JPS6380325A - Information retrieval method and index used therefor - Google Patents

Abstract

PURPOSE:To attain efficient updating and retrieval of data by providing an index corresponding to a key word and a revised time so as to reference the index at the retrieval of updated data. CONSTITUTION:In applying addition and revision to a retrieved data of a retrieval object data storage section 6 based on the data inputted from a terminal device 4, the correspondence between the revised data and the key word in the retrieval index reserved in the retrieval index storage section 7 and the correspondence between the revised data and the key word and data revision time in the reservation retrieval index stored in a reservation retrieval index storage section 8 are executed. Then a reservation retrieval processing section 9 compares the key word in the retrieval condition with the key word in the storage section 8 at the time designated at each retrieval condition of the retrieval condition storage section 3 according to a prescribed logic equation, the revised data satisfying the equation is used to retrieve the revised data taken place after the preceding retrieval under the retrieval condition and the result is transferred to a terminal device 10. The data updating and retrieval are executed efficiently by adopting the system using the key word and index.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an information retrieval method and an index used therein, and in particular, the present invention relates to an information retrieval method and an index used therein, and in particular, the present invention relates to an information retrieval method and an index used therein. The keywords in the conditions and the keywords in the index are compared according to a predetermined logical formula, and updated data that has keywords that satisfy this predetermined logical formula is also searched since the previous search using the search conditions. Search for updated update data,
An information search method suitable for so-called reservation information search (hereinafter referred to as "reservation search") and an index used therein;
That is, it relates to a reserved search index.

[Conventional technology]

Scheduled search is a relatively new field, and several methods have been tried so far.

First, among the search target data, only updated data is duplicated, a scheduled search is performed for the updated data that is duplicated, and after the search is completed, the duplicated updated data is This is a method of erasing it. A second method is to attach a mark to the search target data itself to indicate that it is updated data, and to erase this mark after executing the scheduled search.

[Problem that the invention seeks to solve]

However, in the first method above.

(1) Since there is no correspondence between updated data and updated data, the search has to be executed at the same time for multiple search conditions registered by multiple operators.The operator registers each conditional expression in advance. (2) Updated data is saved overlappingly, regardless of time, so additions or changes to the search target data are saved overlappingly with the search target data. The problem is that the update work is inefficient because it needs to be performed on both the update data and the update data. Furthermore, in the second method described above, (1) since no correspondence is established between update data and update data, a search must be executed at the same time for multiple search conditions registered by multiple operators; (2) Search conditions registered in advance by the operator for all search target data, regardless of whether there is updated data, cannot be executed at the scheduled search time that the operator registers for each conditional expression in advance. After searching for search target data that satisfies the requirements, each search target data is checked for a mark indicating that it is updated data, and those with this mark are taken as search results, resulting in a problem of poor search efficiency.

The present invention has been made in view of the above circumstances, and its first purpose is to solve the above-mentioned problems in the conventional reservation search method and to efficiently update data and search for updated data. The objective is to provide a possible information retrieval method. A second purpose is to provide a reserved search index suitable for use in the above information search.

[Means for solving problems]

A first object of the present invention is to provide an information retrieval system that searches updated data for items that match search conditions registered in advance by an operator, in which an index for registering keywords and data update times of the updated data is provided. By the information search method, the updated data is associated with the keyword and data update time in the index at the time of data update, and the index is referred to at the time of search execution to obtain data updated within the search target period. achieved.

Further, a second object of the present invention is to provide an update period corresponding section of data provided corresponding to keywords included in update data, and an identifier storage section that stores an identifier of update data corresponding to each entry of the update period corresponding section. and a time table that divides update periods corresponding to each entry in the update period corresponding section;
This is achieved by an information retrieval index composed of an index indicating the current time on the time table.

[Effect]

The present invention achieves the above purpose by providing an index (index for scheduled search) for associating updated data with keywords and data update times when updating data, and by referring to this index when executing scheduled searches. This is what I did.

〔Example〕

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

FIG. 2 is a block diagram of an information retrieval system showing one embodiment of the present invention. In the figure, 1 is a terminal device operated by an operator who registers search conditions, 2 is a search condition processing unit that registers search conditions, 3 is a search condition storage unit that stores registered search conditions, and 4 is a search A terminal device operated by an operator who adds or changes target data; 5 a search target data processing unit that adds or changes search target data; 6 a search target data storage unit that stores search target data; 7 a search a search index storage unit that stores a search index that associates all target data with keywords; 8;
9 is a scheduled search index storage unit that stores a scheduled search index that associates update data with keywords and data update times; 9 is a scheduled search processing unit that executes a scheduled search; IO is operated by an operator who receives the results of the scheduled search. A terminal device is shown.

