JPH08314959A - Information retrieval device - Google Patents

Abstract

PURPOSE: To provide an information retrieval device whereby will decision by a retriever is effectively supported even when difference in the satisfaction degree of a retrieval result which is obtained by an ambiguous retrieval condition, is small. CONSTITUTION: A prescribed number of data with a high satisfaction degree with respect to the ambiguous retrieval condition which is set by an inquiry condition input means 21 is extracted. A notice condition designating means 27 changes the ambiguous retrieval condition and a notice satisfaction extent calculating means 28 obtains the satisfaction degree by the retrieval condition after change as to only the prescribed number of data that is previously extracted. A retrieval result output means displays the prescribed number of data extracted first by satisfying both satisfaction degrees based on the retrieval conditions before and after change. Since new data which does not exist in data extracted at first does not appear even if the retrieval condition is changed, the retrieval conditions may be various changed so that data which most matches with a retrieval intention is easily selected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval device for retrieving information according to ambiguous inquiry conditions,
In particular, the present invention relates to an information search device that can easily select the optimum one from a plurality of searched information.

[0002]

2. Description of the Related Art In an information search apparatus, some search conditions are input in order to search for desired information. For example, when searching a database of housing information, a concrete and quantitative value such as "within 10 minutes from a station" or "rent is less than 50,000 yen" is conventionally set as a search condition. In this case, the property that is 11 minutes from the station is excluded from the search results, even though there is no big difference from the station, which is 10 minutes. In addition, properties that are 10 minutes away from the station and properties that are 1 minute away from the station are treated equally as search results. For this reason,
There is a problem that it is difficult to set the search condition and the searcher has a heavy load to select the optimum information from the search results.

Therefore, an information retrieval device that can set ambiguous conditions instead of specific retrieval conditions has attracted attention. In such an information retrieval device, "1 from the station
It is possible to set a search condition such as "close to the station" instead of "within 0 minutes", and ambiguous conditions such as "cheap rent" instead of "rent less than 50,000 yen".

FIG. 15 shows the structure of an information search apparatus that searches information under ambiguous search conditions that have been conventionally used. In order to distinguish ambiguous search conditions from specific search conditions, ambiguous search conditions will be referred to as inquiry conditions hereinafter. The inquiry conditions are
It is set from the inquiry condition input means 101. Multiple inquiry conditions can be specified at the same time, and "and" or "or" conditions can be set between them. The data to be searched is stored in the database 102. The degree-of-coincidence calculating unit 103 is a part for calculating the degree of coincidence, which represents the degree to which each data stored in the database matches the respective inquiry conditions. For example, when three inquiry conditions are set, three coincidences are calculated individually for each data.

The satisfaction degree calculating means 104 is a part for obtaining the degree to which each data satisfies the entire inquiry condition based on each matching degree. Search result extraction means 105
Is a part for extracting data to be output as a search result from the database based on the satisfaction degree obtained by the satisfaction degree calculation means 104. The search result extraction means 105 extracts data based on, for example, the total number of data to be output as a search result and the minimum satisfaction required as the data to be output. The search result output unit 106 is a unit that displays the search results extracted by the search result extraction unit 105 in descending order of satisfaction.

[0006] If the database is data for housing information, the time required from the station is registered as a specific numerical value in the data of each property. Coincidence calculation means 103
Has a function of converting this concrete numerical value into the degree of coincidence with the inquiry condition. By converting the required time into a numerical value of 0 to 1 by a function called a membership function, conversion to the degree of coincidence is performed.

FIG. 4 shows an example of the membership function for "close to the station". As the time required from the station is shorter, the inquiry condition “closer to the station” is satisfied, and thus the degree of coincidence approaches “1”. On the contrary, if the time required from the station is long, the degree of coincidence approaches "0". In this example, when the time is 20 minutes or more, the degrees of coincidence are all “0”. In this way, the membership function is a function that converts specific numerical data such as the time required from the station into the degree of coincidence with an abstract inquiry such as “close to the station”.

The satisfaction degree calculating means 104 obtains the satisfaction degree for all the set inquiry conditions based on the agreement degree for each inquiry condition obtained by the coincidence degree calculating means 103. For example, if the agreement of a property with the conditions of "close to the station" and "cheap rent" is "0.84" and "0.72", add them and divide by 2 It is designed as a degree. When the output condition is set such that the satisfaction level is 0.5 or more and the top five data items are output, the search result extraction unit 105 matches the output condition based on the calculated satisfaction level. Data to be extracted. Then, the retrieved data are displayed in descending order of satisfaction.

Japanese Unexamined Patent Publication (Kokai) No. 3-260769 discloses an information retrieving apparatus that weights the degree of coincidence with respect to each inquiry condition and then obtains the degree of satisfaction with respect to the entire inquiry condition based on these. By weighting each inquiry condition, it is possible to set a search condition closer to the searcher's intention.

Further, in Japanese Patent Laid-Open No. 3-80369, an information retrieval apparatus is shown in which retrieval results for inquiry conditions are classified in descending order of satisfaction level, an outline of the retrieval results is displayed, and then individual detailed results are displayed. Is disclosed. This device is designed to graphically represent the distribution of satisfaction of search results. By displaying such a summary, it is possible to narrow down the information in stages and support the searcher's decision making in the initial stage.

