KR101247739B1 - Apparatus and method for predicting technology recognition change - Google Patents

Abstract

PURPOSE: A device for predicting a technology awareness change and a method thereof are provided to increase efficiency through the selection and concentration of core competencies by minimizing the trial and error of research and development and overlapped research prevention. CONSTITUTION: A technology statistic information extracting unit(110) analyzes documents stored in a multiple literature providing device to extract statistic information about specific technology. A technology awareness calculating unit(120) obtains a credibility value of a document from which the specific technology is extracted, a location information value in the document from which the specific technology is extracted, and an importance value of the specific technology in the document based on the statistic information in order to obtain an awareness value of the specific technology according to each document. An awareness change predicting unit(130) collects awareness values by year, obtains an average value by year, and predicts an awareness change speed of the specific technology. [Reference numerals] (100) Technology awareness change predicting device; (110) Technology statistic information extracting unit; (120) Technology awareness calculating unit; (130) Awareness change predicting unit

Description

Apparatus and Method for predicting technology recognition change

The present invention relates to an apparatus and method for predicting technology awareness, and more particularly, to obtain recognition values of specific technologies for each document based on statistical information on specific technologies, and to collect and average recognition values obtained for each document by year. The present invention relates to an apparatus and method for predicting technology awareness for predicting a rate of awareness change of the specific technology based on a yearly average value.

Knowledge and information are transformed into a knowledge-based industrial society that determines the competitiveness of the country, and national science and technology competitiveness is recognized as a source of national competitiveness.

In this regard, countries around the world are seeking to develop and select core technologies and research projects to survive in future competition.

For this reason, countries are looking for promising research areas that will lead future national science and technology, providing a technology life cycle.

However, the conventional technology life cycle only presents the technologies belonging to each step, and has a disadvantage in that it cannot provide information on the recognition, the speed of change, and the change direction of each technology.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and provides an apparatus and method for predicting technology awareness change that can provide information on technology awareness, the speed of change, and the direction of change.

Another object of the present invention is to estimate the expectations of the public as well as experts on the technology based on the market awareness, interest in the technology, and to detect the change in the estimation, it is possible to predict the future of the technology's awareness / expectation The present invention provides an apparatus and method for predicting change in awareness.

According to an aspect of the present invention in order to achieve the above object, a technical statistics information extraction unit for analyzing the documents stored in the multi-document providing apparatus, extracting statistical information for a specific technology, based on the statistical information of the specific technology A technology recognition calculation unit for obtaining recognition values for each document, a technology including a recognition change prediction unit for collecting and averaging the recognition values obtained for each document for each year, and predicting the rate of change of the specific technology based on the average value for each year. A recognition change prediction apparatus is provided.

The descriptive statistical information extracting unit analyzes the documents stored in the multi-document providing apparatus to obtain documents including the specific technology, and analyzes the obtained documents, the frequency in the document for the specific technology, the whole contained in the document. Statistical information including at least one of a technical name frequency, a location in a document, a frequency for each location, a frequency for each document type, and a document year may be extracted.

The technology recognition calculator calculates a reliability value of a document from which a specific technology is extracted based on the statistical information, a location information value from which a specific technology is extracted from a document, an importance value in a document of a specific technology, and an importance value in the entire document of a specific technology. Each can be obtained, and the obtained values can be used to obtain the recognition value of the specific technology.

The technology recognition calculator may obtain a recognition value of the specific technology by using the following equation.

[Mathematical Expression]

는 문서의 신뢰도 값,

는 문서내에서 특정 기술이 추출된 위치 정보 값,

는 특정 기술의 문서 내 중요도 값,

은 특정 기술의 문서 전체 내 중요도 값, I, J, K, L은

,

,

,

각각의 비중을 의미함.here,

Is the confidence value of the document,

Is the location information value from which a particular technology is extracted from the document,

Is the importance value in the document for a particular technology,

Is the importance value within the entire document of a particular technology, I, J, K, L

,

,

,

Meaning of each specific gravity.

The confidence value of the document may be a value obtained by using a weight obtained based on a weight set by a user or web search statistical information.

