KR101515413B1 - Professional field search supporting method and apparatus - Google Patents

Abstract

A method and apparatus for specialized field search support are disclosed. A method for supporting a specialized field search includes the steps of (a) collecting query-answer data in a web document; (b) extracting words by analyzing the question part and the answer part in the question-answer data; (c) generating a general term-terminology mapping table by calculating a general term-terminology mapping probability by analyzing correlation between words included in the question portion and the answer portion; And (d) extracting and providing a terminology including a word included in the query using the term mapping table.

Description

Field of the Invention < RTI ID = 0.0 > [0001] < / RTI &

The present invention relates to a method and apparatus for learning a mapping probability between a general term, a technical term, and a statute by using Q & A data or precedent data collected from a web site and then using the same to predict a technical term or statute for the query term Field search support method and apparatus thereof.

It is very difficult for the general public who does not have prior knowledge of the specialization field to search for and utilize knowledge in various specialties. For example, it is almost impossible for a general person without medical knowledge to search for and understand medical knowledge about his or her body condition. A general person without legal knowledge can find out the laws and regulations applicable to his / her difficulties without the assistance of a lawyer. It is also very difficult to do. Especially, in the field of law, laws and ordinances consist of legal terminology rarely used in everyday life. As a result, there is a limit to legal information retrieval only by a query made by a general person who does not know legal terminology.

The BEST-Project, which has been carried out since 2005 in the Netherlands, is representative of research to support the general law or the case search. The BEST-Project is developing a case search system that maps the current situation to the general term ontology and provides the case search result to the user. However, building and maintaining these generic ontology and terminology ontologies requires a lot of time and money. Also, in Korean, it is difficult to find a clear line between general terms and technical terms.

The present invention relates to a method and apparatus for learning a mapping probability between a general term, a technical term, and a statute by using Q & A data or precedent data collected from a web site and then using the same to predict a technical term or statute for the query term Field search support method and apparatus therefor.

According to an aspect of the present invention, there is provided a method and apparatus for supporting a search of a specialized field, thereby learning a mapping probability between a general term, a technical term, and a statute by using Q & A data or precedent data collected from a web site, A method for predicting and providing a terminology or a statute for a query word is provided.

According to an embodiment of the present invention, there is provided a method for processing a web document, comprising the steps of: (a) collecting query-answer data in a web document; (b) extracting words by analyzing the question part and the answer part in the question-answer data; (c) generating a general term-terminology mapping table by calculating a general term-terminology mapping probability by analyzing correlation between words included in the question portion and the answer portion; And (d) extracting and providing a terminology including a word included in the query using the term mapping table.

In the step (c), the general term-terminology mapping probability may be calculated using the frequencies of the words simultaneously appearing in the question part and the answer part.

The general term-terminology mapping probability is calculated using PMI (pairwise mutual information)

는 일반 용어 집합을 나타내고,

은 법령 출현 단어 집합을 나타낸다. 또한,

는 전문 용어 집합을 나타내며,

는 일반 용어 집합에는 포함되면서 법령 출현 단어 집합에는 포함되지 않는 단어들을 나타내고,

는 전문 용어에 포함되면서 동시에 법령 키워드에도 포함되는 단어를 나타낸다.here,

Represents a generic term set,

Represents a set of words appearing in a statute. Also,

Represents a terminology set,

Indicates the words included in the general term set but not included in the statutory word set,

Indicates a word included in a terminology keyword while being included in a terminology.

In the step (d), the mapping probability may be calculated and predicted using the mapping table with respect to the n (natural number) terms to be matched with the words included in the query.

In the step (d), the mapping probability is calculated using a naive Bayesian classifier,

The mapping probability is calculated by the following equation,

이되, 상기

는 매핑 테이블에 포함된 매핑 확률을 나타내고,

는 전문용어를 나타내며, X는 질의문을 나타낸다.
here,

However,

Represents the mapping probability included in the mapping table,

Denotes a terminology, and X denotes a query statement.

According to another embodiment of the present invention, there is provided a method comprising: (a) analyzing precedent data and extracting each word; (b) generating a word-law mapping table by calculating a mapping probability between words and laws using the word; And (c) predicting a statute for the query statement using the word-statute mapping table.