[0011]

In an information retrieval apparatus that performs a retrieval based on an ambiguous retrieval condition that has been conventionally used, the retrieval results are displayed in descending order of satisfaction. It is possible to easily grasp high-quality information. Further, in the information search device disclosed in Japanese Patent Laid-Open No. 3-260769, since the inquiry condition can be weighted, the data matching the searcher's intention can be extracted. However, it is difficult to make a decision as to which data should be finally selected when the satisfaction level of several data from the top is the same or the difference is extremely small.

In the information retrieval apparatus disclosed in Japanese Patent Laid-Open No. 3-80369, since the outline of the inquiry result is displayed, it is possible to support the decision making in the initial narrowing down stage. However, when there is little difference in satisfaction of search results in the stage of finally narrowing down the data to one, it is difficult to make a decision as to which data should be finally selected. In such a case, since the conventional information search device changes the search condition and searches the entire database again, data that has not been extracted in the previous search may appear as a new search result. Also,
It is necessary to comprehensively judge the results of multiple searches performed by changing the search conditions. For this reason, there are problems that it becomes difficult to make a final decision, and that it takes a long time to search.

It is therefore an object of the present invention to provide an information retrieving apparatus that can effectively support the decision-making of a searcher even when the difference in satisfaction of search results is small.

[0014]

According to a first aspect of the present invention, there is provided data storage means for storing a plurality of data to be searched, and search condition specifying means for specifying an arbitrary number of ambiguous search conditions. , A satisfaction degree calculation means for obtaining a satisfaction degree indicating a degree of satisfying the search condition designated by the search condition designation means for each data stored in the data storage means, and a satisfaction degree calculated by the satisfaction degree calculation means Data selection means for selecting a predetermined number of data from the data stored in the data storage means, search condition changing means for changing the search condition designated by the search condition designating means, and data selection Change satisfaction calculation means for obtaining the satisfaction degree of the search condition after being changed by the search condition changing means for the data selected by the means. It is not provided in the broadcast search apparatus.

That is, according to the first aspect of the present invention, the satisfaction degree in the changed search condition is obtained only for the predetermined number of data selected based on the satisfaction degree obtained in the first search condition. New data does not appear in the search results even if is changed.

According to the second aspect of the present invention, a data storage means for storing a plurality of data to be searched, a search condition specifying means for specifying an arbitrary number of ambiguous search conditions, and the search condition specifying means. And a data storage means based on the satisfaction degree calculated by this satisfaction degree calculation means. The satisfaction degree calculation means obtains the satisfaction degree indicating the degree of satisfying the search condition designated by the above-mentioned data stored in the data storage means. Data selecting means for selecting a predetermined number of data from the data stored in, the search condition changing means for changing the search condition specified by the search condition specifying means, and the data selected by the data selecting means. About the change satisfaction degree calculating means for obtaining the satisfaction degree of the search condition after being changed by the search condition changing means,
The information retrieving apparatus is provided with display means for displaying together the satisfaction degree calculated by the satisfaction degree calculation means and the satisfaction degree calculated by the change satisfaction degree calculation means for a predetermined number of data selected by the data selection means. There is.

That is, according to the second aspect of the present invention, the satisfaction degree in the changed search condition is obtained only for a predetermined number of data selected based on the satisfaction degree obtained in the first search condition. New data does not appear in the search results even if is changed. Further, since the satisfaction degree according to the initial search condition and the satisfaction degree according to the changed search condition are displayed together, the change in the satisfaction degree can be easily recognized.

According to the third aspect of the invention, a data storage means for storing a plurality of data to be searched, a search condition specifying means for specifying an arbitrary number of ambiguous search conditions, and the search condition specifying means. And a data storage means based on the satisfaction degree calculated by this satisfaction degree calculation means. The satisfaction degree calculation means obtains the satisfaction degree indicating the degree of satisfying the search condition designated by the above-mentioned data stored in the data storage means. Data selecting means for selecting a predetermined number of data from the data stored in, the search condition changing means for changing the search condition specified by the search condition specifying means, and the data selected by the data selecting means. About the change satisfaction degree calculating means for obtaining the satisfaction degree of the search condition after being changed by the search condition changing means,
The satisfaction degree calculated by the satisfaction degree calculation means and the satisfaction degree calculated by the change satisfaction degree calculation means are combined for a predetermined number of data selected by the data selection means in different display types for each rank of the satisfaction degree. The information retrieval device is equipped with a display unit for displaying.

That is, according to the third aspect of the invention, the phenotypes for displaying the satisfaction levels are changed according to the satisfaction ranking. For example, as the degree of satisfaction is higher, the degree of satisfaction is displayed with a higher density, or the colors are displayed in the order of satisfaction.
As a result, it is possible to easily grasp the change in the order of satisfaction of each data before and after the change of the search condition.

