A method for quickly referencing PPT and Word documents.

Abstract

This invention relates to a method for quickly referencing PPT and Word documents, comprising an associated layout module, a chronological reorganization module, a corpus extraction module, a text decomposition module, an import / editing module, and a mapping construction module. For files decomposed and saved to the server in chapter format, quickly referencing a specific chapter's content requires only a global search and a click to insert the desired chapter content into the document, saving users time and improving document writing efficiency.

Description

Technical Field

[0001] This invention relates to the field of computers, and in particular to a method for quickly referencing PPT and Word documents. Background Technology

[0002] In daily applications, each PowerPoint slide is parsed into a single slide document, and the text content is extracted and inserted into an Elasticsearch database for retrieval. Similarly, Word documents are broken down into smaller Word documents according to rules such as headings, lists, or line breaks, and the text from each smaller Word document is extracted and inserted into an Elasticsearch database. Plugins for Office and WPS are also developed to retrieve these smaller files and insert them into local documents.

[0003] Existing technology: Generally, the required content is collected and searched manually by copying and pasting from documents. However, manually collecting and searching documents is labor-intensive, takes a lot of time, and sometimes omissions occur, causing trouble for document organization. Summary of the Invention

[0004] In view of this, the present invention provides a method for quickly referencing PPT and Word documents to solve or partially solve the above problems.

[0005] To achieve the effects of the above technical solutions, the technical solution of the present invention is: a method for quickly referencing PPT and Word documents, comprising: an associated layout module, a temporal reorganization module, a corpus extraction module, a text decomposition module, an import module, and a mapping construction module;

[0006] The related layout module divides the relationships between different chapters into three types of community relationships: direct reference, co-reference, and co-cited. Community relationships are used to divide the document network. Direct reference means that two chapters have the same keywords. Co-reference means that two chapters have the same keywords but are not exactly the same. Co-cited means that two chapters have the same keywords as a third chapter, but the two chapters do not have the same keywords.

[0007] The association layout module is used to classify chapters; it classifies chapters based on community relationships between them; it uses a clustering algorithm to divide chapters into several topic clusters; the clustering algorithm uses aggregation degree to measure the quality of community division; the clustering algorithm is unsupervised learning; topic clusters are the different classes in the clustering analysis results; aggregation degree is an indicator of the classification effect of the clustering algorithm, and the calculation method of aggregation degree is shown in Formula 1:

[0008] Formula 1:

[0009] Where D is the topic cluster; j is the topic cluster number; D j This is the topic cluster numbered j; P is the aggregation degree, which ranges from (0,1). A higher aggregation degree indicates a better classification effect of the clustering algorithm; P(D j ) represents the aggregation degree of the topic cluster numbered j; N represents the number of phrases in the chapter; i represents the chapter number; N ji It is the number of phrases in chapter i within topic cluster j; n is the number of chapters in topic cluster j; n j is the number of chapters in topic cluster j; w is the phrase in topic cluster j; k is the phrase number; m is the number of phrases in topic cluster j; m j It is the number of phrases in the topic cluster numbered j; w k It is the phrase numbered k; p is the probability of the phrase appearing; p ji p is the probability of a phrase appearing in chapter i within topic cluster j; ji (w k ) is the probability of the phrase numbered k appearing in chapter numbered i within topic cluster numbered j;

[0010] The time-series reorganization module is used to group chapters within a topic cluster. It divides the time interval of a topic cluster into several time-shifting windows with equal time intervals. The time interval of a topic cluster is a range of time defined by the minimum and maximum storage times of the chapters within the cluster. Chapters with storage times within the same time-shifting window form sub-topic clusters. The average number of chapters within each topic cluster is maintained at around 8. The number of time-shifting windows is the integer result of dividing the number of chapters within a topic cluster by 8. The equal time interval is the quotient of the time interval of a topic cluster divided by the number of time-shifting windows.

[0011] The corpus extraction module is used to extract chapter information; the process of extracting sub-cluster keywords by the corpus extraction module includes steps P1 and P2:

[0012] Step P1: The corpus extraction module performs word segmentation on the corpus of chapters within the same subtopic cluster to obtain the segmented word groups;

[0013] Step P2: The corpus extraction module performs word frequency statistics on the segmented word groups. The corpus extraction module sets the word group that appears most frequently in each sub-topic cluster as the sub-cluster topic word of that sub-topic cluster; the topic word is the word group that appears most frequently; the sub-cluster topic word is the word group that appears most frequently in the sub-topic cluster.

[0014] The corpus extraction module's process for extracting chapter research questions and methods includes steps T1-T5:

[0015] Step T1: The corpus extraction module breaks down long sentences in the chapter into short sentences;

[0016] Step T2: The corpus extraction module extracts the three semantic elements (S, R, O) from the short sentences, where S represents the subject, R represents the relation, and O represents the object;

[0017] Step T3: The corpus extraction module filters out the target three elements from the three semantic elements; the target three elements are the three semantic elements that contain the research question and research method;

[0018] Step T4: The corpus extraction module performs part-of-speech tagging on the three target elements;

[0019] Step T5: The corpus extraction module performs word frequency statistics on the words after part-of-speech tagging. The corpus extraction module defines the top three words in the word frequency statistics results for research questions and research methods as chapter research questions and chapter research methods, respectively.

