JPH0887509A - Information presentation method - Google Patents

Abstract

(57) [Summary] [Purpose] Information presentation that allows users to search for items based on viewpoints that are not explicitly specified by the user and present solutions and information that may be more appropriate for the user. The purpose is to provide a method. According to an information presenting method of the present invention, a step of inputting a plurality of search condition items and a step of inputting a plurality of search condition items in the information presenting method for presenting a search result by performing an information search from accumulated information At least one of
Of at least one search condition item related to one of the search conditions from the database in which the relationship between the search conditions is stored as knowledge, and adding to the input search condition item; and information based on these search condition items. Performing a search and presenting a search result.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of presenting an optimum solution and information to a user in a computer system for solving problems and retrieving information.

[0002]

2. Description of the Related Art In computer systems for problem solving and information retrieval, how to effectively extract information and solutions suitable for the user from a large amount of information and solutions and to provide them to the user is a problem. Conventionally, in order to narrow down such information and solutions, the user selects and specifies search condition items (keywords) that match each search request on the system, inputs the range of the search conditions, and further It was done by specifying the priority.

However, when a user considers a search condition item that matches a search request, even if the user considers the search condition item to be important or effective, the item based on the viewpoint that the user cannot easily recognize, the prejudice or the prejudice There may be items based on a point of view that is consciously or unconsciously avoided from experience rules. Therefore, with the above-mentioned specification method, although it is possible to make a proposal that reflects the user's wishes regarding items based on the viewpoint that the user is strongly aware of, items based on the viewpoint that the user is not very aware of. There is a problem that the information and the solution about are often overlooked, and the best suggestion for the user is not always made.

[0004]

As described above, in the conventional information presentation method, the search condition items (keywords) used to search for information and solutions are not explicitly specified by the user. If the item (particularly, the item that the user did not specify because the user did not notice it) has the importance or effectiveness, it is not always possible to obtain the best information or solution for the user.

The present invention has been made in consideration of the above circumstances, and searches for an item based on a point of view that is not explicitly specified by the user, and a solution that may be more appropriate for the user. It is an object of the present invention to provide an information presentation method capable of presenting information.

[0006]

According to the present invention, in an information presenting method for presenting a search result by performing an information search from accumulated information, a step of inputting a plurality of search condition items and the input search At least one search condition item related to at least one of the condition items is stored in a database that stores the relationship between the search conditions as knowledge.
It is characterized by including a step of extracting one and adding it to the input search condition item, and a step of performing an information search based on these search condition items and presenting a search result. Further, it is preferable that the method further comprises the step of presenting a basis for adopting the search condition item extracted from the database.

[0007]

According to the present invention, a database is provided which stores the relationship between the search conditions as knowledge, and at least one search condition item related to the search condition item designated by the user is extracted from this database, and this extraction is performed. An information search is performed based on the search condition items thus specified and the search condition items specified by the user, and the search results are presented.

As a result, it is possible to present a proposal different from the proposal created based on the weight of the value (search condition item) designated in advance by the user. Then, it is possible to make an appropriate proposal to the user from the viewpoint that the user does not notice or from a wider viewpoint.

[0009] Further, it is possible to bring out the potential values of the user, remove preconceptions, or eliminate the bias of proposals. Furthermore, it is possible to make an appropriate proposal to the user to some extent without knowing enough items for the user to specify, or to save the user the trouble of re-inputting items by trial and error. There is.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic block diagram of an information presentation system according to an embodiment of the present invention.

The dialogue section 1 is for the system to control dialogue with the user. The dialogue unit 1 is connected to an input unit 2 for a user to input his / her request to the system, an output unit 3 for displaying an inquiry or a search result to the user, and a dialogue history storage unit 4 for storing a history of dialogue. ing.

The control unit 5 controls the operation of the entire system, and the dialog unit 1, the retrieval formula inference unit 7, and the retrieval processing unit 8 are connected to the control unit 5 via the bus 6. A target knowledge base 9 used for inference is connected to the search expression inference unit 7, and a database 10 to be searched is connected to the search processing unit 8.
Is connected.

In addition to the priority and the weight of value explicitly specified by the user, the retrieval formula inference unit 7 refers to the relevance such as input conditions for retrieval and keywords when generating the retrieval formula. However, the target knowledge base 9 describes the knowledge about the mutual association of the description elements that can be used for this condition or keyword.

The search processing unit 8 searches the database 10 for information to be actually presented based on the search formula generated by the search formula inference unit 7. Hereinafter, the information presentation system having the above configuration will be described more specifically by taking a real estate search system as an example.

FIG. 2 illustrates an example of a dialogue screen in such an information presentation system. In this example, the menu format and the free input format are adopted. Figure 2
, 20 is a question sentence that the system gives to the user,
Reference numeral 21 indicates a selected classification name, and 22 and 23 indicate selection items. The user inputs a desired selection item for each selection category by a method such as selecting with a mouse. Here, the item selected by the user is colored as 22 and the unselected item is displayed as white as 23. Reference numeral 24 is a free description field in which the user inputs a character string or the like. Here, the situation of the user who prescribes the search, such as the family structure and work place, is mainly described. The situation content described here is stored in the dialogue history storage unit 4.

