TWI627543B - Research method of mind map generation method - Google Patents

Abstract

A mental map generation method for research topics is implemented by a processing module, and includes the following steps: (A) selecting a maximum amount of data from multiple responses corresponding to multiple respondents Responsive information, wherein each reply data includes a plurality of reply content respectively corresponding to a plurality of different research items related to a research topic, each research item includes at least one question, and each reply content includes a related one-to-one correspondence The part of the reply to the question of the research project; (B) according to the reply data selected in step (A), using a natural language processing method to obtain a plurality of such reply contents respectively corresponding to the reply data Replying to the first summary of the content of the content section; and (C) generating a mental map based on the summary of the first document.

Description

Research method of mind map generation method

The present invention relates to a method for generating graphics, and more particularly to a method for generating a mind map of a research subject.

In recent years, domestic vocational colleges have regarded thematic courses as a very important compulsory course. Students can develop the ability to establish themes, collect materials, and analyze materials in the special courses, and lay the foundation for future advanced research.

However, students often feel confused during the process of data analysis. They don't even know how to conduct data analysis. How to guide students to systematically summarize the research priorities and results from the complicated data is a top priority.

In 1976, Buzan proposed the idea of a mind map, which presents internal thinking through radiological thinking and representation, which helps analysis and memory. In addition, the mind mapping method has the function of visualizing the cognitive concept, which can help the integration of thinking and provide a solution to the problem solving. If the mental map can be applied to data analysis, it should help students to clarify the research focus and results.

Accordingly, it is an object of the present invention to provide a mind map generation method for research topics to assist students in summarizing research priorities and results.

Therefore, the mind map generation method of the research subject of the present invention is implemented by a processing module, and includes the following steps: (A) selecting multiple responses from multiple responses corresponding to multiple respondents The reply data of the maximum amount of data, wherein each reply data includes a plurality of reply contents respectively corresponding to a plurality of different research items related to a research topic, each research item includes at least one question, and each reply content includes a related The part of the reply to the question in a corresponding research project; (B) the reply data selected in step (A), using a natural language processing method to obtain a plurality of such corresponding responses to the reply data Replying to a first document summary of the reply content portion of the content; and (C) generating, according to the first document summary obtained in step (B), a mental map of the research items related to the research subject, wherein The mind map indicates the subject of the study, the research items, and the contents of the first document abstract, the relationship between the research topic and the research projects, and each of the first articles Summary relationship of corresponding research projects.

The effect of the present invention is that the first file summary corresponding to the reply content portion of the reply content is automatically obtained by the processing module, and according to the first file summary, the research subject is indicated, These research projects, and The mental map of the content of the first document abstract, the relationship between the research topic and the research projects, and the relationship between each first document abstract and its corresponding research project to help students summarize the research priorities and results.

11~21‧‧‧Steps

121~125‧‧‧Substeps

151~153‧‧‧ substeps

Other features and effects of the present invention will be apparent from the embodiments of the present invention, wherein: FIG. 1 is a flowchart illustrating an embodiment of a mind map generating method of the subject matter of the present invention; FIG. 2 is a flow FIG. 3 is a flow chart showing how the processing module calculates the first file summary and the second file. A summary of similarity between document summaries; FIG. 4 is a schematic diagram illustrating a mind map generated by the present invention; and FIG. 5 is a schematic diagram illustrating the result of adding the mind map to the second document abstract and at least one research article .

Referring to FIG. 1, an embodiment of a method for generating a mind map of a subject of the present invention can be implemented by a processing module (not shown) and includes the following steps. In this In an embodiment, the processing module may be a processor having computing power included in a personal computer or a server, and the steps of the mind mapping generating method of the research subject of the present invention may be in a software form such as including a plurality of instructions. The application implements and executes the application by the processing module to implement a mind map generation method of the subject matter of the present invention.

In step 11, the processing module selects a response data having the largest amount of data from multiple response data corresponding to multiple respondents, wherein each of the response data includes multiple corresponding to multiple The response content of different research projects related to a research topic, each research project contains at least one question, and each reply content includes a reply content portion of the question related to a corresponding research project. Table 1 is an example questionnaire, and Table 2 is an example of the reply material.

In step 12, the processing module obtains the first part of the reply content portion of the reply content corresponding to the reply data by using a natural language processing method according to the reply data selected in step 11. Summary of the file.

It is worth mentioning that the processing module obtains a first file summary corresponding to the reply content portion of each reply content of the reply data by performing the following sub-steps 121-125 (see FIG. 2).

In sub-step 121, the processing module divides each of the reply content portions into a plurality of content words according to the reply content portion of the reply content of the pen reply data. In this embodiment, the processing module segments each of the replies by using a word-breaking method related to Chinese Knowledge Information Processing (CKIP).

In sub-step 122, the processing module obtains a plurality of content keywords and portions of the reply content portion by using a latent semantic analysis (LSA) method according to the content words of the reply content portion. A semantic matrix of content statements.

