KR101059032B1 - Search Schema Setting Device and Method - Google Patents

Abstract

The present invention relates to an apparatus and method for setting a search schema, comprising the steps of: a client connecting to a server to define a data schema including a plurality of fields; The server recommending a search schema setting value to the client for each field of the defined data schema; Transmitting, by the client, data corresponding to the data schema to the server; Generating an inverted index using the data and the search schema received by the server; And providing, by the server, a search service for the received data by using an inverse index and a search schema.

Data Schema, Search Schema, Inverted Index, Search Engine

Description

Search Scheme Setup Device and Method {APPARATUS AND METHOD OF CONFIGURING SEARCH SCHEMA}

The present invention relates to a search schema setting apparatus and method for recommending a search schema for a customized search service.

A search engine is a system that collects and finds information on the Internet in a broad sense. Nowadays, mainly crawling web pages on the Internet and receiving a specific query, the web pages related to the query are displayed as a result. Refers to the system.

On the other hand, in addition to a search engine for end users such as Google (www.google.com), web sites that have a lot of contents need their own search engines or search services.

Of course, the basic search function, such as searching for posts containing a specific search word for the posts registered on the bulletin board, is widely used, but the characteristics of the search word mainly input to large web sites with various data types In addition, the necessity of building a customized search engine in accordance with the expected search results was raised.

In the past, it was common for experts to analyze the web site and define what data they had, and then develop it individually. However, this is not only very inefficient, but also has a lot of deviations in terms of technical perfection, thereby automating the search engine customization process. Process development was urgently needed.

However, if website operators who do not have expertise in the structure of search engines can directly set the data schema and search schema for customizing the search engine, they may face limitations that reduce user convenience and accessibility. There was a risk.

An object of the present invention is to provide a search schema setting apparatus and method for improving user convenience by recommending an appropriate search schema after setting up a data schema for a customized search service. .

The search schema setting apparatus of the present invention for achieving the above object is an interface unit for receiving a data schema including a plurality of fields from the client via the Internet network, and receives the data;

A search schema analyzer for recommending a search schema for each field of the data schema;

An index generator for generating an inverse index from the received data using a search schema;

And a search unit for searching the received data using the inverted index and the search schema.

On the other hand, the search schema setting method of the present invention for achieving the above object comprises the steps of a client defining a data schema including a plurality of fields connected to the server;

Recommending, by the server, a search schema setting value to the client for each field of the defined data schema;

Transmitting, by the client, data corresponding to the data schema to the server;

Generating an inverted index using the data and the search schema received by the server; And

The server provides a search service for the received data using an inverted index and a search schema.

On the other hand, the search schema setting method of the present invention for achieving the above object comprises the steps of recommending a search schema setting value for each field of the data schema input by the client;

Generating an inverted index using the search schema when data corresponding to the data schema is received from the client;

Performing a search on the received data using an inverted index when a search word is input from a client; And

And reconstructing and returning a search result using the search schema.

According to the present invention as described above, there is an excellent effect that a customized search engine can be obtained with only a few settings for the characteristics of the data held and the characteristics of the search results to be obtained without having to spend a huge development cost and a long development period.

Furthermore, even if the user does not have an expert knowledge about the structure of the search engine, it is possible to greatly improve user convenience and accessibility by recommending the appropriate search schema setting value.

Hereinafter, a configuration of a search schema setting apparatus according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a functional block diagram conceptually showing a configuration of a search schema setting apparatus according to the present invention, and FIG. 2 is a reference diagram conceptually illustrating a data schema and a search schema. 3 and 4 are reference diagrams for describing a search schema preference analysis process for each field of the data schema.

Referring to FIG. 1, the search schema setting apparatus 100 of the present invention includes an interface unit 110, a search schema analysis unit 120, an index generator 130, and a searcher 140.

The interface unit 110 receives a data schema including a plurality of fields from the client 200 through the Internet network, and receives data corresponding to the received schema. In this case, the received data is raw data that is a target of a search service, as described below.

On the other hand, the search schema analysis unit 120 analyzes the search schema preference for each field by using a plurality of pre-collected data as described below, and the search schema value for each field of the input data schema, the client 200 Recommended)

In this case, as shown in FIG. 2, a data schema is defined for a customized search service for each client 200 and includes a plurality of fields.

The data schema exemplarily shown in FIG. 2 has five fields of "URL", "ID", "date", "title", and "content", and each of these fields is "text" or "date". Has the property In the example of FIG. 2, it can be seen that "URL", "content", and "title" are defined as "variable length text" and "ID" as "fixed length text".

On the other hand, the Search Schema checks whether or not the field value is a search target through the search engine, sorts the search results based on the field value, and the field value for each field of the data schema defined in this way. Whether to filter the search results based on the criteria, and whether to display the corresponding field value on the screen of the client 200 as a search result.

