TWI679545B - Domain knowledge base system and method in the same - Google Patents

Abstract

The invention is an automatic establishment domain knowledge base system, which includes: a servo device; a data collection main unit is provided on the servo device and is electrically connected to the servo device; a data exploration main unit (Knowledge Exploration) is provided on the servo device and is electrically connected To the servo device; and the knowledge tree sub-unit (KT organizer), which is set in the data exploration main unit and is electrically connected to the servo device, and the collected data is converted into knowledge data through the knowledge tree sub-unit; the key data captured on the network The data collection main unit is converted into collected data, and the collected data is converted into knowledge data through the data exploration main unit.

Description

Automatic establishment of domain knowledge base system and method

The present invention relates to an automatic establishment of a domain knowledge base system and a method thereof. By using the method of establishing the domain knowledge base automatically through the present invention, a user can more accurately and automatically understand the professional field through the present invention.

In recent years, information exchange on the Internet has become one of the tools that everyone needs and must learn.

As the speed of network transmission is getting faster and faster, and the amount of memory required by servers is getting larger and larger, how to make data an efficient and systematic classification of knowledge management systems is the current society. A big subject.

In view of this, in the application field of the knowledge management system, the classification management of files in a tree structure is often used, generally known as the "Knowledge Tree", which is used to store the files and the basis of their classification and archiving. The acquisition of files is particularly suitable for Knowledge management system. Files are usually files that are automatically retrieved from the network, file system, and database. A tree-structured document classification management knowledge tree is like a file system with classification names and classification attributes. There may be other classifications or documents, and there may be a need to be able to continuously trace back to the overall situation of the past.

Therefore, a good feedback design can not only accurately classify the file content for file management, but also continuously feedback the overall classification knowledge tree for knowledge management, or analyze a series of important keywords for a specific file or classification range. Trends that change over time for analysis of market trends and business intelligence.

In view of the above-mentioned problems of the prior art, an object of the present invention is to provide an automatic establishment of a domain knowledge base system and a method thereof, so as to solve the shortcomings of conventionally manually building a knowledge base to manage an incomplete file database.

According to the purpose of the present invention, an automated domain knowledge base system is proposed, which includes: a servo device; a data collection main unit provided on the servo device and electrically connected to the servo device; a data exploration main unit provided on the servo device and electrically And the knowledge tree sub-unit is set in the data exploration main unit and is electrically connected to the servo device, and the collected data is converted into knowledge data through the knowledge tree sub-unit; the key data captured on the network is transmitted through the data The main collection unit is converted into collected data, and the collected data is converted into knowledge data through the data exploration main unit.

Preferably, the data collection main unit further includes a domain creation subunit, a concept selection subunit, a web crawler subunit, a keyword extraction subunit, and a data classification subunit.

Preferably, the data exploration main unit further includes a concept extraction subunit, a topic analysis subunit, and a knowledge processing subunit.

Preferably, the servo device can be electrically connected to a chat robot device, a search engine system, a key recommendation system, and a document classification system.

According to the purpose of the present invention, a method for automatically establishing a domain knowledge base is proposed, which includes the following steps: fetching data from a network through a server device, selecting key data from a data collection subunit of the data collection main unit, and selecting the key data from the data collection main unit. The data classification sub-unit performs filtering and analysis processing to generate collected data; the concept extraction sub-unit of the data exploration main unit performs data analysis processing on the collected data and retrieves the collected data; the collected data is defined by the theme in the topic analysis sub-unit of the data exploration main unit The data and knowledge tree sub-units use algorithms to calculate the correlation rate between the collected data and the topic data, and then distinguish between the main relationship chain and the special relationship chain between the collected data and the topic data, and set up multiple nodes on the main relationship chain. And a plurality of child nodes are set up on a special relationship chain, and a plurality of nodes and a plurality of child nodes are connected in series to establish a knowledge tree domain.

Preferably, the data collection main unit further includes a domain creation subunit, a concept selection subunit, a web crawler subunit, and a keyword extraction subunit.

Preferably, the data exploration main unit further includes a concept extraction subunit and a data processing subunit.

Preferably, a plurality of nodes and a plurality of child nodes are stored in the field of the knowledge tree and are fed back to another data.

As mentioned above, according to the automatic establishment domain knowledge base system and method of the present invention, it may have one or more of the following advantages:

(1) When capturing data from the Internet, you can filter and filter and build a system that is convenient for users to search or identify on the machine.