In the invention according to claim 4, a data storage means for storing a plurality of data to be searched, and a search condition designating means for designating a compound search condition combining a plurality of ambiguous search conditions, The degree of coincidence representing the degree of coincidence with each ambiguous retrieval condition designated by the retrieval condition designating means is determined by the degree of coincidence calculating means for each data stored in the data storage means and the degree of coincidence calculating means. Satisfaction calculating means for obtaining, for each of the data stored in the data storing means, a satisfaction degree indicating the degree to which the complex search condition is satisfied based on the degree of coincidence with respect to the ambiguous search condition, and the satisfaction degree calculating means. Data selection for selecting a predetermined number of data from the data stored in the data storage means based on the satisfaction level And a plurality of ambiguous search conditions specified by the search condition specifying means, and a coincidence degree storing means for storing the coincidence degree for the data selected by the data selecting means among the coincidence degrees obtained by the coincidence degree calculating means. Among the search conditions, a number of ambiguous search conditions that should be noted more than others are selected, and a focused search condition designating means for designating a focused compound search condition combining these and a matching degree stored in the matching degree storage means. The information retrieval apparatus is provided with an attention satisfaction calculation means for obtaining the satisfaction degree in the attention composite search condition designated by the attention search condition designation means for the data selected by the data selection means based on the above.

That is, according to the invention of claim 4, the degree of coincidence with respect to the data extracted by the composite search condition is stored. Satisfaction with the target compound condition for the first extracted data based on the stored matching degree by selecting the ambiguous search condition that is more noticeable than the others and specifying the target compound search condition Are seeking. Even if the number of ambiguous search conditions was reduced, only the first extracted data was satisfied with the changed conditions. Therefore, the data obtained from the first search should meet the intention of the searcher. Can be easily found.

According to the fifth aspect of the present invention, a data storage unit for storing a plurality of data to be searched and a search for designating a compound search condition by weighting a plurality of ambiguous search conditions and combining them. The condition specifying means, the coincidence degree calculating means for obtaining the degree of coincidence representing the degree of coincidence with each ambiguous search condition designated by the search condition specifying means for each data stored in the data storing means, and the coincidence degree. First satisfaction degree calculating means for weighting the degree of coincidence with respect to the ambiguous search condition obtained by the calculating means to obtain the degree of satisfaction indicating the degree of satisfying the composite search condition for each data stored in the data storage means. And stored in the data storage means based on the satisfaction degree calculated by the first satisfaction degree calculation means. Data selecting means for selecting a predetermined number of data from the data, and coincidence degree storing means for storing the coincidence degree for the data selected by the data selecting means among the coincidence degrees obtained by the coincidence degree calculating means. A weighting changing means for changing the weighting specified by the search condition specifying means, and a weight after the matching degree stored in the matching degree storage means is changed by the weighting changing means to satisfy the composite search condition. The information retrieving apparatus is provided with a second satisfaction degree calculating means for obtaining a degree for the data selected by the data selecting means.

That is, in the fifth aspect of the invention, the degree of coincidence before weighting the data extracted by the composite search condition is stored. Then, the degree of satisfaction of the composite retrieval condition in which the stored degree of coincidence is weighted after the change is obtained for the first extracted data.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to embodiments.

FIG. 1 shows the outline of the configuration of an information retrieval apparatus according to an embodiment of the present invention. This information retrieval device is a CPU that plays a central role in performing various processes.
11 is provided. Various circuit devices are connected to the CPU 11 through various buses 12 such as a data bus.
Of these, the ROM (Read Only Memory) 13 is
It is a read-only memory that stores various fixed data in addition to programs. A RAM (Random Access Memory) 14 is a memory that stores data that needs to be temporarily stored when executing a program. The magnetic disk device 15 is a storage device that stores data to be searched and various data. The keyboard 16 and the mouse 17 are input devices for inputting various instructions such as inquiry conditions from a searcher. The display device 18 is a device for displaying search results and various information, and here, a CRT display is used.

FIG. 2 shows an outline of the functional structure of the information retrieval apparatus in one embodiment of the present invention. The inquiry condition input means 21 is a circuit portion for inputting inquiry conditions. The database 22 is a part that stores data to be searched. The degree-of-coincidence calculating unit 23 has a function of obtaining the degree of matching of each data with the input inquiry condition for each inquiry condition. The satisfaction degree calculation unit 24 obtains a satisfaction degree that represents the degree to which the entire inquiry condition to which each data is input is comprehensively satisfied based on the agreement degree obtained for each inquiry condition by the agreement degree calculation unit 23. Is.

The search result extracting means 25 is a part for selecting data to be extracted as a search result based on the degree of satisfaction. As the selection condition, the number of data to be extracted,
The minimum degree of satisfaction required for the data to be extracted is set. The search result output means 26 plays a role of displaying the data name extracted by the search result extracting means 25 and the degree of satisfaction thereof.

The attention condition designating means 27 is a part for changing the search condition inputted from the inquiry condition inputting means 21. It is provided with a function of designating one or more search conditions that are more noticeable than others among the search conditions input from the query condition input means, and a function of changing the weight of each search condition. The attention satisfaction degree calculating unit 28 is a unit that obtains the satisfaction degree of the search condition after the data extracted by the first search is changed by the attention condition specifying unit 27. The attention satisfaction degree calculating unit 28 is configured to obtain the satisfaction degree based on the changed search condition based on the matching degree for each inquiry condition obtained by the matching degree calculation unit 23. The satisfaction degree obtained by the attention satisfaction degree calculation unit 28 is displayed together with the satisfaction degree obtained for the first search condition.