The importance value in a document of the specific description is obtained through the frequency AF of the specific description among the total description frequency TF included in the document.

The importance value in the entire document of the specific technology is obtained as an inverse ratio value of the total number of documents N to the number of documents d containing the specific technology.

The recognition change prediction unit collects the recognition values obtained for each specific document for each year, obtains the average of the collected recognition values by year, and predicts the recognition value of the predicted year based on the average value for each year, The predicted recognition value can be used to determine the rate of change of recognition.

The recognition change prediction unit may predict the recognition value of the predicted year by using at least one of a regression analysis method, a moving average analysis method, and a combination of the two methods (ARMA).

The recognition change predicting unit may compare the predicted change rate with a preset change direction step, and determine a change in recognition rate according to the comparison result.

According to another aspect of the present invention, a method for predicting a change in recognition of a specific technology by the technology awareness change prediction apparatus, the method comprising: (a) analyzing statistical documents stored in a multi-document providing apparatus, and extracting statistical information about a specific technology; (b) obtaining recognition values of the specific technology for each document based on the statistical information, and (c) collecting and averaging the recognition values obtained for each document by year, and calculating the specific technology based on the average value for each year. A technology awareness change prediction method comprising predicting a rate of change of awareness is provided.

In the step (a), the documents stored in the multi-document providing apparatus may be acquired to obtain documents including the specific technology, and the obtained documents may be analyzed to include the frequency in the document for the specific technology and included in the document. Statistical information including at least one of the overall technical name frequency, the frequency within the document and the location-specific frequency, the frequency of each document type, the year of the document may be extracted.

In the step (b), the reliability value of the document from which the specific technology is extracted based on the statistical information, the location information value from which the specific technology is extracted from the document, the importance value in the document of the specific technology, and the importance in the entire document of the specific technology Each value may be obtained and the recognition values of the specific technology may be obtained using the obtained values.

Recognition value of the specific technology can be obtained by using the following equation.

[Mathematical Expression]

는 문서의 신뢰도 값,

는 문서내에서 특정 기술이 추출된 위치 정보 값,

는 특정 기술의 문서 내 중요도 값,

은 특정 기술의 문서 전체 내 중요도 값, I, J, K, L은

,

,

,

각각의 비중을 의미함.here,

Is the confidence value of the document,

Is the location information value from which a particular technology is extracted from the document,

Is the importance value in the document for a particular technology,

Is the importance value within the entire document of a particular technology, I, J, K, L

,

,

,

Meaning of each specific gravity.

The confidence value of the document may be a value obtained by using a weight obtained based on a weight set by a user or web search statistical information.

The importance value in the document of the specific technology may be obtained through the frequency AF of the specific technology among all the technology name frequency TF included in the document.

The importance value in the entire document of the specific technology may be obtained as an inverse ratio value of the total number of documents N to the number of documents d containing the specific technology.

In the step (c), the recognition value obtained for each document for the specific technology is collected for each year, and the average of the collected recognition values is calculated for each year. The prediction value of the predicted year is predicted based on the average value for each year. The method may include obtaining a rate of change of the recognition using the predicted recognition value.

The technology awareness change prediction method may further include comparing the predicted change rate with a preset change direction step, and determining the change in recognition degree according to the comparison result.

According to another aspect of the present invention, by analyzing the documents stored in the multi-document providing apparatus, extracting statistical information for a specific technology, obtaining each recognition value for each specific document based on the statistical information, There is provided a computer-readable recording medium having a technology awareness change prediction method comprising collecting and averaging the recognition values obtained for each document for each year, and predicting the rate of change of the recognition of the specific technology based on the yearly average value. .

According to the present invention, it is possible to provide information on the recognition of the technology, the speed of change and the direction of change of the recognition.

In addition, it is possible to estimate the expectation of not only experts on the technology but also the general public based on the market awareness and interest in technology, and to detect the change of the estimation to enable future prediction of the technology's awareness / expectation.