The word-law mapping table includes reliability according to a mapping between words and laws, and the reliability is calculated by the following equation,

는 판례 데이터내 출현한 단어의 집합을 나타내고,

는 법령명에 대한 집합을 나타내며,

은 판례 데이터내 출현 단어 집합에 포함되는 단어들 중 법령명에 포함되지 않는 단어를 나타내고,

는 판례 데이터내에 출현하는 단어들 중 법령명의 집합에 포함되는 법령들을 나타낸다.
here,

Represents a set of words appearing in precedent data,

Represents a set of legal names,

Represents a word not included in the legal name among the words included in the appearing word set in the precedent data,

Represent laws and ordinances included in the set of statute names among the words appearing in the precedent data.

According to another aspect of the present invention, after learning the mapping probability between a general term, a technical term, and a statute by using Q & A data or precedent data collected from a web site, it is possible to predict a terminology or a statute A device is provided that can provide.

According to an embodiment of the present invention, there is provided an information processing apparatus including a collection unit for collecting query-answer data in a web document; An extracting unit for extracting a word by analyzing a question part and an answer part in the query-answer data; A mapping table generation unit for generating a general term-terminology mapping table by calculating a general term-terminology mapping probability through analyzing the correlation between each of the words included in the question unit and the answer unit; And a prediction unit for extracting and providing a terminology including a word included in the query using the term mapping table.

According to another embodiment of the present invention, an extracting unit extracts words by analyzing precedent data; A mapping table generation unit for calculating a mapping probability between a word and a statute by using the word to generate a word-statute mapping table; And a predictor for predicting a statute for a query statement using the word-statistical mapping table.

A method and apparatus for supporting a specialized field search according to an embodiment of the present invention can learn a mapping probability between a general term, a technical term, and a statute by using Q & A data or precedent data collected from a web site, And can predict and provide a jargon or a statute for the query.

Therefore, the present invention has an advantage that it is possible to provide search convenience to a user who is relatively inexperienced about a jargon or a statute.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 schematically shows the configuration of a specialized field search support apparatus according to an embodiment of the present invention; FIG.
FIG. 2 illustrates a mapping table according to an embodiment of the present invention. FIG.
3 is a block diagram schematically illustrating an internal configuration of a specialized field search support apparatus according to another embodiment of the present invention.
FIG. 4 illustrates a word-statutory mapping table according to another embodiment of the present invention; FIG.
FIG. 5 is a diagram illustrating a result of a statute predicted for a query according to an embodiment of the present invention; FIG.
FIG. 6 is a flowchart illustrating a method of providing a terminology for a query term according to an embodiment of the present invention. FIG.
FIG. 7 is a flowchart illustrating a method of providing a statute for a query term according to another embodiment of the present invention; FIG.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. It is to be understood, however, that the invention is not to be limited to the specific embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

The present invention is to provide a statutory search service to a general person whose legal knowledge is relatively weak. The present invention can provide a general term used by a general person as a mapped term.

In addition, the present invention can generate a mapping table between words and statutes based on the mapping probability between words and laws based on precedent data, and can predict and provide a statute for the query using the generated mapping table.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a diagram schematically showing a configuration of a specialized field search supporting apparatus according to an embodiment of the present invention. FIG. 2 is a diagram illustrating a mapping table according to an embodiment of the present invention. In FIG. 1, the Q & A data is collected from the website and the Q & A data is divided into the question part and the answer part, and then the word is extracted. Then, the probability of mapping the general term and the technical term is calculated, Let's talk about a possible device.

Referring to FIG. 1, a specialized field search supporting apparatus 100 according to an exemplary embodiment of the present invention includes a collecting unit 110, an extracting unit 115, a mapping table generating unit 120, a memory 130, 135).

The collection unit 110 is a means for collecting query answer data through a communication network.

For example, the collecting unit 110 may collect Q & A data (e.g., Naver intellectuals, etc.) from a web document and store the Q & A data in a database.

The extracting unit 115 is a unit for extracting words in the question part and the answer part after analyzing the Q & A data to distinguish the question part and the answer part, respectively.

For example, the extracting unit 115 can morphologically analyze the Q & A data to extract each word in morpheme units. Assume that the question in the Q & A data is the same as "I ask about the image copyright of the product." The extracting unit 115 can extract the words "product", "image", "copyright" have.

In addition, due to the nature of Q & A data, it is highly possible that those who answer questions in comparison to the questioner are more likely to be experts in the field. On a particular website, a service is provided to respond to experts in the field such as doctors, lawyers and patent attorneys have.