The search condition processing unit 2 stores search conditions input from the terminal device 1 in the search condition storage unit 3. The search target data processing unit 5 adds or changes the search target data stored in the search target data storage unit 6 based on the data input from the terminal device 4, and also adds or changes the search target data stored in the search target data storage unit 6 and the updated data and the search index storage unit 7. a scheduled search processing unit that associates keywords in the search index to be saved, and associates updated data with keywords and data update times in the reserved search index saved in the reserved search index storage unit 8; 9 compares the keywords in the search conditions with the keywords in the reserved search index at the time specified for each search condition stored in the search condition storage unit 3 according to a predetermined logical formula, and calculates this logical formula. To search for updated data that has a keyword that satisfies and that has occurred since the previous search under the search conditions being searched. The processing results are sent to the terminal device rn10.

FIG. 3 is a diagram conceptually showing the contents of the reserved search index stored in the reserved search index storage section 8. As shown in FIG. In the figure, 11 is a keyword storage section for storing keywords, and 12 is an update period correspondence section for storing a table having entries corresponding to data update periods, and this table exists in the keyword storage section 11 corresponding to the keywords. Further, 13 is an update data ID storage section that stores an identifier of update data, and the identifier of each update data is stored in the update period correspondence section 12 associated with the same keyword as the keyword of each update data in the keyword storage section 11. It is associated with the entry corresponding to the update time of each update data in the table.

Further, 14 is a time table section in which each entry in the table of the update period correspondence section 12 indicates a corresponding data update period, and the entry number of the entry in this table is the entry number of the entry in the table of the update period correspondence section 12. They match, and the value of each entry is an initial value of NULL, and the time is set at the time of starting the reserved search. Reference numeral 15 denotes a current entry number storage section that indicates to which entry in the table of the update period correspondence section 12 the identifier of update data should be associated when updating data.

FIG. 4 is a diagram conceptually showing a table storing the previous search time for each search condition registered in advance. In the figure, 16 is a search time storage unit that stores the previous search condition for each search condition written down in advance, and the initial value is the registration time of the search condition.

FIG. 5 is a diagram showing the relationship between the update time of search target data and the scheduled search time. In FIG. 0, tit to ti3 are the scheduled search times under search condition 1, t2. .

t2□ indicates the time of scheduled search under search condition 2, and tx indicates the time when the search target data of identifier E was updated. Furthermore, A, B, C, D, and E are identifiers A, B, and E, respectively.
This indicates that the search target data having C, D, and E has been updated.

Hereinafter, using the example shown in FIG. 5, the mechanism and usage of the scheduled search index will be explained in accordance with FIGS. 3, 4, and the flowchart in FIG. 1 showing the overall flow of the operation of this embodiment. After explaining the overall flow of operations using FIG. 1, we will explain how to associate update data with keywords and update times.

The process is divided into the following steps: a search method using pre-registered search conditions, and a method for canceling the association between update data, keywords, and update times.

(1) First, the overall flow of the operation of this embodiment will be outlined with reference to FIGS. 1(a) and 1(b).

FIG. 1(a) shows the operation of the search target data processing section 5 shown in FIG. The search target data processing unit 5 receives the update data from the terminal device 5!4 using the keyword, ir
When the update data is obtained together with the a benzushi (process 21), the updated data is stored in the search target data storage unit 6 (process 22), and then the keyword and identifier of the data are registered in the search index storage item 7. (Process 23) Similarly, the keyword and identifier of the data are registered in the reserved search index storage section 8 (Process 24).

Subsequently, the search target data processing unit 5 associates the update data with the keyword and update time in accordance with the procedure detailed in the next section. When this association is completed, the reserved search processing unit 9 performs a search using the pre-registered search conditions according to the operation flow shown in FIG. 1(b).

The scheduled search processing unit 9 obtains the search conditions from the search condition storage unit 3 (process 25), obtains the identifier of the search target data that matches the search conditions (process 26), and obtains the search target data from this. (Process 27), and the retrieved data is sent to the terminal device 10 (Process 28).

The process of associating the update data with keywords and update times will be described in detail below.

(2) The method for associating update data with keywords and update times is as follows (see Figure 3).

First, a keyword that matches the keyword included in the update data is obtained from the keyword storage section 11. A table in the update period correspondence section 12 that is associated with this keyword is found. If there is no keyword in the keyword storage section 11 that matches the keyword included in the update data, the keyword is added to the keyword storage section 11, and the update period corresponding section 1
Add a table associated with this keyword in 2.

Next, the entry number of the table in the update period correspondence section 12 is obtained from the current entry number storage section 15, which links the identifier of the update data. The identifier of the update data in the update data ID storage unit 13 that is already associated with the entry with this entry number is obtained, and the identifier of the new update data is added to this. If the update data has multiple keywords, the above process is repeated for the number of keywords.

In FIG. 3, the update data of identifier A has keyword 1 and keyword 2, and identifiers B and D.

It is assumed that the update data of E has keyword 1, and the update data of va besushi C has keyword 2.

#1. $2. ...#n indicates the entry number, and entry numbers #1 and #2 of the table in the time table section 14
The value of each entry is tzzt t2□, which indicates that the identifier of update data updated by time t1□ is associated with the entry with entry number #1 in the table of the update period correspondence unit 12.