In addition, by establishing a global monitoring system for technology, industry, market, and policy, and providing emerging signals and forward and backward information based on this, enhancing the success rate of technology commercialization and technological competitiveness of innovative SMEs, and advancement of researchers and innovative SMEs. Global research trends can be grasped to prevent duplicate research and minimize R & D trials and errors, which enables the selection and concentration of core competencies, increasing efficiency and increasing the likelihood of success in technology projects.

1 is a view schematically showing the configuration of the technology awareness change prediction apparatus according to the present invention.
2 is a flowchart illustrating a method of predicting a change in technology awareness by the apparatus for predicting change in technology awareness according to the present invention.

The foregoing and other objects, features, and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG.

1 is a view schematically showing the configuration of the technology awareness change prediction apparatus according to the present invention.

Referring to FIG. 1, the technology awareness change prediction apparatus 100 includes a technology statistical information extractor 110, a technology awareness calculator 120, and a awareness change predictor 130.

The descriptive statistical information extracting unit 110 extracts statistical information on the descriptive name by analyzing the documents. Here, the document includes a paper, a patent, a report, a web document, and the like. The statistical information is the name of the technology extracted from the document, the frequency of the technology name in the document, the frequency of the entire technology name contained in the document, the frequency and location-specific frequency in the document from which the technology name is extracted, the frequency of each document type, the statistics of individual technologies in the document, and the technology name. And the year of the document.

The descriptive statistics information extracting unit 110 analyzes documents stored in the multi-document providing apparatus to obtain documents including a descriptive name defined in a descriptive dictionary, and extracts descriptive names from the documents. The technical term dictionary stores technical terms in various fields such as medicine, engineering, and natural science, and documents such as papers, patents, and reports are stored in the multi-document providing device.

That is, the descriptive statistical information extracting unit 110 recognizes a term by performing morphological analysis, syntax analysis, and semantic analysis on the document, and recognizes the term as a descriptive name when the recognized term is registered in the descriptive term dictionary. At this time, the descriptive statistics information extracting unit 110 determines the location where each descriptive name is extracted and stores the descriptive name and the location together. Locations include titles, subtitles, intermediate titles, content, abstracts, and so on. In addition, the descriptive statistics information extracting unit 110 stores the frequency for each location of the document from which the descriptive name is extracted.

In addition, the descriptive statistical information extracting unit 110 accumulates the frequency of extracting the technical name from each document, and stores the year of documents from which the technical name is extracted.

In addition, the descriptive statistics information extracting unit 110 extracts the context between the descriptive names extracted based on the pre-stored context pattern information, and recognizes the relationship between descriptive names by normalizing the extracted context based on the provided normalization dictionary database. do.

When two or more description names are recognized, the descriptive statistics information extracting unit 110 extracts the context between the description names based on the context pattern information, and normalizes the extracted contexts based on the normalization dictionary database. Create

For example, in the case of the phrase "notebook consists of monitor, SSD, and mobile processor", the descriptive statistics information extractor 110 recognizes the technical names of the notebook, the monitor, the SSD, and the mobile processor.

In this case, since two or more descriptive names are recognized, the descriptive statistics information extracting unit 110 extracts only the contexts corresponding to pre-stored context pattern information, for example, "Be verb + past participle type + preposition".

Then, the descriptive statistics information extracting unit 110 unifies the tense of the context, for example, "was" into "be" based on the normalization dictionary, and similar expressions, for example, "similar to" and "similar with". Contextual normalization, such as "similar to", is performed to create a relationship between technical names.

The technology recognition calculator 120 calculates the recognition value of each technology based on the statistical information extracted by the technology statistics information extractor 110. Technology awareness is a numerical value of information on how important a particular technology is to experts, the public, etc. In this case, the technology recognition calculator 120 obtains a recognition value for each technology for each individual document.

The technology awareness calculation unit 120 may determine a reliability value of a document from which a specific technology is extracted based on statistical information, a location information value from which a specific technology is extracted from a document, an importance value in a document of a specific technology, and importance in the entire document of a specific technology. Each value is obtained, and the obtained values are used to determine the recognition value of a specific technology.