If there is a further selection category for the selected selection item, a submenu such as 25 may be displayed and the selection item may be input in the same manner. This system searches and presents information suitable for the user from the database based on the user's input obtained through such a dialogue. Since this example is a real estate search system, the presentation result presents detailed information on the property contents, floor plans, etc. using character strings and images.

FIG. 3, FIG. 4 and FIG. 5 are diagrams showing examples of search expressions in such a search system. In these search expressions, 31 is a search command, and the keyword "select" is used in this example. 32
Is the name of the data to be searched, and in this case indicates that the property data in the real estate search system is searched. The search expression can be defined by a logical expression, and 33 indicates an and expression which is the keyword. In addition to the and expression, an or expression, a not expression, etc. can be described. 34 is a variable indicating a search condition item, 35 is a search condition that the search condition item must satisfy, and 36 is a numerical value indicating the weight of this search condition item. The search process is designed so that the search condition item having a higher weight is preferentially searched. Also,
Reference numeral 37 is a description when there are a plurality of value candidates to be satisfied by the search condition item, and in this case, the keyword "&some" is used.

In FIG. 3, the weight of all search condition items is 0.
And when no weighting is done,
The search process is performed equally for all search items.
The search formula is "The price is 70 million or less, and the size is 70
~ 90 square meters, a new type of detached house, along the Nambu Line,
Please search for a property that meets the conditions that the nearest station is Musashikosugi or MusashiMizonokuchi and that shopping is convenient near the station. "
It means that.

FIG. 4 is a description in which each search condition item in FIG. 3 is weighted. Here, the weight is represented by a numerical value of 0 to 10. The price and size are weighted with 10, which is an absolute requirement.

FIG. 5 is a description in which the search condition items are weighted as in the case of FIG. 4, and the search condition items are rearranged by the rearrangement operation of the search conditions described later. Next, FIG. 6 conceptually shows an example of the description of the target knowledge base 9 that is referred to when a search expression is inferred and generated in such a search system. This target knowledge base is for determining a search condition according to the degree of relevance, and in FIG. 6, 41 represents a concept that can be a search condition, and 42 represents a group of related concepts. For example, concepts that are common in terms of convenience, such as “close to the station”, “shopping convenience”, and “commuting convenience” are grouped and described. Reference numeral 43 represents an element that defines a concept, and 44 is a link that represents a relationship of which concept is a condition defining condition.

The contents of the target knowledge base 9 may be fixedly set or may be appropriately edited such as addition / deletion / change. Next, FIG. 7 shows a schematic process flow of the system performed by the control unit 5.

First, the dialogue unit 1 causes the user to input a selection item by using the inquiry interface as shown in FIG. 2, and the search formula inference unit 7 refers to the selection item and first searches without weight setting. An expression is generated, and the search processing unit 8 executes search processing based on the search expression (step S5).
1). At this stage, the search formula generated by referring to the selection items input in the example of FIG. 2 is as shown in FIG.

As a result of executing the search process, a solution is obtained, and when the user is satisfied with the solution as a result of presenting the obtained solution to the user through the dialogue unit 1 (step S52), the process is terminated. If not, change the search expression as follows and perform the next search.

First, with the search condition items as they are, a weighting operation is performed on each search condition item to generate a new search expression, and a search process is executed (step S53). As a procedure of the weight operation at this time, for example, a method of increasing the weight value of the search condition item which is an absolute constraint condition such as price or size and decreasing the weight of other search condition items is used. is there. The retrieval formula obtained as a result is as shown in FIG.

As a result of executing the search process, a solution is obtained, and when the user is satisfied with the solution as a result of presenting the obtained solution to the user through the dialogue unit 1 (step S54), the process is terminated. If not, change the search expression as follows and perform the next search.

Next, the search condition is recombined, a new search expression is generated, and the search process is executed (step S).
55). As a result, the process ends if a solution satisfying the user is obtained (step S56), but if not, a search condition rearrangement operation is performed again, a search process is executed using a new search expression, and a solution is obtained. Or this is repeated until the user interrupts.

Next, FIG. 8 shows a processing flow of the search condition rearrangement operation shown in step S55 of FIG. First, an operable search condition item whose weight is not 10 is selected (step S61). In the example of FIG. 4, other than price and size are selected. Next, referring to the target knowledge base 9, it is inspected whether or not there is an element belonging to the same concept group in the search condition (step S62), and if there is such an element, the one with the highest weight is left, Other search conditions are deleted from the search formula (step S63). At this time, if all have the same weight, the one specified by the user's situation content stored in the dialogue history storage unit 4 is left, or if not, it is randomly selected and one is left. FIG.
In the example, “close to the station” and “shopping convenience” belong to the same concept group and have the same weight, so “shopping convenience” is deleted and the result is as shown in FIG.