According to the example of FIG. 2, the "Title" field is a search target, but not a sort target, and is not a filtering target, but a display target. The "ID" field is a search target, a display target, and a sorting target and a filtering target. It is not known.

Since the values belonging to the "Title" field among the data transmitted from the client 200 are search targets, the inverse index is generated as described below. However, the search results are not sorted based on the "ID" field value, and filtering is not performed such as excluding a result value having or not having a specific "ID" value. When displaying a search result on the screen, it means that the "ID" field value is displayed.

Meanwhile, the search schema may further include an analysis method for a search target field. According to the example of FIG. 2, the analysis method for the search target field is “morphological analysis” using a morpheme analyzer, “bigram analysis” by dividing a field value into two syllables, or “no analysis” using a field value as a search target. There are three of ".

As shown in FIG. 2, when the client 200 inputs a data schema having five fields of "URL", "ID", "date", "title", and "content", the search schema analyzer 120 ) Recommends values for whether to search, sort, filter, or display each field.

Furthermore, an analysis method may be recommended for the search target field, in which the analysis method may have any one of a morphological analysis, a bigram analysis, and no analysis.

In the example of FIG. 2, the search schema analysis unit 120 may search for "subject" and "display" for the "title" field input by the client 200, and "FALSE" for sorting and filtering. It can be seen that the recommended value of ". (In the example of FIG. 2, "O" and "X" are represented, but it may be understood to mean "TRUE" and "FALSE", respectively.)

If so, referring to FIGS. 3 and 4, the search schema analysis unit 120 goes through a process to recommend a search schema value.

3 illustrates a process of analyzing a search schema preference from a plurality of clients for each field. Referring to FIG. 3, the search schema analyzer 120 first analyzes data schemas of a plurality of pre-collected clients, and compares field names and attributes to separate fields recognized as the same or highly related.

In the example of FIG. 3, a plurality of clients include fields "title", "title", and "title" in the data schema, and since all of them have an attribute of "variable length text", the search schema analyzer 120 The conclusion is that these are the same or related fields.

Then, preference can be obtained by analyzing what search schema attributes these same or related fields have. In the example of FIG. 3, 70% of the fields titled "Title", "Title", and "title" from the data schemas of a plurality of clients are set to be searched, and 30% of the cases are not. If it is shown as the search schema analysis unit 120 will recommend the "TRUE" value related to the search object for the "title" field of the data schema input by the client 200. This is because preference is 70% higher than "FALSE".

Similarly, we recommend "FALSE" with a 65% preference for sorting, a recommendation of "FALSE" with 90% of preference for filtering, and "FALSE" with a 55% preference for display. Recommended.

Meanwhile, in the example of FIG. 4, the preference is analyzed by a different method. First, the search schema analyzer 120 compares the names and attributes of fields based on collected data, and is recognized as having the same or higher relevance. The same is true of separating fields.

However, group information is classified using the collected information of a plurality of riders and the preferences related to each item of the search schema are collected for each group. The process of collecting the preference for each item is as described in the example of FIG. 3.

In the example of Figure 4, the client 200 has entered the personal information "twenties, office workers, men", the search schema analysis unit 120 using a number of collected data schemas and search schemas It analyzes the preferences of search schema items of the group to which the client 200 belongs and recommends them.

The personal information of the client 200 may be collected through a separate membership registration process or received from the client 200 through various other processes, and a specific collection process is excluded.

When the search schema analysis unit 120 recommends a high preference value for each search schema item through the above process, the client 200 may accept it or change it to another value. At this time, if the client 200 accepts it, the recommended value is stored, and if the client 200 changes, the changed value is stored and the stored search schema is used for a customized search service for the client 200 as described below. .

Meanwhile, the index generator 130 generates an inverse index by analyzing the data received from the client 200. Inverted Index is information that stores where a specific index is located. The search engine uses this inverted index to search only the search results that contain the index. You can quickly extract it and return it as a search result.

The index generator 130 generates an inverted index for a field to be searched by referring to a search schema among the fields of the received data.

In the example of FIG. 2, since the “URL”, “ID”, “Title”, and “Content” fields correspond to a search object according to the search schema defined by the client 200, the index generator 130 receives the request from the client. Inverse indexes are generated by analyzing these field values against the data.

At this time, these field values are analyzed using an analysis method defined by the search schema. According to the example of FIG. 2, since the "Title" and "Content" fields are defined to perform morphological analysis, the corresponding field values are analyzed using the morpheme analyzer, but the "URL" and "ID" fields are not analyzed and the field values are kept as they are. Make it a search target.