)를 문서별로 구한다. That is, the technology recognition calculator 120 recognizes each technology using the equation (1).

Get document by document.

는 기술명이 추출된 소스 문서의 신뢰도 값,

는 문서내에서 기술명이 추출된 위치 정보 값,

는 기술명의 문서 내 중요도 값,

은 기술명의 문서 전체 내 중요도 값을 의미한다. I, J, K, L은

,

,

,

각각의 비중을 의미하는 값으로, 그 값들은 미리 설정되어 있다. here,

Is the confidence value of the source document from which the technical name was extracted,

Is the location information value extracted from the technical name in the document,

Is the importance value in the document of the technical name,

Means the importance value in the entire document of the technical name. I, J, K, L

,

,

,

A value representing each specific gravity, the values of which are preset.

)는 문서가 작성된 사이트 또는 데이터베이스의 신뢰도를 의미하고, 뉴욕 타임즈, CNN 등 다양한 사이트별 신뢰도를 가중치로 표현한 수치이다.

가 큰 값을 갖게 되는 경우 소스의 신뢰도는 높은 것이고, 반대로 낮은 값을 갖게 되는 경우 신뢰도가 낮은 것을 의미한다. Reliability of the source document from which the technical name was extracted

) Refers to the reliability of the site or database where the document is written, and is a numerical expression representing the reliability of various sites such as the New York Times and CNN.

When the value is large, the reliability of the source is high. On the contrary, when the value is low, the reliability is low.

는 수학식 2를 이용하여 구해진다.

Is obtained using equation (2).

은 사람의 직관과 경험을 통해 수동적으로 계산된 가중치로, 미리 테이블 형태로 정의된다.

는 웹 검색 통계 정보를 기반으로 자동으로 얻어진 가중치로, 전처리를 통해 미리 또는 주기적으로 계산된 값일 수 있다.

는

의

에서의 비중,

는

의

에서의 비중을 의미한다.here,

Is a weight manually calculated through human intuition and experience, and is defined in the form of a table in advance.

Is a weight automatically obtained based on web search statistical information, and may be a value calculated in advance or periodically through preprocessing.

The

of

Specific gravity,

The

of

It means the specific gravity in.

Hereinafter, a method of obtaining w _IA will be described.

The weight for each resource is basically obtained from the search engine's search results. The weight (people's preference / awareness) for a particular resource, for example cnn.com, is based on the ratio of document frequency (w _IAT ) including cnn.com within the web document. The frequency of this document refers to the number of documents on the cnn site (g) in the search results document when the technical name or (technical name + relation name) is searched through the search engine as a query, and cnn in the non-documents of the cnn site (link) Calculate by adding the number of documents (h). For efficiency, the document frequency is calculated only for the top n documents.

That is, the ratio (w _IAT ) of document frequency including cnn.com for each query (descriptive name or descriptive name + relation name) is calculated using Equation 3.

Here, a and b reflect the relative importance of g and h by adjusting their values by considering the specific gravity of g and h. Since w _IAT is calculated once for each query, it is extracted from the search results of all or some query sets and averaged.

That is, performing each query search results in w _IAT for each resource (cnn, abc, bbc, etc.), and repeats the search by changing the query terms to collect and average the values of w _IAT while retrieving all queries (technical name or technical name + relation The average value of the w _IAT of the resource cnn, i.e., the preference / perception (w _IA ).

)은 기술 통계정보 추출부에서 추출된 기술명의 위치에 따라 가중치를 부여한 값일 수 있다. 위치는 제목, 부제, 중간 제목, 문단 등으로 구분할 수 있고,

는 기술명이 추출된 위치별로 값이 미리 정의되어 있다. 예를 들면, 제목은 10, 부제는 8,중간제목은 5 등과 같이 위치별로 값이 정해져 있다. 기술 인지도 계산부는 기술명이 추출된 위치가 다를 때, 각 추출된 문서내의 위치별로 서로 다른 가중치를 부여하여 위치 정보 값(

)을 구한다. Geolocation value from which the technical name is extracted within the document (

) May be a weighted value according to the position of the description name extracted by the description statistical information extractor. Locations can be divided into headings, subtitles, middle headings, paragraphs, and so on.