In sub-step 123, the processing module calculates a semantic similarity between each content statement relative to each of all other content statements based on the semantic matrix.

In sub-step 124, the processing module calculates, according to all semantic similarities and a semantic threshold corresponding to each content statement, an association in which all similarities in the semantic similarities of each content statement are greater than the semantic threshold. Quantity.

In sub-step 125, the processing module obtains the content statement with the associated number of top N high according to the number of associations corresponding to the content sentences, respectively, as the first file digest corresponding to the reply content portion. In this embodiment, the value of N can be adjusted according to the number of content sentences of the content part of the reply. For example, the more the content statement, the larger the value of N; the smaller the content statement, the smaller the value of N. .

Table 3 illustrates a first file summary of the portion of the reply content corresponding to each reply content obtained according to the reply data shown in Table 2.

In step 13, the processing module generates a mental map (see FIG. 4) of the research items related to the research subject according to the first file digests obtained in step 12, wherein the mental map indicates The research subject, the research items, and the contents of the first document abstract, the relationship between the research topic and the research projects, and the relationship between each first document summary and its corresponding research project, and still in each The first file summary corresponding to the reply content portion indicates one of the respondent identification codes corresponding to one of the respondents associated with the first file summary.

It is worth mentioning that the processing module first draws the content corresponding to the research topic, the content of the research projects, and the plurality of nodes of the contents of the first file summary related to the research projects, Next, draw a link between the research topic and each research project, and a link between each first document abstract and its corresponding research project. In detail, each research item corresponds to the reply content including at least one reply content part, and the reply content part corresponds to a first file abstract, so the correspondence between each research item and the reply content part is And the corresponding relationship between the reply content part and the first file summary, the corresponding relationship between each first file summary and each research item is obtained, and then the corresponding relationship between each research item and each first file summary is drawn. A link to the first document summary and its corresponding research project. In addition, the process The module also adds the respondent identification code associated with the first file summary to the node corresponding to the first file summary.

In step 14, the processing module obtains a second file summary of each of the reply content portions of the reply content corresponding to each of the remaining reply data according to the remaining reply data having the non-maximum data amount.

In step 15, the processing module calculates a summary similarity between the first file summary and the second file summary according to the first file summary and the second file summary related to the problem of each research item. .

It is worth mentioning that the processing module calculates the similarity between the first file summary and the second file summary by performing the following sub-steps 151-153 (see FIG. 3).

In sub-step 151, the processing module divides the first file digest into a plurality of first words according to the first file digest and the second file digest related to the question of each research item, and The second file digest is divided into a corresponding plurality of second words. In this embodiment, the processing module divides the first file digest and the second file digest by using the word segmentation method related to Chinese knowledge information processing.

In sub-step 152, the processing module obtains a plurality of first keywords of the first file summary and a plurality of second keywords of the second file summary. In this embodiment, the processing module uses a word frequency reverse file frequency method to obtain the first keywords and the second keywords.

In sub-step 153, the processing module calculates the digest similarity between the first file digest and the second file digest according to the first keyword and the second keywords. In this embodiment, the processing module calculates the digest similarity using a cosine similarity formula.

In step 16, the processing module determines whether the digest similarity between the first file digest and the second file digest is less than a digest threshold. When the determination result is YES, proceed to step 17; otherwise, proceed to step 18.

In step 17, the processing module adds the content of the second file summary to the mind map, and the mind map further indicates the relationship between the second file summary and its corresponding research item, and on the two file abstracts. Indicates one of the respondent identification codes corresponding to one of the respondents associated with the second file summary.

In step 18, the processing module adds one of the respondent identification codes corresponding to one of the respondents associated with the second file summary to the first file summary.

For example, if respondent B is concerned with "users' understanding of escape packages", the second summary of the research project is "First aid kits are escape kits, which can contain some first-aid supplies and medicines." A's "Escape bag is something that is eaten inside, like cookies, high-calorie food, used, like a radio, a flashlight." The summary of the abstract of the first document is less than the summary threshold, then the second file summary. The first aid kit is the escape kit, which can be used to put some first-aid supplies and medicines. It will be added to the mind map (see Figure 5). If the respondent B is concerned with the function of the escape vest, this research project The summary of the second document and the respondent A's summary of the summary of the first document corresponding to the function of the escape vest are larger than the summary threshold, and the function of the escape vest is The first document summaries corresponding to the research project are added to one of the respondent identification codes "B" corresponding to the respondent B (see FIG. 5). If Respondent B has no response to the "Recommended Improvements" project, the first summary of the research project on the mental map corresponding to the "Recommended Improvements" will not be one of the respondents' B Respondent identification code "B".

In step 19, the processing module stores the first keywords of the first file summary, the second keywords of the second file summary, and one of the second keywords according to the problem related to each research item. A plurality of databases related to the research articles of the research subject determine whether at least one research article includes one of the first keywords and the second keywords. When the determination result is YES, proceed to step 20; otherwise, proceed to step 21.

