identifies each information. It is managed as possible. Note that various known methods can be used as a method for the information management unit 1 to manage each information. For example, each information may be attached with an identifier and managed by this. Further, the information management unit 1 may be configured as a part of a database, and may manage the name of each information together with the content of each information. The access history management unit 2 manages the past number of accesses and the access date and time for each piece of information as an access history corresponding to the information management unit 1, and the information managed by the information management unit 1 is accessed every time the information is managed. Update access history for information. The access history feature extraction unit 3 extracts the time change of the access count from the access history of each piece of information as a feature of the access history, and as shown in FIG. 2, the access frequency (the number of accesses) associated with each piece of information. The access history feature is retained as temporal transition data of.

The access history input unit 4 is composed of an input means having a keyboard or the like, and inputs the time when the user accessed when searching for desired information and the access frequency at that time as a search key. . The information retrieval unit 5
Based on the input search key, the access history feature extraction unit 3 searches the access history feature for each piece of information, and extracts information having similar features as a search result. The display device 6 is composed of a display device and displays a list of names of retrieved information. The display device 6 may be composed of a printer, and the search results may be printed out.

Next, by the information retrieval apparatus having the above configuration,
For information that the user has accessed in the past,
The process when trying to access again will be described. When the user inputs from the access history input unit 4 by using the time when the information is accessed and the access frequency at that time as the search key, the information search unit 5 holds the characteristics of the input search key in the access history feature extraction unit 3. Information having similar characteristics is searched for by comparing with the characteristics of the access history.

For example, the user thought "I would like to see again the documents that were frequently accessing the database when I was studying C language before." And said, "I was studying C language in 5 of this year." It was from June to June. Since I saw it many times, I think that I have been accessing it at least 100 times or more. "{Access time: Access frequency} = {1995-5 to 19955.6: "OVER
The information "100"} "is input as a search key.
As shown in FIG. 2, the information search unit 5 selects the features of the access history stored in the access history feature extraction unit 3 as shown in FIG.
Information having a history characteristic that has been accessed 100 times or more between May 1995 and June 1995 is searched, and the searched information name is displayed on the display device 6 as shown in FIG. For example, information names that match the search key are "patent writing", "used car market trends", and "How To"
In the case of "Programming C" and "technical report creation procedure", these document names are displayed in a list.

Therefore, the desired information "How To
"Programming C" can be easily found by a search key including ambiguity based on memory. When the user selects a desired information name from the displayed information names by the input means, the information management unit 1 extracts the information of the location where the information exists and the content of the information, and the display device 6 is displayed and output.

FIG. 4 shows the configuration of the information retrieval apparatus according to the second embodiment of the present invention. The same parts as those in the first embodiment described above are designated by the same reference numerals, and duplicate description will be omitted. In the information retrieval apparatus of this embodiment, instead of the access history input unit 4 of the first embodiment, an access frequency transition input means for inputting a temporal transition of past access frequency for desired information by a user in a sentence format. 14, and an access frequency transition analysis unit 17 that analyzes the temporal transition of the input past access frequency and extracts the access count and access date and time.
And an access transition feature extraction means 18 for extracting the temporal change of the access count of information from the extracted access count and access date and time as a feature of access transition. In addition, the information retrieval means 15 provided in this embodiment
Searches the information having the access history feature similar to the feature of the access transition extracted by the access transition feature extraction unit 18 from the access history feature extraction unit 3 by the comparison process.

Next, by the information retrieval device having the above configuration,
For information that the user has accessed in the past,
The process when trying to access again will be described. When the user accesses the information, the user can access the access frequency input section 4 by a sentence or a sentence to show the temporal transition of the access frequency.
When input from, the access frequency transition analysis unit 17 analyzes the sentence or the sentence to extract the access count and the access date and time, and the access transition feature extraction unit 18 uses the access count and the access date and time to change the access transition over time. It is extracted as a feature of. Then, the information search unit 15 compares the characteristics of the access transition with the characteristics of the access history stored in the access history characteristic extraction unit 3 and searches for information having similar characteristics.