FIG. 3 shows an example of a screen displayed on the display when an inquiry condition is input. In the upper left area of the screen, a title 31 indicating that this screen is an inquiry condition input screen is displayed. In the database name input field 32, the name of the database to be searched is entered. here,
Housing information database is set as the database name. The result extraction condition input field 33 includes a result extraction number input field 34 for designating the number of data to be output as a search result and a result extraction satisfaction input field for designating a lower limit of the satisfaction level of the data to be output. And 35 are provided. Whether the lower limit of the output number or the satisfaction level is adopted as the result output condition by pointing the mouse in the area of the result extraction number selection button 36 and the result extraction satisfaction selection button 37 and then clicking the mouse button, respectively. You can choose whether or not.

When "x" is displayed in the selection buttons 36 and 37, it means that this item is adopted as the result output condition. When the selection buttons 36 and 37 are blank, this item is not adopted as a result output condition. In addition, the numerical value of the result output condition can be input directly from the keyboard,
It can be set by clicking the increase / decrease buttons 38 and 39 with a mouse.

The inquiry condition input column 41 is a column for inputting inquiry conditions used for retrieval. Here, the inquiry condition is "new, cheap, close to the station."
Is set. This condition is input as a logical expression in which the individual conditions of "new", "cheap", and "close to the station" are combined by a logical product. The symbol “∧” in the figure is a logical symbol representing a logical product. The logical conditions that can be input include logical sum and negation in addition to logical product, which are displayed on the screen by the logical symbols “∨” and “¬”, respectively. The number added after the delimiter ":" in the inquiry condition represents the weight of each inquiry condition. The weighting is relatively set between each inquiry condition and is specified by a positive integer. In this example,
The weight of each condition is set to "1".

To start searching the database according to the set inquiry conditions, the start button 42 at the upper right of the screen is clicked with the mouse. The execution of the search can be canceled by designating the cancel button 43 with the mouse.

The coincidence degree calculating means 23 shown in FIG. 2 calculates the degree of coincidence with each inquiry condition set for each data in the designated database. The ambiguous search condition is represented by a membership function that maps the feature quantity of the data into a numerical value from 0 to 1.

FIG. 4 shows an example of the membership function. This is a membership function for "close to the station", and returns a value closer to "1" when the time required from the station is closer to "0", and a value closer to "0" when the time required from the station is longer. It is like this. For example,
The degree of coincidence with the property “close to the station” for a property that is a 10-minute walk from the station is “0.5”. Similarly, for the inquiry condition "new", a membership function with the building age as a parameter is defined, and for the inquiry condition "cheap", a membership function with the rent as a parameter is defined.
These functions are stored in the ROM 13 in advance. In the housing information database, the time required as data for each property,
The building age or rent is registered as a concrete numerical value. The membership function converts these concrete numerical values into the degree of agreement with respect to ambiguous query conditions. By mapping the range of the converted numerical value from 0 to 1 for any inquiry condition, it is possible to compare the degree of coincidence with different inquiry conditions on the same scale.

FIG. 5 shows an example of the degree of coincidence obtained for each inquiry condition. "New", "Cheap", "Close to the station" for each of the retrieved data
The degree of coincidence is calculated for each individual inquiry condition. For example, for property A, the degree of agreement for “new” is “0.9”, the degree of agreement for “cheap” is “0.6”,
The degree of agreement of “close to the station” is “0.3”. From now on, it can be seen that Property A is new and the rent is moderate, but it is relatively far from the station.

The satisfaction degree calculating means 24 calculates, for each data, a total satisfaction degree for the entire inquiry condition from the coincidence degree with respect to each inquiry condition. Satisfaction when each inquiry condition is connected by a logical product as in the screen shown in FIG. 3 is obtained by the following formula. Satisfaction = (sum of weighting) ÷ Σ (weight given to each condition ÷ degree of agreement of conditions) (1) In the case of the property A shown in FIG. 5, the degree of agreement with respect to each inquiry condition is “new” Is "0.9", which is "cheap"
Is "0.6", "close to the station" is "0.3",
Since the weightings that have been set are all "1", the degree of satisfaction is as follows. (1 + 1 + 1) ÷ {(1 ÷ 0.9) + (1 ÷ 0.6) + (1 ÷ 0.3)} = 3 ÷ (1.111 + 1.667 + 3.333) = 3 ÷ 6.111 = 0. 491 (2) In this way, the degree of satisfaction with respect to all the inquiry conditions for the property A is 0.491.

FIG. 6 shows the degree of satisfaction calculated based on the degree of coincidence of each data shown in FIG.
Equation (1) is a harmonic mean of each degree of coincidence, and even if there is one degree of low degree of coincidence in the inquiry conditions, it does not control the degree of satisfaction. . In the case of a simple logical product, the result of the logical operation is governed by its minimum value. For example, in Property A, the minimum degree of coincidence is "0.3", so the degree of satisfaction is "0.3" with a simple logical product, but by taking the harmonic mean, the effect appears gently. The degree of satisfaction is “0.491”.

The search result extraction means 25 shown in FIG. 2 extracts data to be output according to the result output condition set based on the calculated satisfaction level. Search result extraction means 2
In step 5, each data to be searched is ranked in descending order of satisfaction. Then, the data that matches the set result output condition is selected in order from the data with the highest rank.