The value is predefined for each location where the technical name is extracted. For example, the title is 10, the subtitle is 8, the middle title is 5, and so on. When the technology name extraction position is different, the technology recognition calculation unit assigns different weights to the positions in each extracted document, thereby providing location information values (

).

)는 문서내의 빈도수 기반 중요도를 의미하는 것으로, 문서내에 포함된 전체 기술명 빈도수(TF) 중에 해당 기술명의 출현 빈도수(AF)를 통해 계산된다. Importance of document in technical name (

) Refers to the frequency-based importance in the document, and is calculated from the frequency of appearance of the technology name (AF) among the entire technology name frequency (TF) included in the document.

That is, the technical guidance calculation unit 120 calculates the importance in the document of the technical name by using Equation 4.

)는 문서내의 역문헌 빈도수를 기반으로 한 중요도 값으로, 전체 문헌수(N)에서 특정 기술명이 포함된 문서수(d)의 역(Inverse) 비율 값을 통해 계산된다. Importance of the entire document in the technical name (

) Is an importance value based on the inverse document frequency in the document, and is calculated through an inverse ratio value of the number of documents (d) including a specific technical name in the total number of documents (N).

That is, the technical recognition calculator 120 calculates the importance of the entire document of the technical name using Equation 5.

Technical names appearing in this document are much smaller w _L, w _L tends to emerge only some of the documents increases.

As described above, the technology recognition calculation unit 120 uses the reliability of the document from which the technology is extracted, the location information from which the technology is extracted from the document, the importance of the document in the technology, and the importance of the technology in the entire document. Obtain In this case, the technology recognition calculator 120 obtains the recognition of each technology for each document. For example, if there are 10 documents including a specific technology, the technology recognition calculator 120 obtains 10 recognitions for the specific technology.

The recognition change prediction unit 130 collects the technology recognition values obtained for each technology for each technology, obtains an average of the collected technology recognition values for each year, and then based on the average value for each year, the recognition change speed of each technology. And predict the change direction. In this case, since the statistical information extracted by the descriptive statistical information extracting unit includes a year, the recognition change predicting unit 130 may check the year of the document and may divide the descriptive recognition value for each year based on the year.

That is, the recognition change prediction unit 130 collects the recognition values obtained for each document for each specific technology for each year, and calculates the average of the collected recognition values for each year. Thereafter, the recognition change prediction unit 130 predicts the recognition value of the prediction year based on the average value for each year, and calculates the change rate of the recognition using the predicted recognition value. In this case, the recognition change predicting unit 130 may predict the recognition value of the predicted year by using a regression analysis method, a moving average analysis method, or a combination of the two methods (ARMA). Here, the prediction year means a year for which the recognition value is to be predicted. In addition, the rate of change in the awareness may be calculated using the slope of the predicted awareness recognition value compared with the previous year's awareness average value. Here, the previous year refers to the previous year of the forecast year, and the awareness value of the forecast year refers to the forecast value predicted based on the average value for each year.

For example, for each technology, the recognition values obtained for individual documents from 2001 to 2011 are collected for each year, and the average of the recognition values for each year in 2001, 2002, ..., 2011 is obtained. Then, using the regression analysis method, moving average analysis method, and a combination of the two methods (ARMA), the 2012 awareness value is predicted, and the slope of the 2012 awareness estimate is compared to the 2011 awareness average value. Obtain The obtained technology may be the rate of change of awareness of a specific technology. For example, when the recognition average value in 2011 is 4 and the prediction predicted value in 2012 is 6, the slope becomes (6-4) / 1 = 2, and thus the speed of technology awareness change may be '2'.

When the awareness change rate for a particular technology is calculated as described above, the awareness change prediction unit 130 determines a change direction by determining a step to which the calculated change rate belongs. At this time, the step and the change direction according to the change rate of the recognition is set in advance, the recognition change prediction unit may determine the change direction by determining the step corresponding to the change rate.