Next, a search condition other than the search conditions already described in the search formula is selected, and the search condition is rearranged. As one of the selection methods, the target knowledge base is referred to, whether or not there is an element defined by the user's situation content is checked (step S64), and if there is, it is selected. Further, in that case, it is checked whether or not the element belongs to the same concept group as the element which is already the search condition (step S6).
5) If not, that element is selected as a search condition and weighted (step S66). As one of the weighting strategies, for example, a method of increasing the weight over the search condition corresponding to the item selected by the user, aiming at unexpectedness in order to remove the user's preconceptions, can be considered. Figure 2
From the example of dialogue, the user situation is that the couple work together,
Since it is known that there are children, referring to the target knowledge base in FIG. 6, “good sun exposure” is selected as a search condition from the elements defined by the element “child”, and the weight is set to “Station”. Set to higher than “close to” and rearrange the formula of FIG. 4 into the formula of FIG.

On the other hand, if there is no element that matches the user's situation contents in step S64, and step S64
If it belongs to the same concept group as the element already selected as the search condition in 65, select an element that does not belong to the same concept group as the element already the search condition,
Weighting is performed as a new search condition (step S6).
7). As one of the element selection methods at this time, for example, a method of randomly selecting elements that do not belong to the same concept group as the element that is already the search condition can be considered.

A series of these processes is repeated as shown in steps S55 and S56 of the flowchart of FIG. At that time, a method of selecting an element that has not been selected as a search condition in the past after the process is repeated several times by storing the search expression and the result thereof in the dialogue history storage unit 4 is also conceivable. .

Next, FIG. 9 shows a flow of processing for explaining to the user the basis for selecting the information retrieved as a result of the reassembling operation of the retrieval conditions. First, the retrieval condition selected as a result of performing the rearrangement operation is retrieved from the dialogue history storage unit 4 (step S71). If there are multiple search conditions, multiple search conditions are retrieved. Next, with reference to the target knowledge base 9, it is checked whether or not there is a concept that defines the conceptual element that has become the search condition, and it is further checked whether or not there is a content of the user situation that matches this (step S72). If there is this, the concept is extracted (step S73).
In the example of FIG. 5, “good sun exposure” is selected as a new search condition, but the “child” defining this is retrieved with reference to the target knowledge base 9. Finally, an explanatory note is generated using the search condition and the defined concept thus extracted (step S74). For example, in the case of this example, the explanation sentence "I have a child, so I will propose a property of XX as a result of considering the conditions of good sunlight" is presented.

As described above, in this embodiment,
It becomes possible to present a proposal different from the proposal created based on the weight of the value designated in advance by the user. Then, it is possible to make an appropriate proposal to the user from the viewpoint that the user does not notice or from a wider viewpoint.

Further, it is possible to bring out the potential values of the user, remove preconceptions, or eliminate the bias of proposals. Furthermore, it is possible to make an appropriate proposal to the user to some extent without knowing enough items for the user to specify, or to save the user the trouble of re-inputting items by trial and error. There is.

Further, as described above, since the mutual knowledge of the concepts that can be the search condition and the relationship between the concept and the context contents defining the concept are described in the target knowledge base, refer to this and rearrange the search conditions. The rationale for the retrieval of the information presented as a result of the operation can be explained using the relationship between the concepts of the target knowledge base. The present invention is not limited to the above-described embodiments, and various modifications can be carried out without departing from the scope of the invention.

[0035]

As described above, according to the present invention, it is possible to make a proposal created based on a value weight previously designated by the user, and at the same time, make a proposal based on a value weight not explicitly designated by the user. , It is possible to make an appropriate proposal from the viewpoint that the user is not aware of.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of an embodiment of the present invention.

FIG. 2 is a diagram for explaining an example of an interactive screen used in the same embodiment.

FIG. 3 is a diagram for explaining an example of a search formula used in the same embodiment.

FIG. 4 is a diagram for explaining another example of the search formula used in the embodiment.

FIG. 5 is a view for explaining still another example of the search formula used in the same embodiment.

FIG. 6 is a conceptual diagram for explaining one expression example of a target knowledge base used in the embodiment.

FIG. 7 is a flowchart showing a processing flow of a control unit of the embodiment.

FIG. 8 is a flowchart showing a processing flow of a search condition rearrangement operation.

FIG. 9 is a flowchart showing the flow of processing for explaining the reason for information selection to the user.

[Explanation of symbols]

1 ... Dialogue unit, 2 ... Input unit, 3 ... Output unit, 4 ... Dialogue history storage unit, 5 ... Control unit, 6 ... Bus, 7 ... Search expression inference unit, 8 ...
Search processing unit, 9 ... Target knowledge base, 10 ... Database

Claims (2)

[Claims] 1. An information presenting method for presenting a search result by performing an information search from accumulated information, and a step of inputting a plurality of search condition items, and at least one of the input search condition items. Extracting at least one search condition item related to the above from the database storing the relevance between each search condition as knowledge, and adding to the input search condition item; and information search based on these search condition items. And presenting a search result. 2. The information presenting method, further comprising the step of presenting a basis for adopting the search condition item extracted from the database.