Accordingly, the words extracted from the question part by analyzing the Q & A data in this specification are also referred to as general terms according to the syntax, and the words extracted from the answer part can be described as technical terms according to the syntax.

The mapping table generation unit 120 is a means for calculating a mapping probability through correlation analysis between words extracted from the Q & A data and generating a term mapping table based on the mapping probability.

For this, the mapping table generator 120 may calculate the mapping probability using the frequencies of the words appearing in the question part and the answer part for the words extracted from the question part and the answer part, respectively.

For example, the mapping table generation unit 120 may calculate the mapping probability by calculating the pairwise mutual information (PMI) between the general term and the terminology.

The mutual information amount (PMI) can be calculated using the following equation (1).

는 질문내 출현 단어 집합을 나타내고,

은 법령내 출현 단어 집합을 나타내며,

는 답변내 출현 단어 집합을 나타낸다. here,

Represents a set of emerging words in a query,

Represents the set of words appearing in the statute,

Represents the set of emerging words in the answer.

는 질문내 출현 단어 집합에 포함되는 단어들 중 법령내 출현 단어 집합에 포함되지 않는 단어들을 나타내며,

는 답변내 출현 단어 집합에 포함되면서 법령내 출현 단어 집합에도 포함되는 단어를 나타낸다.Also,

Indicates the words not included in the set of words appearing in the statute among the words included in the set of words appearing in the question,

Represents the words that are included in the set of emerging words in the answer and also included in the set of emerging words in the statute.

FIG. 2 illustrates a mapping table according to an embodiment of the present invention. Referring to FIG. 2, the mapping table includes the number of simultaneous occurrences of words included in an answer part corresponding to a word included in a question part, and a mapping probability calculated based on the number of occurrences.

The predicting unit 125 is a means for predicting and providing a related terminology by calculating a terminology mapping probability of a query sent by the user using the mapping table. The predicting unit 125 may extract and provide the top n (natural number) terms having a high mapping probability among the terms included in the query sentence by referring to the mapping table.

For example, the predictor 125 may calculate the probability of what the user's input query statement implies using the Naïve Bayesian classifier, which will be described in more detail below with reference to FIG. 3 .

The memory 130 is a means for storing various algorithms, mapping tables, and the like necessary for operating the specialized field search support apparatus 100 according to an embodiment of the present invention.

The control unit 135 includes internal components (for example, the collecting unit 110, the extracting unit 115, the mapping table generating unit 120) of the apparatus for supporting a specialized field search 100 according to an exemplary embodiment of the present invention, The memory 130, and the like).

In FIG. 1, the mapping probability between a general term and a terminology according to an exemplary embodiment of the present invention is calculated and a mapping table for the term is generated. FIG. 2 is a block diagram of an apparatus for predicting and providing a statute for a query language based on a mapping probability between query statements and statutes according to another embodiment of the present invention.

FIG. 3 is a block diagram schematically illustrating an internal configuration of a specialized field search support apparatus according to another embodiment of the present invention, FIG. 4 illustrates a word-law mapping table according to another embodiment of the present invention, FIG. 5 is a diagram illustrating a result of a statute predicted for a query according to an embodiment of the present invention.

3, the apparatus 300 for searching a special field according to another embodiment of the present invention includes a collecting unit 310, an extracting unit 315, a mapping table generating unit 320, a predicting unit 325, (330) and a control unit (335).

The collecting unit 310 is a means for collecting precedent data.

For example, the collection unit 310 may collect precedent data at a specific web site that provides precedent data.

The extracting unit 315 is means for analyzing precedent data and extracting each word.

For example, the extracting unit 315 can extract the words by analyzing the precedent data in morpheme units. The method of extracting words in a specific sentence from a morpheme unit itself is already known to those skilled in the art, so a separate explanation will be omitted.

The mapping table generating unit 320 is a unit for generating a mapping table by calculating a mapping probability between a word extracted from the precedent data and a statute. Hereinafter, the mapping table will be referred to as a word-statute mapping table in order to distinguish it from the mapping table described in FIG.

For example, the mapping table generation unit 320 may use reliability as a measure for calculating a mapping probability between a word extracted from the precedent data and a statute.

The mapping table generation unit 320 may calculate the mapping probability between words and laws using the following equation (2).