For example, the rate of change corresponding to the range of 0 ~ 1 is 1 step, the rate of change corresponding to the range of 1 ~ 3 is 2 steps, the rate of change corresponding to 3 or more is divided into 3 steps, and 1 Steps can be set to change direction as slow rising, step 2 rising, and step 3 rising sharply. If the rate of change in technology awareness is '2', since it corresponds to step 2, the direction of change may be determined by rising.

In addition, the change rate corresponding to the range of 0 ~ -1 is 1 step, the change rate corresponding to the range of -1 ~ -3 is 2 steps, and the change rate corresponding to -3 or less is divided into 3 steps. , The first stage is gentle descent, the second stage is falling, and the third stage is descent.

2 is a flowchart illustrating a method of predicting a change in technology awareness by the apparatus for predicting change in technology awareness according to the present invention.

Referring to FIG. 2, the technology recognition change prediction apparatus analyzes a document to extract a technical name (S202), and extracts statistical information about the extracted technical name (S204). That is, the technology recognition change prediction apparatus analyzes documents stored in the multi-document providing apparatus to obtain documents including the technical name defined in the technical term dictionary, and extracts the technical name from the documents. Then, the technology awareness change prediction apparatus extracts the location of each technology name, the year, etc., and extracts the technology name extracted from the document, the frequency of the technology name in the document, the frequency of the entire technology name contained in the document, and the location in the document from which the technology name is extracted. And statistical information including the frequency per location, the frequency per document type, the statistics of individual descriptions in the document, the year of the document from which the technical name is extracted, and the like.

After performing step S204, the technology awareness change prediction apparatus obtains a technology awareness value based on the statistical information (S206). That is, the technology awareness change prediction device is based on statistical information of each technology, the reliability value of the document from which the technology is extracted, the location information value from which the technology is extracted from the document, the importance value in the document of the technology, and the importance value in the entire document of the technology. We obtain each of them and use the obtained values to find the recognition value for each document by individual documents.

After performing step S206, the technology awareness change prediction device collects and averages technology recognition values obtained for each document for each technology (S208), and calculates the speed and direction of change in the recognition of each technology based on the average value for each year. Predict (S210). That is, the technology awareness change prediction device collects the recognition values obtained for each document for each specific technology for each year, and calculates the average of the collected recognition values for each year. Then, the technology awareness change prediction apparatus predicts the recognition value of the prediction year based on the average value for each year, and calculates the change rate of the awareness using the predicted recognition value. Then, the technology awareness change prediction device determines the change direction by determining the stage to which the obtained change rate belongs.

The technology awareness change prediction method may be written by a program, and codes and code segments constituting the program may be easily inferred by a programmer in the art.

Thus, those skilled in the art will appreciate that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the embodiments described above are to be considered in all respects only as illustrative and not restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

100: technology recognition change prediction device 110: technology statistical information extraction unit
120: technology awareness calculation unit 130: awareness change prediction unit

Claims (20)