(2) The present invention is compatible with document classification or recommendation systems, and can be quickly and objectively and transparently implemented for users to understand and analyze.

101‧‧‧Domain Knowledge Management System

102‧‧‧ Establish domain subunit

103‧‧‧Data Processing Subunit

104‧‧‧Concept selection subunit

105‧‧‧Web crawler subunit

106‧‧‧Collecting Information

107‧‧‧Keyword Extraction Subunit

108‧‧‧Data Classification Subunit

109‧‧‧Concept extraction subunit

110‧‧‧Subject Analysis Subunit

111‧‧‧ Knowledge Tree Subunit

112‧‧‧ Domain Knowledge Base

113‧‧‧Data Collection Master Unit

114‧‧‧Data Exploration Main Unit

115‧‧‧Automatic domain knowledge structure

116‧‧‧Forum

117‧‧‧Community

118‧‧‧News

119‧‧‧General dictionary

120‧‧‧ professional dictionary

121‧‧‧Cloud Dictionary

122‧‧‧Application level

123‧‧‧Search Engine System

124‧‧‧File Classification System

125‧‧‧Key Recommendation System

126‧‧‧chat robot device

201‧‧‧ Establishing domain subunits

202‧‧‧Concept selection subunit

203‧‧‧Web crawler subunit

204‧‧‧Keyword extraction subunit

205‧‧‧Data Classification Subunit

206‧‧‧Data Exploration Subunit

207‧‧‧Data Collection Master Unit

301‧‧‧Data Exploration Main Unit

302‧‧‧Concept extraction subunit

303‧‧‧Thematic Analysis Subunit

304‧‧‧ Knowledge Tree Subunit

305‧‧‧Data Processing Subunit

306‧‧‧Data Collection Master Unit

FIG. 1 is a schematic diagram of the relationship of the automatic establishment of a domain knowledge base system according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of a data collection main unit of a method for automatically establishing a domain knowledge base according to an embodiment of the present invention.

FIG. 3 is a schematic diagram of a main unit for automatically establishing a domain knowledge base method and data exploration according to an embodiment of the present invention.

In order to help the review committee understand the technical features, contents and advantages of the present invention and the effects that can be achieved, the present invention is described in detail in conjunction with the accompanying drawings in the form of embodiments, and the drawings used therein are The main purpose is only for the purpose of illustration and supplementary description. It may not be the actual proportion and precise configuration after the implementation of the invention. Therefore, the attached drawings should not be interpreted and limited to the scope of rights of the present invention in actual implementation. He Xianming.

As shown in Figure 1, it is a schematic diagram of the relationship between the automatic establishment of a domain knowledge base system of the present invention, which includes: a domain knowledge management system 101, a domain creation sub-unit 102, a data processing sub-unit 103, a concept selection sub-unit 104, and a network. Crawler subunit 105, new data 106, keyword extraction subunit 107, data classification subunit 108, concept extraction subunit 109, topic analysis subunit 110, knowledge tree subunit 111, domain knowledge base 112, data collection main unit 113, Data Exploration Main Unit 114, Automatic domain knowledge structure 115, forum 116, community 117, news 118, general dictionary 119, professional dictionary 120, cloud dictionary 121, application level 122, search engine system 123, file classification system 124, key recommendation system 125.

The domain knowledge management system 101 includes: establishing a domain sub-unit 102 and a data processing sub-unit 103. Establishing the domain sub-unit 102 explains that key data is edited and input into the data collection main unit 113 in the automatic domain knowledge structure 115, and Use the domain knowledge base 112 to set the connectivity between the data to generate collected data. The data exploration main unit 114 will find the concept / keyword of the domain from the data in the collected data, and use the topic analysis subunit 110 The concept / keyword definition topic knowledge tree subunit 111 uses the domain concept / keyword and topic to establish a domain knowledge tree system to convert collected data into knowledge data, and outputs the knowledge data through the data processing subunit 103.

It is further explained that when the key data enters the data collection main unit 113, it will first select the key data through the concept selection sub-unit 104 and the web crawler sub-unit 105 and establish the integrity and correctness of the data itself, and then use the domain knowledge base 112 and new The data 106 complements and further strengthens the integrity of the collected data. The collected data is obtained from the online forum 116, community 117, and news 118. Finally, the key data is extracted through the keyword extraction subunit 107 and the data classification subunit 108. The new data 106 is compatible to generate the collected data.