는 판례 데이터내 출현 단어 집합을 나타내고,

는 법령명 집합을 나타내고,

은 판례 데이터내 출현 단어 집합에 포함되는 단어들 중 법령명에 포함되지 않는 단어를 나타내고,

는 판례 데이터 내에 출현하는 단어들 중 법령명의 집합에 포함되는 법령들을 나타낸다.here,

Represents a set of emerging words in case data,

Indicates a set of statute names,

Represents a word not included in the legal name among the words included in the appearing word set in the precedent data,

Represent laws and ordinances included in the set of statute names among the words appearing in the precedent data.

FIG. 4 illustrates a word-statute mapping table. As shown in FIG. 4, the word-statutory mapping table includes each word and a statistic mapped to the corresponding word, and reliability according to the statute. That is, the higher the reliability of the statute mapped to each word, the higher the reliability of the statute mapped to the word.

The predicting unit 325 is means for predicting the statutory mapping probability of the input query sentence of the user using the keyword-statutory table.

For example, the predictor 325 may use the Naïve Bayesian classifier to compute the probability of what the user's input query statement implies.

Generally, since each query is often associated with several laws, the predictor 325 can calculate the query-statistical mapping probability using the number 3 of the Naive Bayes classifier with the MAX function removed.

This is expressed in Equation 3.

는 매핑 확률을 나타내며, 이는 수2와 동일하다. 예측부(325)는 질의문-법령 매핑 확률을 예측시 매핑 확률이 지나치게 감소하는 현상을 방지하기 위해 키워드-법령 테이블을 참조하여 법령 C_i와 매핑될 확률이 높은 상위 n개의 키워드-법령 매핑만을 이용하여 매핑 확률을 계산할 수 있다. 이를 수식으로 표현하면 수 4와 같다.here,

Represents the mapping probability, which is equal to the number 2. In order to prevent the mapping probability from being excessively reduced when predicting the query-statistical mapping probability, the predicting unit 325 refers to the keyword-statutory table and stores only the top n keyword-statutory mappings having a high probability of being mapped to the statute C _i The mapping probability can be calculated. This can be expressed as Equation 4.

FIG. 5 shows the statute predicted corresponding to the query statement and the mapping probability according to the statute.

As shown in FIG. 5, the predicting unit 325 calculates a mapping probability of a statute for a query statement using a word-statute mapping table, and calculates a statute for the query word in descending order of mapping probability according to the calculated mapping probability Can be predicted and provided.

The memory 330 is a means for storing various algorithms, mapping tables, and the like necessary for operating the specialized field search supporting apparatus 300 according to an embodiment of the present invention.

The control unit 335 includes internal components (for example, a collecting unit 310, an extracting unit 315, a mapping table generating unit 320) of the specialized field search supporting apparatus 300 according to an embodiment of the present invention. A prediction unit, a memory 330, and the like).

6 is a flowchart illustrating a method of providing a terminology for a query term according to an embodiment of the present invention.

In step 610, the specialized search support device collects Q & A data from a web site and stores it in a database.

In step 615, the specialized field search support apparatus divides the collected Q & A data into question parts and answer parts, and extracts words by analyzing each of them.

For example, the specialized field search support apparatus can extract the words in morphological units by analyzing the question part and the answer part respectively in terms of morpheme as described above. It will be obvious to those skilled in the art that a separate explanation will be omitted.

In step 620, the specialized field search support apparatus generates a mapping table through correlation analysis between words extracted from the question part and the answer part.

As described above, the mapping table can be generated by calculating the mapping probability through the correlation analysis between the words extracted from the question part and the answer part (the words extracted from the question part and the words extracted from the answer part).

In step 625, the specialized field search support apparatus predicts and provides a terminology corresponding to the query term using the mapping table.

For example, the specialized field search support apparatus can calculate the mapping probability for the inputted query word based on the Naïve Bayesian classifier by referring to the mapping table, and then provide the top n terms of the mapping probability with a high probability.

FIG. 7 is a flowchart illustrating a method of providing a statute for a query term according to another embodiment of the present invention.

In step 710, the specialized search support device collects precedent data. For example, a specialized search appliance can collect precedent data from a website and store it.

In step 715, the specialized search support apparatus analyzes the case data and extracts each word. For example, a specialized field search support device can extract words in morphological units by analyzing precedent data in units of form. It will be obvious to those skilled in the art that a separate explanation will be omitted.

In step 720, the specialized field search support apparatus calculates a mapping probability between a word extracted from the precedent data and a statute to generate a word-statute mapping table.