A descriptive statistics information extracting unit extracting statistical information on a specific technique by analyzing documents stored in the multi-document providing apparatus;
Obtaining at least one of a reliability value of a document from which a specific technology is extracted, a location information value from which a specific technology is extracted from a document, an importance value in a document of a specific technology, and an importance value in the entire document of a specific technology based on the statistical information; A technology recognition calculator for obtaining document recognition values of the specific technology for each document using the obtained values; And
Includes a recognition change prediction unit for obtaining the recognition value obtained by the document for each year and the average value, and predicting the rate of change of the specific technology based on the average value for each year;
And the importance value in the document of the specific technology is obtained through the frequency (AF) of the specific technology among all the technology name frequencies (TF) contained in the document .
The method of claim 1,
The descriptive statistical information extracting unit analyzes the documents stored in the multi-document providing apparatus to obtain documents including the specific technology, and analyzes the obtained documents, the frequency in the document for the specific technology, the whole contained in the document. Technical recognition change prediction apparatus, characterized in that for extracting statistical information, including at least one of the technical name frequency, the location within the document and the frequency by location, frequency by document type, document year.
delete The method of claim 1,
The technology recognition calculator calculates the recognition value of the specific technology by adding the reliability value of the document, the location information value of the specific technology extracted from the document, the importance value in the document of the specific technology, and the importance value in the entire document of the specific technology. A technology awareness change prediction device.
The method of claim 1,
And a reliability value of the document is a value obtained by using a weight obtained based on a weight set by a user or web search statistical information.
delete The method of claim 1,
The importance value of the entire document of the specific technology is a technology awareness change prediction device, characterized in that it is obtained as an inverse ratio value of the number of documents (d) containing a specific technology in the total number of documents (N).
The method of claim 1,
The recognition change prediction unit collects the recognition values obtained for each specific document for each year, obtains the average of the collected recognition values by year, and predicts the recognition value of the predicted year based on the average value for each year, A technology awareness change prediction device, characterized in that the rate of change of recognition is calculated using the predicted recognition value.
9. The method of claim 8,
The recognition change predicting unit predicts a recognition value of the predicted year using at least one of a regression analysis method, a moving average analysis method, and a method of mixing the two methods (ARMA).
The method of claim 1,
And the recognition change predicting unit compares the predicted change rate with a preset change direction step, and determines a change in recognition rate according to the comparison result.
In the method of predicting the change in awareness of the technology recognition technology,
(a) analyzing the documents stored in the multi-document providing apparatus to extract statistical information on a specific technology;
(b) at least one of a reliability value of a document from which a specific technology is extracted based on the statistical information, a location information value from which a specific technology is extracted from a document, an importance value in a document of a specific technology, and an importance value in the entire document of a specific technology Obtaining a recognition value of the specific technology using the obtained values; And
(c) collecting and averaging the recognition values obtained for each document for each year, and predicting a recognition rate of change of the specific technology based on the yearly average value;
The importance value in the document of the specific technology is a technology recognition change prediction method, characterized in that is obtained through the frequency (AF) of the specific technology of the total technology name frequency (TF) contained in the document .
The method of claim 11,
The step (a)
Analyze the documents stored in the multi-document providing apparatus to obtain documents containing the specific technology, and analyze the obtained documents, the frequency in the document for the specific technology, the frequency of the entire description name contained in the document, the location in the document And extracting statistical information including at least one of location frequency, document type frequency, and document year.
delete The method of claim 11,
The recognition value of the specific technology is a value obtained by adding the reliability value of the document, the location information value of the specific technology extracted from the document, the importance value in the document of the specific technology, and the importance value in the entire document of the specific technology. Awareness change prediction method.
The method of claim 11,
And a confidence value of the document is a value obtained by using a weight obtained based on a weight set by a user or web search statistical information.
delete The method of claim 11,
The importance value of the entire document of the specific technology is a technology awareness change prediction method, characterized in that it is obtained as an inverse ratio value of the number of documents (d) including the specific technology in the total number of documents (N).
The method of claim 11,
The step (c)
Collecting recognition values obtained for each specific document by document for each year, and obtaining a yearly average of the collected recognition values;
Predicting a recognition value of a predicted year based on the average value for each year;
And obtaining a rate of change of the recognition rate using the predicted recognition value.
The method of claim 11,
And comparing the predicted recognition rate of change with a preset change direction step, and determining the recognition change direction according to the comparison result.
When executed by the technology awareness change prediction device,
Analyzing the documents stored in the multi-document providing apparatus, extracting statistical information on a specific technology;
Obtaining at least one of a reliability value of a document from which a specific technology is extracted, a location information value from which a specific technology is extracted from a document, an importance value in a document of a specific technology, and an importance value in the entire document of a specific technology based on the statistical information; Obtaining recognition values of the specific technique using the obtained values; And
(c) collecting and averaging the recognition values obtained for each document for each year, and predicting a recognition rate of change of the specific technology based on the yearly average value;
A technology recognition change prediction method is recorded by a program and read by a computer, characterized in that the importance value in the document of the specific technology is obtained through the frequency of the specific technology among the total technology name frequencies (TF) contained in the document. Possible recording media.

Images