When the searcher 140 receives a search word from the client 200, the search unit 140 performs a search using the generated inverted index using the search word, and sorts, filters, and displays search results using the search schema. The determination is made and returned to the client 200.

Hereinafter, a method of recommending and setting a search schema in the search schema setting apparatus having the above configuration will be described with reference to FIG. 5. 5 is a flowchart illustrating a method of setting a search schema according to the present invention in time series.

Referring to FIG. 5, the client 200 first accesses the search schema setting device schema setting device 100 according to the present invention through an internet network to define a data schema (S110). In this case, the data schema includes a plurality of fields as shown in FIG. 2, and preferably may further include attributes of each field.

On the other hand, when the data schema is defined, the schema setting device 100 receiving the input recommends a search schema for each field of the data schema (S120). As shown in the example of FIG. 2, the search schema analyzing unit 120 analyzes the collected data and recommends appropriate values with respect to whether to search, sort, filter, or display each field. . In the example of Figure 2 is expressed as having only two values of O or X, but this is only exemplary and may have a variety of values depending on the implementation. In the example of FIG. 2, the schema setting device 100 recommends a value of "O" as a value of "TRUE" as a value for sorting object for the "ID" and "date of creation" fields. However, if you first sort by the "ID" field, then sort by the "Date" field, then "1" as the recommendation for the "ID" field and the recommendation for the "Date" field. "2" may be recommended.

On the other hand, the search schema analysis unit 120 examines the preferences by analyzing data schemas, search schemas, and personal information of a plurality of previously analyzed clients, and recommends an appropriate value to the client 200 using the result of the preferences survey. do.

Depending on the implementation, the schema setting device 100 may automatically determine the recommended value as the search schema of the client 200, or if the client 200 confirms, the recommended value may be used as the search schema. You might be given permission to change it to a value.

Thereafter, when the client 200 transmits data corresponding to the data schema to the server (S130), the schema setting apparatus 100 generates an inverse index using the received data and the search schema (S140).

The schema setting apparatus 100 provides the client 200 with a search service for the received data by using the generated inverted index and the search schema.

Meanwhile, referring to FIG. 6, a method of recommending and setting a search schema in a search schema setting apparatus having the above configuration will be described in terms of a server. 6 is a flowchart illustrating a method of setting a search schema according to the present invention in time series.

First, the search schema setting value is recommended for each field of the data schema input by the client 200 (S210). When the data schema is input as illustrated in FIG. 2, the process is performed by presetting values for sorting, searching, filtering, and display for each field, and recommending them to the client 200. .

In addition, an analysis method may be further recommended for the search target field, and preferably, the analysis method may have any one value of morphological analysis, bigram analysis, or no analysis.

Depending on the implementation, the recommended value may be determined as the search schema of the client 200 as it is. In addition, if the client 200 confirms the recommended search schema setting value, the client may change the recommended search schema setting value. If you change the setting, the changed setting can be confirmed by the search schema.

On the other hand, the process of setting the recommended value for each field in advance through the process in detail, as shown in Figure 3, the method recommended by referring only to the data and the defined data schema retained by the client 200, and Figure 4 As shown in the figure, there may be a method of recommending with reference to the personal information of the client 200.

According to the example of FIG. 3, the search schema analyzer 120 first analyzes a plurality of collected client-specific data schemas and a search schema to distinguish relevant fields. In the example of FIG. 3, it can be seen that fields having field names of "title", "title", and "title" are divided into identical or at least highly related fields.

Thereafter, other clients examine the preferences of how the search schema is set, and recommend appropriate values according to the preferences. In the example of FIG. 3, about 70% of the “title” fields are set to a value of “TRUE”. Therefore, a value of “TRUE” having high preference is recommended.

On the other hand, according to the example of Figure 4 by referring to the personal information of the client 200 to further refine the recommendation. The personal information of a number of other collected clients is divided into a plurality of groups by each item of personal information, and each group is examined for preference as described in the example of FIG.

Thereafter, the client 200 determines which group the user 200 belongs to based on the personal information and recommends an appropriate value using the preference of the group.

When the search schema is determined through the recommendation process of the search schema, the client 200 transmits data corresponding to the data schema to the schema setting apparatus 100. When the schema setting apparatus 100 receives the data, the schema setting apparatus 100 generates an inverted index using the search schema (S220). Specifically, an inverse index for searching is generated for fields set as a search target among the set search schemas.

Subsequently, when a search word is input from the client 200, a search for the received data is performed using an inverse index (S230).

Meanwhile, the search results are reconstructed using the search schema. The reconstruction of the search results refers to filtering and sorting and deleting field values that are not displayed on the screen of the client 200. When the reconstruction is completed, the reconstruction of the search results is returned to the client 200 (S240).