For example, the user thinks "I want to search for an e-mail about how to set up a printer", and the previous access status to this e-mail is described as "I used to refer frequently, but recently I only accessed once". When input, the access frequency transition analysis unit 17 analyzes words that mean time and number of times from this sentence, and {access time: access frequency} = {"before", "recent": "frequent", " The analysis result relating to the access frequency transition of once "}" is passed to the access transition feature extraction unit 18, and the access transition feature extraction unit 18 extracts the access transition feature from the access frequency transition as shown in FIG. In other words, the access transition feature extraction unit 18 displays, for example, ““ before ”= 1 year or more ago”, ““ recent ”= 1 week or less”, ““ frequent ”= 100.
Corresponding information that numerically represents a word, such as “more than once”, is set in advance, and the access transition feature is extracted based on the relationship between the time and the number of accesses as illustrated in FIG.

Then, the information retrieving unit 15 compares the access history feature of each information held in the access history feature extracting unit 3 with the access transition feature extracted by the access transition feature extracting unit 18, Information having similar characteristics is searched for, and the search result is displayed and output on the display device 6. For example, an information name having a feature A indicated by a dotted line in FIG. 6 and an information name having a feature B indicated by a dashed line in FIG. 6 are searched for as having a history feature similar to the feature shown in FIG. Note that the features shown in FIG. 5 and the line A shown in FIG.
Or, there is a slight difference from the characteristic of B, but a difference of a preset degree is allowed, and it is set to search for those having similar characteristics.

As a result of the search processing, for example, when the searched information names are “network security”, “printer setting”, and “power supply arrangement list”, FIG.
As shown in FIG. 5, these information names are displayed in a list on the display device 6, and the desired mail “printer setting” can be easily found by inputting a sentence format including ambiguity based on the memory.

FIG. 8 shows the configuration of an information retrieval apparatus according to the third embodiment of the present invention. The same parts as those in the first embodiment described above are designated by the same reference numerals, and duplicate description will be omitted. In the information retrieval apparatus of the present embodiment, instead of the access history input unit 4 of the first embodiment, an access frequency distribution input means 24 for inputting a time distribution of past access frequencies for desired information by a user in a graph is provided. An access frequency distribution feature extraction unit 27 that analyzes the input distribution of past access frequencies and extracts the feature of the access transition is provided. In addition, the information retrieval means 25 provided in this embodiment
Searches the access history feature extraction unit 3 for information having access history features similar to the access transition features extracted by the access frequency distribution feature extraction means 27 by comparison processing.

Next, by the information retrieval device having the above configuration,
For information that the user has accessed in the past,
The process when trying to access again will be described. The access frequency distribution input unit 24 allows the user to draw a graph showing the temporal distribution of access frequency when accessing the information.
When the input graph is input from, the access frequency distribution feature extraction unit 27 analyzes the graph and extracts the time change of the access count as the feature of the access transition. Then, the information search unit 25 compares the characteristics of the access transition with the characteristics of the access history stored in the access history characteristic extraction unit 3, and searches for information having similar characteristics.

For example, the user thought "I want to refer to the information regarding the usage status of copy paper", and the previous access status to that information was roughly: "I accessed several times in the past. Based on this, when the user inputs the graph as shown in FIG. 9 from the access frequency distribution input unit 24, the access frequency distribution feature extraction unit 27 analyzes the input graph and passes the characteristics to the information retrieval unit 25. That is, when the user inputs the graph (FIG. 9) showing the change in access frequency with time,
As indicated by the thick line in FIG. 10, the access frequency distribution feature extraction unit 27 extracts the contour line of this graph as a feature of the temporal distribution of access frequency.

Then, the information retrieval unit 25 compares the characteristics of the access history of each information held in the access history characteristic extraction unit 3 with the characteristics of the access transition extracted by the access frequency distribution characteristic extraction unit 27. , Information having similar characteristics is searched, and the search result is displayed and output on the display device 6. For example, an information name having the characteristic C indicated by the dotted line in FIG. 11 is searched for as having the history characteristic similar to the characteristic indicated by the thick line in FIG. Although there is a slight difference between these features, a preset degree of difference is allowed, and settings are made so as to search for those having similar features.

As a result of this search processing, for example, when the searched information names are "copy paper usage status", "Web access frequency distribution map", and "address book",
As shown in FIG. 2, these information names are displayed on the display device 6 as a list, and desired information “copy paper usage status” can be easily found by inputting in a graph format including ambiguity based on memory. .