FIG. 7 shows a state in which the data shown in FIG. 6 is rearranged in descending order of satisfaction. For example, it is assumed that data is extracted according to the result output condition set on the screen shown in FIG. Since the total number of data to be output at this time is “5” and the lower limit of satisfaction is “0.5”, the property C, the property F, the property H, the property J, and the property B are selected. If the lower limit of satisfaction is “0.8”,
Property B is not extracted, Property C, Property F, Property H, Property J
Only 4 cases will be selected.

FIG. 8 shows an example of display contents on the search result output screen. The title 51 of this screen is displayed in the upper left area of the screen. A database name 52 that is a search target is displayed on the right side thereof. The set result extraction condition is displayed in the result extraction condition display field 53. Here, it is shown that the data to be displayed is selected under the conditions of the maximum display number “5” and the lower limit of satisfaction “0.5”. The inquiry condition display field 54 displays the inquiry condition used when obtaining the displayed search result. In the search result display field 55 displayed on the left side of the lower part of the screen, the data names extracted as a result of the search and their satisfaction are displayed in descending order of satisfaction. In the figure, "new, cheap,
The housing information database is searched under the inquiry condition of “close to the station” with equal weighting, and five properties with a satisfaction level of “0.5” or more are displayed.

The attention condition designation screen call button 56 arranged in the upper right portion of the screen is a button for calling the attention condition designation screen for changing the inquiry condition. It is used when changing the query condition when the searcher cannot make a decision based on the search result based on the query condition initially set. In the search result shown in FIG. 8, the satisfaction levels of the property C and the property F are both “0.964”, which are equal. Therefore, the searcher cannot finally decide which property should be used. In such a case, the inquiry condition is changed by designating the attention condition designation screen call button 56. When the decision can be made based on the search result, the search can be ended by selecting the end button 57.

FIG. 9 shows an example of the attention condition designation screen. The title 61 of this screen is displayed in the upper area of the screen. The start button 62 arranged in the upper right part of the screen is an area for designating with a mouse the start of the search based on the changed search conditions. The cancel button 63 is
This is an area for specifying the interruption of the started search with the mouse. In the condition display field 64, the inquiry condition specified on the inquiry condition input screen of FIG. 3 is displayed. Corresponding to each inquiry condition, select button 6 on the right side
5 _{1 to} 65 ₃ are arranged. Select button 65 _1-6
5 _3, whether to use the corresponding query condition in the next search an area for selecting a mouse. By clicking the select buttons 65 _{1 to} 65 ₃ with a mouse, it is possible to alternately switch whether or not to select. The state in which "x" is displayed in the selection buttons 65 _{1 to} 65 ₃ indicates that this inquiry condition is selected as the search condition. The state where "x" is not displayed indicates that the search condition is not selected.

Select buttons 65 _{1 to} 6 corresponding to each of the inquiry conditions displayed in the condition display field 64.
Weighting designation fields 66 _{1 to} 66 ₃ are provided on the right side of 5 ₃ . The weighting can be changed directly from the keyboard, or can be changed by clicking the weight increase / decrease button 67 with a mouse.

FIG. 10 shows an example of the attention condition designation screen after changing the inquiry condition.
Here, the weights 71 and 72 of the inquiry conditions of "cheap" and "close to the station" are changed to "2". This makes it possible to obtain a search result that emphasizes these two conditions among the three inquiry conditions. The inquiry condition after being changed in this way will be referred to as an attention condition. The attention satisfaction calculation means 28 shown in FIG. 2 calculates the satisfaction degree under the attention condition for the data extracted by the inquiry condition initially set. Among the matching degrees calculated by the matching degree calculating means 23 for the initially set inquiry condition, those corresponding to the data extracted as the search result are stored in a predetermined area of the RAM 14 shown in FIG. The attention satisfaction degree calculation unit 28 obtains the satisfaction degree for the attention condition based on the stored coincidence degree. The degree of satisfaction obtained for the attention condition will be called attention satisfaction.

After changing the weighting as shown in FIG. 10, the attention satisfaction calculating means 28 calculates the attention satisfaction according to the equation (1). However, the changed value is substituted for the weighting. For example, in the case of the property C that has the highest degree of satisfaction in the first search, the matching degree of each inquiry condition is as shown in FIG.
As shown in, the “new” is “1.0”, the “cheap” is “0.9”, and the “close to the station” is “1.0”. The attention satisfaction level of the property C is calculated by the following formula. (1 + 2 + 2) ÷ {(1 ÷ 1.0) + (2 ÷ 0.9) + (2 ÷ 1.0)} = 5 ÷ (1 + 2.222 + 2.0) = 5 ÷ (5.222) = 0. 957 (3)

FIG. 11 shows the satisfaction level of attention for the data selected in the first search. The noticeable satisfaction levels of the property C and the property F, which both had a satisfaction level of “0.964”, are “0.957” and “0.978”, respectively. In the first search, it was not possible to determine which of the property C and the property F to be selected, but as a result of changing the weighting and searching, it can be seen that the property F is the most suitable.