This is the same as that described with reference to FIG. 3, so duplicate descriptions will be omitted.

In step 725, the specialized search support apparatus predicts the statute of the query statement using the word-statute mapping table.

That is, the apparatus for searching a specialized field can calculate a statute mapping probability of a query word by referring to a word-statute mapping table, and predict a statute for the query word based on the calculation. To this end, the expert search support apparatus can use the Naive Bayesian classifier, which is a concept classification technique, which is the same as that described with reference to the number 4 in FIG. 3, so that a duplicate description will be omitted.

Meanwhile, the expert search support method according to an embodiment of the present invention may be implemented in a form of a program command that can be executed through a variety of means for processing information electronically and recorded in a storage medium. The storage medium may include program instructions, data files, data structures, and the like, alone or in combination.

Program instructions to be recorded on the storage medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of software. Examples of storage media include magnetic media such as hard disks, floppy disks and magnetic tape, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, magneto-optical media and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as devices for processing information electronically using an interpreter or the like, for example, a high-level language code that can be executed by a computer.

The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention as defined in the appended claims. It will be understood that the invention may be varied and varied without departing from the scope of the invention.

100, 300: Specialized Field Search Support Device
110, 310: collecting unit
115, 315:
120, and 320: a mapping table generation unit
125, 325:
130, 330: Memory
135, 335:

Claims (10)

A method of supporting a terminology search support device for terminology search for a query,
(a) collecting query-answer data in a web document;
(b) extracting each word of the question portion as a general term by analyzing the question portion and the answer portion, and extracting each word of the answer portion as a jargon in the question-answer data; ;
(c) calculating a general term-terminology mapping probability using the frequencies of the words in the question portion and the answer portion appearing simultaneously in the question portion and the answer portion, - generating a terminology mapping table, wherein the mapping table includes a mapping probability between the general term extracted from the query unit and the terminology; And
(d) using the mapping table to extract and provide n term terms having a high mapping probability corresponding to the general term matching the words included in the query, and providing a terminology search for the query, How to support. delete The method according to claim 1,
Wherein the general term-terminology mapping probability is calculated using PMI (pairwise mutual information).

here,

Represents a generic term set,

Represents a set of words appearing in a statute. Also,

Represents a terminology set,

Indicates the words included in the general term set but not included in the statutory word set,

Indicates words included in the terminology keyword while being included in the terminology. delete The method according to claim 1,
The step (d)
Wherein the mapping probability is calculated using a naive Bayesian classifier. ≪ RTI ID = 0.0 > 18. < / RTI > Claim 6 has been abandoned due to the setting registration fee. 6. The method of claim 5,
Wherein the mapping probability is calculated by the following equation: < EMI ID = 17.0 >

here,

However,

Represents the mapping probability included in the mapping table,

Denotes a terminology, and X denotes a query statement. A method of supporting a statutory search for a query language,
(a) extracting words by analyzing case data;
(b) generating a word-law mapping table by calculating a mapping probability between words and laws using the word; And
(c) predicting a statute for the query term using the word-statute mapping table. 8. The method of claim 7,
Wherein the word-law mapping table includes reliability according to a mapping between a word and a statute, and the reliability is calculated by the following equation.

here,

Represents a set of words appearing in precedent data,

Represents a set of legal names,

Represents a word not included in the legal name among the words included in the appearing word set in the precedent data,

Represents the statutes included in the set of statutory names among the words appearing in the precedent data. A collection unit for collecting query-answer data in a web document;
An extracting unit for extracting from the query-answer data the question part and the answer part, extracting each word of the question part as a general term by analyzing the question part and the answer part, and extracting each word of the answer part as a jargon;
The general term and the terminology included in the question portion and the answer portion are used to calculate the general term to terminology mapping probability using the frequencies of the words simultaneously appearing in the question portion and the answer portion, A mapping table generation unit for generating a mapping table, the mapping table including a mapping probability between the general term extracted from the query unit and the terminology; And
And a prediction unit for extracting and providing n term terms having a high mapping probability corresponding to a general term matching a word included in the query using the mapping table. An extracting unit for analyzing precedent data and extracting each word;
A mapping table generation unit for calculating a mapping probability between a word and a statute by using the word to generate a word-statute mapping table; And
And a predicting unit for predicting a statute for the query statement using the word-statistical mapping table.

Images