The collected data will then enter the cloud dictionary 121 in the data exploration main unit 114 to retrieve important keywords. The cloud dictionary 121 includes: a general dictionary 119 and a professional dictionary 120, accumulating dictionary keywords commonly used or related to the professional field; and then After the collected data is collected through the theme analysis sub-unit 110, a dendritic analysis of the knowledge tree sub-unit 111 will be performed. The analysis method uses the minimum set method to calculate the correlation rate between the collected data and distinguish the correlation rate between the data. After the relationship between the main connected chain and the special connected chain, nodes are set on the main connected chain and child nodes are set on the special connected chain. Knowledge nodes are generated by connecting nodes and sub-nodes in series, which is the knowledge data, where the high correlation rate value is the main connection chain and the low correlation rate value is the special connection chain.

The internal formula of the knowledge tree subunit 111 is calculated as follows:

k _b )>Conf(k _b

k _a )→k _a

SK,k _b

GK k _a ,k _b

DT,k _a ≠k _b Conf ( k _a

k _b )> Conf ( k _b

k _a ) → k _a

SK , k _b

GK k _a , k _b

DT , k _a ≠ k _b

The collected data k _a and the knowledge data k _b are obtained by dividing the intersection of the collected data k _a and the knowledge data k _b by the collected data k _a needs to be greater than or equal to the subject analysis TH. The collected data k _a and the knowledge data k _b Domain knowledge tree DT, and the collected data k _a and knowledge data k _b belong to different data. The knowledge data k _b is linked through the main relationship chain GK, and the collected data k _a is linked through the special relationship chain SK.

When a user wants to capture a set of key data from the Internet, the key data is converted into collected data through the data collection main unit 113, and the collected data is converted into knowledge data through the data exploration main unit 114.

As shown in FIG. 2, it is a schematic diagram of a data collection main unit of the method for automatically establishing a domain knowledge base according to the present invention, including: a domain sub-unit 201, a concept selection sub-unit 202, a web crawler sub-unit 203, and a keyword extraction sub-unit. Unit 204, data classification sub-unit 205, data exploration sub-unit 206, and data collection main unit 207.

Among them, after the domain-building sub-unit 201 has determined the keywords entered by the user, the keywords are selected from the important keywords as keyword seeds through the concept selection sub-unit 202 and combined and provided to the web crawler. Unit 203.

The web crawler sub-unit 203 uses the keyword seed to collect multiple domain-related data on the Internet and saves it to the domain knowledge base 112. The keyword extraction sub-unit 204 will key the newly entered data. The words are retrieved, and the data classification sub-unit 205 is used to re-categorize the collected data into domain and non-domain through the keyword correlation rate. The classified data and keywords are provided to the data exploration main unit 114 to construct domain knowledge.

Further explain that the keyword extraction sub-unit 204 uses user-defined extracted keyword lengths to perform keyword extraction on newly collected data. The extracted keywords will become candidate keywords and be stored in the domain knowledge base 112 and the key will be The data is compatible with the new data 106 through the data classification subunit 205 to collect data.

As shown in FIG. 3, it is a schematic diagram of the main unit of data exploration of the method for automatically establishing a domain knowledge base in the present invention, which includes: a data exploration main unit 301, a concept extraction subunit 302, a topic analysis subunit 303, a knowledge tree subunit 304, and data The processing subunit 305 and the data collection main unit 306.

From the above-mentioned collected data generated by the data collection main unit 207, the concept extraction subunit 302 finds out the concept / keyword of the domain from the domain knowledge base 112 provided by the data collection main unit 207, and the topic analysis subunit 303 The keywords define the topic. The knowledge tree sub-unit 304 uses the domain concepts / keywords and topics to build a domain knowledge tree. The important keywords in the domain knowledge tree are fed back to the data collection main unit 207 for the next keyword cycle. By making the present invention searchable through each keyword, more keywords can be supported in series with each other.

Further explain that the exploration main unit 301 uses the formula as follows:

Where k _t is the number of topics, DK is the domain keyword, and k _d is the number of domains. The number of domains k _d defines the boundary of the domain keywords DK to increase the linkability between the keywords.