When there are data having the same degree of satisfaction in the first retrieval result, the retrieval result extraction means 25 changes the order of the data in the order of the degree of satisfaction of interest. Compared with the order shown in FIG. 7, the order of the property C and the property F in FIG. 11 is exchanged. In addition, the ranking of the data can be set to be sorted in the order of the degree of satisfaction of attention.

FIG. 12 shows an example of the search result output screen in which the attention satisfaction degree according to the attention condition is displayed together with the satisfaction degree according to the first inquiry condition. The same display contents as those in FIG. 8 are designated by the same reference numerals, and the description thereof will be appropriately omitted. At the lower side of the inquiry condition display field 54, an attention condition display field 81 for displaying the attention condition is provided. In the attention condition display field 81, numbers are assigned in the order specified as the attention condition. The search result display field 82 includes a data name display field 8 for displaying a data name.
3, a satisfaction degree display column 84 for displaying the satisfaction degree, and a attention satisfaction degree display column 85 for displaying the attention satisfaction degree are provided.
The attention satisfaction degree display column 85 is assigned the same number as the corresponding attention condition display column. In this way, by displaying the satisfaction level when the search is performed under the original inquiry condition and the attention satisfaction level side by side, it is possible to easily grasp the change in the satisfaction level when the search condition is changed.

If the result of the search under the attention condition cannot be decided yet, the attention button 56 is designated again,
Different attention conditions can be set. The attention condition designated the second time is assigned the number “2”, and the attention condition 1
Are displayed side by side below. Further, the number “2” is assigned to the attention satisfaction regarding the attention condition designated the second time, and the numbers are displayed on the right side of the attention satisfaction 1. After that, each time a search is performed by designating the attention condition, these fields with increased numbers are additionally displayed.

FIG. 13 shows the flow of processing when such an information retrieval apparatus retrieves data.
When the search start is designated, the CPU 11 clears and initializes the stored contents of the search result storage area prepared in the RAM 14 for holding the search results (step S10).
1). Next, the inquiry condition input screen shown in FIG. 3 is displayed on the display 18, and the designation of the database name to be searched, the inquiry condition, and the result extraction condition is accepted (step S102). When the start button 53 is clicked, the search according to the set inquiry condition is started. At this time, the CPU 11 secures, in the area initialized in step S101, an area for storing the degree of coincidence and the degree of satisfaction obtained in the subsequent process, corresponding to the number of output data designated by the result extraction condition.

The CPU 11 checks whether or not the processing for obtaining the degree of satisfaction has been performed for all the data in the designated database (step S103). When there is unprocessed data (step S103; Y), one of the unprocessed data is read from the database (step S104). The CPU 11 obtains the degree of coincidence of the read data with respect to each set inquiry condition (step S105). The degree of agreement is
It is obtained using the membership function corresponding to each inquiry condition.

Next, based on the degree of coincidence with respect to each inquiry condition, the total satisfaction degree with respect to all the inquiry conditions is calculated (step S106). The calculated satisfaction is compared with the satisfaction stored in the search result storage area to rank the calculated satisfaction (step S).
107). In the search result storage area, for example, when "5" is specified as the output data number of the search result, the five data names from the top, the degree of coincidence and the degree of satisfaction are stored in descending order of satisfaction. ing. That is, the ranking performed in this process corresponds to checking whether or not the degree of satisfaction is within the top five, and if the degree of satisfaction is within the top five, checking which rank from the top. .

The CPU 11 checks whether or not the calculated satisfaction level matches the result output condition (step S10).
8). For example, the number of output data of the result output condition is "5"
If the lower limit of satisfaction is "0.5", the satisfaction is within the top five places, and if the value is 0.5 or more, it means that the result output condition is met. Become. When the result output condition is met (step S109), the data name, the matching degree, and the satisfaction degree of the matched data are stored in the search result storage area. The CPU 11 examines whether the new data matches the result output condition, so that among the data which previously match the result output condition, the data that no longer matches appears (step S110).

For example, if the number of output data is set to 5 and the number of data registered in the search result storage area is 4 or less, even if new matching data appears,
The four cases up to that point will not fail. But,
If a new match appears when five cases have already been registered, the data that had been 5th in satisfaction level until then falls to 6th. Therefore, the result output condition is not satisfied. If such data exists (step S110; Y), the CPU 11 deletes the data that does not match from the search result storage area (step S111). Even if the new data matches, if the result output condition does not match (step S110; N), the data is not deleted.

In this way, after it is determined that the satisfaction level is calculated and the result output condition is matched for one data, the process returns to step S103. When the process of obtaining the satisfaction level for all the data is completed (step S103;
N), the data matching the result output condition stored in the search result storage area and the degree of satisfaction thereof are displayed as the search result (step S112). For example, the search result is displayed as shown in FIG. After displaying the search result, the CPU 11 waits for the attention button 56 or the end button 57 to be clicked (step S1).
13, step S114).

When the attention button is clicked (step S113; Y), the attention condition designation screen shown in FIG. 9 is displayed on the display, and designation of the attention condition is accepted (step S115). Then, when the start button 62 on the attention condition designation screen is clicked, the CPU 11 calculates attention satisfaction based on the degree of coincidence of the data stored in the search result storage area (step S116), and calculates the attention satisfaction. The degree of satisfaction previously obtained is displayed together with the degree of satisfaction (step S117). Here, as shown in FIG. 12, the satisfaction degree and the attention satisfaction degree are displayed side by side. When the end button is clicked (step S114; Y), the search process ends.