Next, an example of a method for searching for information having similar characteristics by allowing a slight difference from the access history characteristics in the above-described search processing will be described. FIG.
In the example shown in (1), the characteristic D of the access history is given a range of access times (for example, 5 times), and the range E within the range is searched and extracted as similar. Also, FIG.
In the example shown in FIG. 4, the inflection points P0 to
The time intervals t1 to t4 are obtained for P4, and the ratios of these time intervals are compared to search and extract as similar ones. It should be noted that in the present invention, various methods other than these methods can be adopted, and the point is that at least one of the temporal element and the number-of-times element should have a width in the comparison process.

[0031]

As described above, according to the information retrieval apparatus of the present invention, the characteristic is extracted from the relationship between the access frequency and the access time, and the information is retrieved based on the characteristic of the access history. The necessary information can be quickly retrieved by the user's vague memory about the past access frequency. In addition, since it can be used to detect information having a certain periodicity with the passage of time, and it is also possible to collect information having the same tendency of change in access frequency. It is also possible to improve the access efficiency by classifying the information items according to the information having the same tendency of access frequency.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an information search device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating characteristics of an access history according to the first embodiment of the present invention.

FIG. 3 is a diagram showing a display example of search results according to the first embodiment of the present invention.

FIG. 4 is a configuration diagram of an information search device according to a second embodiment of the present invention.

FIG. 5 is a diagram illustrating a feature extraction result of access transition according to the second embodiment of the present invention.

FIG. 6 is a diagram illustrating characteristics of an access history according to the second embodiment of the present invention.

FIG. 7 is a diagram showing a display example of search results according to the second embodiment of the present invention.

FIG. 8 is a configuration diagram of an information search device according to a third embodiment of the present invention.

FIG. 9 is a diagram illustrating a search key according to a third embodiment of the present invention.

FIG. 10 is a diagram showing an analysis result of a search key according to the third embodiment of the present invention.

FIG. 11 is a diagram illustrating access history features according to the third embodiment of the present invention.

FIG. 12 is a diagram showing a display example of search results according to the third embodiment of the present invention.

FIG. 13 is a diagram illustrating an example of a similarity determination principle according to the present invention.

FIG. 14 is a diagram illustrating another example of the similarity determination principle according to the present invention.

[Explanation of symbols]

1 ... Information management unit, 2 ... Access history management unit,
3 ... Access history feature extraction unit, 4 ... Access history input unit, 5, 15, 25 ... Information search unit, 6 ...
..Display devices, 14 ... Access frequency transition input unit,
17 ... Access frequency transition analysis unit, 18 ... Access transition feature extraction unit, 24 ... Access frequency distribution input unit, 27 ... Access frequency distribution feature extraction unit,

Claims (3)

[Claims] 1. An information management means for managing a plurality of information, an access history management means for managing the number of times of access and access date and time for each piece of information managed by the information management means as an access history of each piece of information, said access An access history characteristic extracting means for extracting a time change of the number of times of access of individual information from the history as an access history characteristic, and an access history input means for inputting a past access time to desired information and an access frequency at the time as a search key. An information retrieval means for retrieving information having an access history characteristic similar to the input retrieval key from the information management means, and an output means for outputting a result retrieved by the information retrieval means. Information retrieval device. 2. An information management means for managing a plurality of information, an access history management means for managing the access count and access date and time of each information managed by the information management means as an access history of each information, and the access. An access history feature extraction means for extracting a time change of the access count of each information from the history as an access history feature, and an access frequency transition input means for inputting a temporal transition of the past access frequency for desired information in a sentence format, An access frequency transition analysis unit that analyzes the input temporal changes in the access frequency and extracts the access count and access date and time, and the time transition of the access count of information from the extracted access count and access date and time. Access transition feature extraction means to extract as a feature, and extracted access estimation Information retrieval apparatus characterized by comprising: the information retrieval means for retrieving information with access history characteristics from the information management means for similar features, and output means for outputting the results retrieved by the information retrieval means, the. 3. Information management means for managing a plurality of information, access history management means for managing the number of times of access and access date and time for each piece of information managed by the information management means as an access history of each piece of information, said access An access history feature extraction means for extracting the time change of the access count of each information from the history as an access history feature, and an access frequency distribution input means for inputting the time distribution of the past access frequency for the desired information in a graph format, Access frequency distribution feature extraction means for analyzing the temporal distribution of past access frequencies extracted to extract features, and information for searching the information management means for information having access history features similar to the extracted access frequency distribution features A search means, and an output means for outputting the result searched by the information search means. An information retrieval device characterized in that