Although the present invention has been described in detail with reference to several embodiments, the present invention is not limited to these embodiments and should not be interpreted, but should be freely modified and interpreted within the scope of the technical idea described in the claims.

1 is a functional block diagram conceptually showing the configuration of a search schema setting apparatus according to the present invention;

2 is a reference diagram conceptually illustrating a data schema and a search schema,

3 is a reference diagram illustrating an example of a preference analysis process of a search schema for each field.

4 is a reference diagram for explaining a process of analyzing preferences of personal information items by search schema by way of example.

5 is a flowchart illustrating a method of setting up a search schema according to the present invention in time series.

6 is a flowchart illustrating a method of setting a search schema according to the present invention in time series from the server's point of view.

<Explanation of symbols for the main parts of the drawings>

110: interface unit 120: search schema analysis unit

130: index generator 140: search unit

Claims (21)

Connecting the client to the server to define a data schema including a plurality of fields; The server recommending a search schema setting value to the client for each field of the defined data schema; Transmitting, by the client, data corresponding to the data schema to the server; Generating an inverted index using the data and the search schema received by the server; And And a server providing a search service for the received data by using an inverse index and a search schema. The method of claim 1, The recommendation of the search schema setting value is a search schema setting method, characterized in that is made by recommending the values for whether to sort, search, filter, or display for each field. 3. The method of claim 2, A search schema setting method, characterized in that the method for recommending an analysis method for a field recommended as a search object when recommending a search schema setting value is further recommended. The method of claim 3, The analysis method is a search schema setting method characterized in that it has any one of a morphological analysis, a bigram analysis, no analysis. 3. The method of claim 2, Analyzing the data schemas and search schemas of a plurality of collected clients and examining preferences for sorting, searching, filtering, and display for each field of the data schema; The search schema setting method of claim 1, wherein values for sorting, searching, filtering, and displaying are recommended for each field of the data schema using the preference. 3. The method of claim 2, The client is connected to the server to enter the personal information; further comprising, Analyzing data schemas, search schemas, and personal information of multiple collected clients, values for sorting, searching, filtering, and displaying each field of the data schema based on the client's personal information Search schema setting method, characterized in that for recommending. Recommending a search schema setting value for each field of the data schema input by the client; Generating an inverted index using the search schema when data corresponding to the data schema is received from the client; Performing a search on the received data using an inverted index when a search word is input from a client; And And reconstructing and returning a search result by using the search schema. The method of claim 7, wherein The step of recommending the search schema setting value is a search schema setting method characterized in that it is made by recommending by setting in advance the values for whether to sort, whether to search, filtering or display for each field. The method of claim 8, A search schema setting method characterized by further recommending an analysis method for a field recommended as a search target. The method of claim 9, The analysis method is a search schema setting method characterized in that it has any one of a morphological analysis, a bigram analysis, no analysis. The method of claim 7, wherein And if the client confirms the recommended search schema setting value, storing the changed setting value if the client changes the recommended search schema setting value to another setting. The method of claim 7, wherein Analyzing related schemas and a plurality of other client-specific data schemas to distinguish relevant fields; And And analyzing preferences of sorted fields, search targets, filtering, and display. The method of claim 12, And searching for, sorting, filtering, and displaying sorted fields for each field of the data schema input by the client using the analyzed preferences. The method of claim 12, The client is connected to the server to input the personal information; Search schema setting method further comprising the. 15. The method of claim 14, Analyzing personal information of a number of other collected clients to divide into a plurality of groups, analyzing the search schema preference for each group; The search schema setting method of claim 1, wherein the search schema preference of the group to which the client belongs is recommended for each field of the data schema input by the client. An interface unit for receiving a data schema including a plurality of fields from a client and transmitting data; A search schema analyzer for recommending a search schema for each field of the data schema; An index generator for generating an inverse index from the received data using a search schema; And a search unit for searching the received data using the inverted index and the search schema. The method of claim 16, The search schema analysis unit, characterized in that for each field of the data schema search schema setting device, characterized in that for recommending the value of whether to sort, whether to search, filtering or display. The method of claim 17, The search schema analyzing unit further recommends an analysis method for a field recommended as a search target. The method of claim 18, The analysis method is a search schema setting apparatus, characterized in that it has any one of a morphological analysis, a bigram analysis, no analysis. The method of claim 17, The search schema analysis unit analyzes data schemas and search schemas of a plurality of collected clients and analyzes preferences for each field to recommend sorting, searching, filtering, and display. Setting device. The method of claim 20, The search schema analyzing unit analyzes schema preference for each item of personal information by using the collected personal information of a plurality of other clients, and sorts, searches, filters, and displays each field according to the personal information of the client. Search schema set up device, characterized in that for recommending.

Images