Modification

In the embodiments described so far, the attention satisfaction is only displayed side by side next to the satisfaction, but in the modified example, the display color is changed according to the rank of the attention satisfaction. .

FIG. 14 functionally shows the structure of the information retrieval apparatus in the modification. The same parts as those in the embodiment shown in FIG. 2 are designated by the same reference numerals, and the description thereof will be omitted as appropriate. In the modification, the display color determining unit 9 is provided between the attention satisfaction calculating unit 28 and the search result output unit 26.
1 is provided. The display color determination unit 91 determines the display color of the search result data displayed on the search result display unit 26 according to the attention satisfaction level. Here, the display is made closer to red as the attention satisfaction becomes higher, and closer to blue as the attention satisfaction becomes lower. Taking FIG. 12 as an example, the article F is displayed in red and the article H is displayed in blue. When the search results based on two or more attention conditions are displayed, they are sequentially displayed side by side, such as attention satisfaction degree 1, attention satisfaction degree 2 ... Since the ranking of attention satisfaction differs for each search, the attention satisfaction of each data is examined for each attention condition from the top.
At this time, if the display colors are different in descending order of satisfaction, it is possible to easily understand the rank of attention satisfaction.

In the embodiment and the modified examples described above, the satisfaction degree is obtained by obtaining the harmonic mean from the agreement degree, but the method of obtaining the satisfaction degree from the agreement degree is not limited to this. For example, the geometric mean may be used. Further, the case where the original inquiry condition is limited as the attention condition and the case where the weighting is changed have been described, but other changes may be made. For example, a new inquiry condition may be added. Even in this case, the satisfaction degree based on the changed attention condition is obtained only for the data obtained by the first search. Further, although the satisfaction level and the attention satisfaction level are displayed side by side, the display type is not limited to the horizontal display type as long as they can be easily compared. For example, it may be displayed as a graphic. Further, not only the satisfaction level and the attention satisfaction level are displayed at the same time, but in addition to these, the result of comprehensively judging the satisfaction level and the attention satisfaction levels for a plurality of times, for example, an average value may be displayed.

In the modification, the color is changed, but the displayed light and shade may be changed. In addition, it goes without saying that the size of the character indicating the degree of satisfaction, the font, and the thickness of the character may be different.

[0062]

As described above, according to the first aspect of the present invention, even if the search condition is changed, new data that does not appear in the first search result does not appear, so that the data that best matches the search intention can be easily obtained. You can choose.

According to the second aspect of the invention, since the satisfaction degree according to the initial search condition and the satisfaction degree according to the changed condition are displayed together, the change in the satisfaction degree can be easily recognized.

Further, according to the third aspect of the invention, the phenotypes for displaying this are made different according to the rank of satisfaction, so that the rank of satisfaction of each data can be easily changed by changing the search condition. Can be recognized.

According to the invention described in claim 4, a search condition which is more noticeable than the others is specified among the search conditions specified first, and only the data extracted by the first search condition is searched again. Since this is done, it is possible to easily find the data obtained in the first search that is more in line with the searcher's intention. Moreover, since the degree of coincidence with respect to the data extracted by the compound search condition is stored, the satisfaction degree with the compound search condition of interest can be easily obtained in a short time. Further, since the degree of coincidence only needs to be stored for the first extracted data, a small storage area is sufficient.

Further, according to the invention of claim 5, it is possible to obtain the satisfaction level by changing the weighting of the search condition with respect to the first searched data. Can be found in.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing an outline of a functional configuration of an information search device according to an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing an example of a screen displayed on a display when an inquiry condition is input.

FIG. 4 is a characteristic diagram showing an example of a membership function.

FIG. 5 is an explanatory diagram showing an example of the degree of coincidence obtained for each inquiry condition.

FIG. 6 is an explanatory diagram showing the degree of satisfaction obtained based on the degree of coincidence of each data shown in FIG.

FIG. 7 is an explanatory diagram showing a state in which the data of FIG. 6 is rearranged in descending order of satisfaction.

FIG. 8 is an explanatory diagram showing an example of a search result output screen.

FIG. 9 is an explanatory diagram showing an example of an attention condition designation screen.

FIG. 10 is an explanatory diagram showing an example of an attention condition designation screen after changing an inquiry condition.

FIG. 11 is an explanatory diagram showing the attention satisfaction level for the data selected in the first search.

FIG. 12 is an explanatory diagram showing an example of a search result output screen in which the search result based on the attention condition is displayed together with the search result based on the first inquiry condition.

FIG. 13 is a flowchart showing a flow of processing when the information search device according to the embodiment of the present invention searches for data.

FIG. 14 is a block diagram functionally showing a configuration of an information search device in a modified example.

FIG. 15 is a block diagram showing the configuration of an information search device that searches information according to ambiguous search conditions that have been conventionally used.