The entry with entry number #3 is associated with an identifier of update data that will be updated from time t2□ to time t1□ when the next scheduled search starts. According to the example in FIG. 5, identifiers A and B are associated with the entry number #1 in the update period correspondence section 12 table associated with keyword 1,
The entry with entry number #2 has no corresponding identifier, the entry with entry number #3 is associated with the identifier E, and the identifier E is added at time tx.

Furthermore, the entry with entry number #1 in the table of update period correspondence unit 12 associated with keyword 2 is associated with identifier A, and the entry with entry number #2 is associated with identifier C.

As described above, this embodiment has the advantage that it is possible to easily associate update data with keywords and update times.

(3) Next, a supplementary explanation will be given regarding the search method using pre-registered search conditions (see FIG. 3).

First, the entry number is obtained from the current entry number storage section 15, and the value of the current entry number storage section 15 is changed to the entry number of the entry in the table of the update period correspondence section 12 to be used next. The identifier of the update data updated after this point is associated with the entry in the table of the update period correspondence unit 12 having the entry number after the change. A scheduled search start time is set in the entry in the time table section 14 that has the entry number before the change in the current entry number storage section 15. In Fig. 3, the current entry number storage section 1
The value of 5 is changed from #3 to #4, and the value of tzz is set in the entry number #3 of the time table section 14.

Next, the identifier of the search condition that is currently being searched is compared with the search condition identifier in the search time storage section 16, and the previous search time corresponding to the matching search condition identifier is obtained. An entry number in the time table section 14 having a value greater than this previous search time is obtained.

Next, keywords that match the keywords in the above-mentioned search conditions are obtained from the keyword storage section 11. A group of identifiers of update data associated with the entry number obtained above in the table of the update period correspondence unit 12 associated with this keyword is obtained.

If the search condition is composed of a logical formula of a plurality of keywords, this process is repeated for the number of keywords, and from the obtained update data identifier group, the update data identifiers that satisfy the logical formula are obtained. Update data having the identifiers of these update data is obtained from the search target data storage unit 6.

As described above, according to this embodiment, it is possible to efficiently search for data that matches the search conditions and that has been updated since the previous search.

(4) Finally, a method for canceling the association between update data, keywords, and update times will be explained.

Deletion of the update data identifier is performed for those for which scheduled searches have been completed using all registered search conditions.In other words, the oldest previous search time is obtained from the search time storage unit 16, and the value of this entry is set to NULL. At the same time, the update data identifier group associated with the entry with this entry number in the table of the update period correspondence unit 12 is deleted.

As described above, this embodiment has the advantage that the identifier of updated data for which a scheduled search has been completed can be easily removed from the scheduled search index.

〔Effect of the invention〕

As described above, according to the information retrieval method of the present invention, in an information retrieval system that searches for updated data that matches search conditions registered in advance by an operator, keywords of the updated data and data update An index for registering time is provided, and when updating data, the updated data is associated with the keyword and data update time in the index, and when executing a search, the index is referenced to obtain data updated within the search target period. As a result, an information retrieval method that can efficiently update data and search for updated data can be realized.
This has the remarkable effect of being extremely effective in realizing the above-mentioned reserved search method.

Further, the index of the present invention includes a data update period corresponding section provided corresponding to a keyword included in the update data, an identifier storage section storing an identifier of the update data corresponding to each entry in the update period corresponding section, and the above-mentioned. Since it is composed of a time table that divides update periods corresponding to each entry in the update period corresponding section and an index indicating the current time on the time table, it provides an information retrieval index suitable for the above information retrieval method. It is something that can be done.

[Brief explanation of the drawing]

FIGS. 1(a) and (b) are operational flowcharts of an information retrieval method that is an embodiment of the present invention, FIG. 2 is a block diagram of an information retrieval system that is an embodiment of the present invention, and FIG. 3 is the above-mentioned Diagram conceptually showing the contents of the reserved search index, No. 4
The figure is a diagram conceptually showing a table storing the previous search time for each search condition registered in advance, and FIG. 5 is a diagram showing the relationship between the update time of the search target data and the reserved search time. 1: terminal device, 2: search condition processing unit, 3: search condition storage unit, 4: terminal device, 5: search target data processing unit, 6: search target data storage unit, 7: search index storage unit,
8: Scheduled search index storage unit, 9: Scheduled search processing unit, 10: Terminal device, 11: Keyword storage unit, 12:
Update period correspondence section, 13: Update data ID storage section, 14:
Time table section, 15: Current entry number storage section, 16:
Search time storage section. Figure 1 (a) (b)
Figure 2 Figure 3 Figure 4 Figure 1 Day

Claims (2)

[Claims] (1) In an information retrieval system that searches newly added or changed search target data (updated data) for items that match search conditions registered in advance by an operator, keywords of the updated data and data updates are used. An index for registering time is provided, and when updating data, the updated data is associated with the keyword and data update time in the index, and when executing a search, the index is referenced to obtain data updated within the search target period. information retrieval method. (2) a data update period correspondence section provided for keywords included in the update data; an identifier storage section for storing update data identifiers corresponding to each entry in the update period correspondence section; An information search index consisting of a time table that divides update periods corresponding to each entry, and an index that points to the current time on the time table.