To further explain, the knowledge tree subunit 304 can be further described by the domain key setting difference and the domain key setting difference, where the domain key setting difference is:

DK is a domain keyword, and DK is another set of keywords in the domain. Therefore, using the key setting of the domain is poor, the domain keyword can be used to define its relationship chain.

Finally, the user can use the present invention to make the keyword system repeat the self-learning for smoother and more frequent keyword updates.

To sum up, the method for automatically establishing a domain knowledge base of the present invention includes the following steps:

Step 1: Capture data through the servo device and select key data through the data collection sub-unit of the data collection main unit.

Step 2: The key data is filtered and processed in the data classification sub-unit of the main data collection unit to generate collected data.

Step 3: Collecting the data The concept of the main unit of data exploration extracts the sub-units for data analysis and processing, and simultaneously collects the collected data.

Step 4: Collect data Define the topic data in the topic analysis sub-unit of the data exploration main unit.

Step 5: The knowledge tree sub-unit uses an algorithm to calculate the correlation rate between the collected data and the topic data, and then distinguishes between the main relationship chain and the special relationship chain between the collected data and the topic data, and sets up plural numbers on the main relationship chain. Nodes, and a plurality of child nodes are set up on the special relationship chain, and a plurality of nodes and a plurality of child nodes are connected in series to establish a knowledge tree field.

The above description is only exemplary, rather than limiting, and any equivalent modification or change made without departing from the spirit and scope of the present invention shall be included in the scope of the attached patent application.

Claims (8)

A system for automatically establishing a domain knowledge base includes: a servo device configured to capture a data; a data collection main unit disposed on the servo device and including a data collection subunit and a data classification subunit, the The data collection sub-unit obtains key data from the servo device. The key data is filtered and analyzed in the data classification sub-unit to generate a collected data. A data exploration main unit is set on the servo device and includes a concept extraction sub-unit. A topic analysis subunit and a knowledge tree subunit; wherein the concept extraction subunit performs a data analysis process on the collected data and retrieves the collected data, and the topic analysis subunit defines a topic data on the collected data, and The knowledge tree sub-unit uses an algorithm to calculate the correlation rate between the collected data and the topic data, and further distinguishes a main relationship chain and a special relationship chain between the collected data and the correlation ratio of the topic data, and Set up multiple nodes on the relationship chain and set up multiple children on the special relationship chain Point, and the plurality of nodes and a plurality of sub node in series, so as to establish a knowledge tree fields. As described in item 1 of the scope of the patent application, the domain building knowledge base system is automated, wherein the data collection main unit further includes a building domain sub-unit, a concept selection sub-unit, a web crawling sub-unit, and a keyword extraction sub-unit . The automatic establishment of a domain knowledge base system as described in the first item of the patent application scope, wherein the data exploration main unit further includes a topic analysis sub-unit and a data processing sub-unit. The automatic establishment of a domain knowledge base system as described in the first patent application scope, wherein the servo device is electrically connected to a chat robot, a search engine system, a key recommendation system, and a document classification system. A method for automatically establishing a domain knowledge base includes the steps of: capturing a piece of data through a servo device, a data collection sub-unit of a data collection main unit, selecting a key piece of data, and doing it in a data classification sub-unit of the data collection main unit. Filtering analysis processing to generate a collection of data; and using a concept extraction sub-unit of a data exploration main unit to perform a data analysis process on the collected data and retrieve the collected data, the collected data is analyzed through a theme of a data exploration main unit A sub-unit is defined in a sub-unit. A knowledge tree sub-unit uses an algorithm to calculate the correlation rate between the collected data and the subject data, and distinguishes between a main relationship chain and a correlation between the collected data and the correlation rate of the subject data. A special relationship chain, a plurality of nodes are set up on the main relationship chain and a plurality of child nodes are set up on the special relationship chain, and the plurality of nodes and the plurality of child nodes are connected in series, thereby establishing a knowledge tree field. The method for automatically establishing a domain knowledge base as described in item 5 of the scope of patent application, wherein the main data collection unit further includes a domain creation subunit, a concept selection subunit, a web crawler subunit, and a keyword extraction subunit. . The method for automatically establishing a domain knowledge base as described in item 5 of the scope of patent application, wherein the data exploration main unit further includes a concept extraction subunit and a data processing subunit. The method for automatically establishing a domain knowledge base as described in item 5 of the scope of the patent application, wherein the knowledge tree domain saves the plurality of nodes and the plurality of child nodes, and feeds back to another key material.