In step 20, the processing module adds an identification data of the research article including one of the first keywords and the second keywords to the mind map, and the mind map also indicates the Identifying the relationship between the data and one of the first file summary and the second file summary corresponding thereto. In this embodiment, the identification material contains the author of the research article and the date of publication (see Figure 5). By correlating the research article related to one of the first document abstract and the second document abstract on the mind map, the correlation between the research result and the literature discussion can be further verified.

In step 21, the processing module does not add any identification data to the mind map.

In summary, the mind map generation method of the research subject of the present invention automatically obtains the first file summary corresponding to the reply content portion of each reply content by the processing module, and according to the research topic, the first A document abstract automatically generates the mental map indicating the relationship between the research subject and the research project, and the relationship between each research project and the corresponding first document abstract, thereby effectively assisting the student to summarize the research focus And results, and can save students time to analyze data. In addition, by marking the identification data of the research article related to the first document abstract or the second document abstract on the mind map, it is also convenient for the student to refer to the first document abstract or the first in the research process. The research article related to the second document summary, and further, the related interviewee identification code is marked on the first document abstract and the second document abstract, and the first document abstract and the second may also be indicated at a glance The summary of the document and the relationship of the respondents, indeed, can achieve the object of the present invention.

However, the above is only the embodiment of the present invention, and the scope of the invention is not limited thereto, and all the equivalent equivalent changes and modifications according to the scope of the patent application and the patent specification of the present invention are still The scope of the invention is covered.

Claims (5)

A mental map generation method for research topics is implemented by a processing module, and includes the following steps: (A) selecting a maximum amount of data from multiple responses corresponding to multiple respondents Responsive information, wherein each reply data includes a plurality of reply content respectively corresponding to a plurality of different research items related to a research topic, each research item includes at least one question, and each reply content includes a related one-to-one correspondence The part of the reply to the question of the research project; (B) according to the reply data selected in step (A), using a natural language processing method to obtain a plurality of such reply contents respectively corresponding to the reply data Replying to a first file summary of the content portion; (C) generating, according to the first document summary obtained in step (B), a mind map of the research items related to the research subject, wherein the mind map indicates The subject matter of the study, the research projects, and the contents of the summary of the first documents, the relationship between the research subject and the research projects, and the summary of each of the first documents (D) a second summary of each of the replies of the replies corresponding to each of the remaining replies, based on the remaining replies with a non-maximum amount of information; (E) calculating a summary similarity between the first document summary and the second document summary based on the first document summary and the second document summary related to the problem of each research item, step (E) comprising The following steps, (E-1) according to the question related to each question of each research item a file summary and the second file summary, dividing the first file digest into a plurality of first words, and dividing the second file digest into corresponding plurality of second words, (E-2) utilizing one a word frequency reverse file frequency method, obtaining a plurality of first keywords of the first file summary and a plurality of second keywords of the second file summary, and (E-3) according to the first keywords and the first keywords a second keyword, using a cosine similarity formula to calculate the digest similarity between the first file digest and the second digest; (F) determining the digest between the first digest and the second digest Whether the similarity is less than a summary threshold value; and (G) when the determination result is yes, adding the content of the second file summary to the mind map, and the mind map also indicates the second file summary and its corresponding study The relationship of the project. The mind map generating method of the research subject of claim 1, wherein in step (G), the mind map further indicates on the two file summary that one of the respondents associated with the second file summary corresponds to One of the respondent IDs. The mind map generation method of the research subject of claim 1, after the step (G), further comprising the following steps: (H) according to the first file summary related to the problem of each research item a first keyword, the second keywords of the second file summary, and a database storing a plurality of research articles related to the research topic, determining whether at least one research article includes the first keywords And one of the second keywords; and (I) when the determination result is YES, an identification data of the research article including one of the first keywords and the second keywords is added to the mind map, and the mind map also indicates The relationship between the identification data and one of the first file summary and the second file summary corresponding thereto. The mind map generating method of the research subject of claim 1, wherein the step (B) comprises the following steps: (B-1) each reply content according to the reply content portion of the reply content of the reply data Partially dividing into a plurality of content words; (B-2) obtaining, by using a potential semantic analysis method, a plurality of content keywords and a plurality of content sentences of the reply content portion according to the content words of the reply content portion a semantic matrix; (B-3) calculating a semantic similarity between each content statement and each of all other content statements according to the semantic matrix; (B-4) according to all semantic meanings corresponding to each content statement Similarity and a semantic threshold, calculating an associated quantity of all semantic similarities corresponding to each content statement that is greater than the semantic threshold; and (B-5) corresponding associations corresponding to the content statements respectively The quantity obtains the content statement of the first number N of the associated quantity as the first file summary corresponding to the part of the reply content. The mind map generating method of the research subject of claim 1, wherein in step (C), the mind map is further indicated on the first file summary corresponding to the part of the reply content A file summary relates to one of the respondents' identifiers corresponding to one of the respondents.