[Explanation of symbols]

11 ... CPU, 12 ... Bus, 13 ... ROM, 14 ... RA
M, 15 ... Magnetic disk device, 16 ... Keyboard, 17
... mouse, 18 ... display device, 21 ... inquiry condition input means, 22 ... database, 23 ... coincidence degree calculation means, 2
4 ... Satisfaction calculating means, 25 ... Search result extracting means, 26 ...
Search result output means, 27 ... Attention condition designating means, 28 ... Attention satisfaction calculating means

Claims (5)

[Claims] 1. A data storage means for storing a plurality of data to be searched, a search condition designating means for designating an arbitrary number of ambiguous search conditions, and a search condition designated by the search condition designating means. And a satisfaction degree calculation means for obtaining a satisfaction degree representing a degree of satisfaction for each data stored in the data storage means, and a satisfaction degree calculation means for storing the satisfaction degree calculated by the satisfaction degree calculation means in the data storage means. Data selecting means for selecting a predetermined number of data from the existing data, search condition changing means for changing the search condition specified by the search condition specifying means, and data selected by the data selecting means. Change satisfaction degree calculating means for obtaining the satisfaction degree of the search condition after being changed by the search condition changing means. That the information retrieval apparatus. 2. A data storage means for storing a plurality of data to be searched, a search condition designating means for designating an arbitrary number of ambiguous search conditions, and a search condition designated by the search condition designating means. And a satisfaction degree calculation means for obtaining a satisfaction degree representing a degree of satisfaction for each data stored in the data storage means, and a satisfaction degree calculation means for storing the satisfaction degree calculated by the satisfaction degree calculation means in the data storage means. Data selecting means for selecting a predetermined number of data from the existing data, search condition changing means for changing the search condition specified by the search condition specifying means, and data selected by the data selecting means. A change satisfaction calculating means for obtaining a satisfaction degree of the search condition after being changed by the search condition changing means, and the data selecting means. Therefore, the information retrieval is characterized by including display means for displaying the satisfaction degree calculated by the satisfaction degree calculation means and the satisfaction degree calculated by the change satisfaction degree calculation means for a predetermined number of selected data. apparatus. 3. A data storage means for storing a plurality of data to be searched, a search condition designating means for designating an arbitrary number of ambiguous search conditions, and a search condition designated by the search condition designating means. And a satisfaction degree calculation means for obtaining a satisfaction degree representing a degree of satisfaction for each data stored in the data storage means, and a satisfaction degree calculation means for storing the satisfaction degree calculated by the satisfaction degree calculation means in the data storage means. Data selecting means for selecting a predetermined number of data from the existing data, search condition changing means for changing the search condition specified by the search condition specifying means, and data selected by the data selecting means. Change satisfaction calculation means for obtaining satisfaction of the search condition after being changed by the search condition changing means, and rank of satisfaction degree Display means for displaying together the satisfaction degree calculated by the satisfaction degree calculation means and the satisfaction degree calculated by the change satisfaction degree calculation means for a predetermined number of data selected by the data selection means in different display types for each An information retrieval apparatus comprising: 4. A data storage means for storing a plurality of data to be searched, a search condition specifying means for specifying a compound search condition combining a plurality of ambiguous search conditions, and a search condition specifying means. About the coincidence degree calculating means for obtaining the degree of coincidence representing the degree of coincidence with each designated ambiguous search condition for each data stored in the data storage means, and the ambiguous search condition obtained by this coincidence degree calculating means Satisfaction calculating means for obtaining, for each data stored in the data storage means, a satisfaction degree indicating a degree of satisfying the composite search condition based on the degree of coincidence of the satisfaction degree calculated by the satisfaction degree calculating means. Data selection means for selecting a predetermined number of data from the data stored in the data storage means based on A matching degree storage means for storing the matching degree for the data selected by the data selecting means among the matching degrees obtained by the matching degree calculating means; and a plurality of ambiguous search conditions specified by the search condition specifying means. Among the other, select an arbitrary number of ambiguous search conditions that are more noticeable than others, and specify a noticeable search condition designating unit that specifies a noticeable compound search condition, and a match stored in the matching degree storage unit. Information including a notice satisfaction degree calculating means for obtaining a satisfaction degree in the notice composite search condition designated by the notice search condition specifying means for the data selected by the data selecting means based on the degree. Search device. 5. A data storage means for storing a plurality of data to be searched, and a search condition designating means for weighting a plurality of ambiguous search conditions and designating a composite search condition combining these. The degree of coincidence representing the degree to which each ambiguous search condition designated by the search condition designating means is obtained for each data stored in the data storage means, and the degree of coincidence calculating means First satisfaction degree calculating means for weighting the degree of coincidence with respect to the ambiguous search condition to obtain a satisfaction degree representing a degree of satisfying the composite search condition for each data stored in the data storage means, The data stored in the data storage means is based on the satisfaction degree calculated by the first satisfaction degree calculation means. Data selecting means for selecting a predetermined number of data from among, and a matching degree storing means for storing the matching degree for the data selected by the data selecting means among the matching degrees obtained by the matching degree calculating means, Weighting changing means for changing the weighting specified by the search condition specifying means, and weighting after the matching degree stored in the matching degree storing means after being changed by the weighting changing means is applied to the composite search condition. And a second satisfaction degree calculating means for obtaining the satisfaction degree of the data selected by the data selecting means.