[0020] The corpus extraction module draws a horizontally flowing river to the right for each topic cluster, with each river drawn in a different color. Equidistant vertical lines representing time are drawn along the river; the number of these lines equals the number of sub-topic clusters within the corresponding topic cluster. Timestamps are marked on these vertical lines, increasing sequentially from left to right, with the timestamp representing the midpoint of the time movement window. The timestamps are located above the river. Each equidistant vertical line corresponds one-to-one with the sub-topic clusters within the time window marked on that vertical line. The vertical lines also display information corresponding to the corresponding sub-topic clusters. The keywords for the corresponding subtopic clusters are displayed on the equidistant vertical lines, between the timestamp and the river. The width of the water in the horizontally flowing river to the right varies; the more chapters in the subtopic clusters corresponding to the equidistant vertical lines, the wider the water flow at the intersection with the equidistant vertical lines. Chapter nodes are marked on the equidistant vertical lines, and the number of chapter nodes equals the number of chapters in the subtopic clusters corresponding to the equidistant vertical lines. Each chapter node corresponds one-to-one with a chapter, and the keywords for the corresponding chapter are displayed next to the chapter node. When the user moves the mouse over a chapter node, the chapter research question and chapter research method will be displayed in a pop-up window.

[0021] The import module provides document import and query interfaces; users upload documents through the document import interface; users enter keywords through the query interface to search for chapters in the database; users can view the rivers of each topic cluster drawn by the corpus extraction module through the query interface.

[0022] The text splitting module is used to split user-uploaded documents into several chapters and save them in the database; the text splitting module parses each slide in a PPT document into a single slide document, and extracts the text content of a single slide document as a chapter; the text splitting module splits a Word document into several fragment documents according to the rules of headings, lists and line breaks, and extracts the text content of fragment documents as a chapter.

[0023] The mapping module is used to search for chapters associated with user-input keywords. It measures the similarity between user-input keywords and sub-cluster keywords by calculating cosine similarity. The mapping module sends the sub-cluster information of the sub-cluster with the highest similarity to the user-input keywords to the import module. Sub-cluster information includes sub-cluster keywords, chapter IDs of all chapters within the sub-cluster, chapter keywords, time window, chapter research questions, and chapter research methods. The chapter ID is the chapter's identifier in the database; it is a unique identifier for each chapter.

[0024] The import module provides feedback on sub-cluster information to the user. The user can select several chapters to reference or discard sub-cluster information through the interface provided by the import module. If the user selects several chapters to reference by clicking, the import module uses the obtained chapter ID of the selected chapter to query the database for all document information of the selected chapter and inserts all the retrieved document information into the document position specified by the user.

[0025] The beneficial results of this invention are as follows: This invention provides a method for quickly referencing PPT and Word documents, including an associated layout module, a chronological reorganization module, a corpus extraction module, a text decomposition module, a writing and import module, and a mapping construction module. For files that have been decomposed and saved to the server in chapter format, quickly referencing the technical solutions provided in a specific chapter requires only a global search and a click on the reference operation to insert the desired chapter content into the document, saving users time and improving document writing efficiency. Detailed Implementation

[0026] To make the technical problem to be solved, the technical solution, and the beneficial effects of this invention clearer, the invention will be described in detail below with reference to embodiments. It should be noted that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention. Products that can achieve the same function are equivalent substitutions and improvements, and are all included within the protection scope of this invention. The specific method is as follows:

[0027] Example 1: This example specifically introduces a common method for quickly referencing PPT and Word documents, as follows:

[0028] The methods for quickly referencing PPT and Word documents include: related layout module, time sequence reorganization module, corpus extraction module, text decomposition module, writing and import module, and mapping construction module;

[0029] The related layout module divides the relationships between different chapters into three types of community relationships: direct reference, co-reference, and co-cited. Community relationships are used to divide the document network. Direct reference means that two chapters have the same keywords. Co-reference means that two chapters have the same keywords but are not exactly the same. Co-cited means that two chapters have the same keywords as a third chapter, but the two chapters do not have the same keywords.

[0030] The association layout module is used to classify chapters; it classifies chapters based on community relationships between them; it uses a clustering algorithm to divide chapters into several topic clusters; the clustering algorithm uses aggregation degree to measure the quality of community division; the clustering algorithm is unsupervised learning; topic clusters are the different classes in the clustering analysis results; aggregation degree is an indicator of the classification effect of the clustering algorithm, and the calculation method of aggregation degree is shown in Formula 1:

[0031] Formula 1:

[0032] Where D is the topic cluster; j is the topic cluster number; D j This is the topic cluster numbered j; P is the aggregation degree, which ranges from (0,1). A higher aggregation degree indicates a better classification effect of the clustering algorithm; P(D j ) represents the aggregation degree of the topic cluster numbered j; N represents the number of phrases in the chapter; i represents the chapter number; N ji It is the number of phrases in chapter i within topic cluster j; n is the number of chapters in topic cluster j; n j is the number of chapters in topic cluster j; w is the phrase in topic cluster j; k is the phrase number; m is the number of phrases in topic cluster j; m j It is the number of phrases in the topic cluster numbered j; w k It is the phrase numbered k; p is the probability of the phrase appearing; p ji p is the probability of a phrase appearing in chapter i within topic cluster j; ji (w k ) is the probability of the phrase numbered k appearing in chapter numbered i within topic cluster numbered j;

[0033] The time-series reorganization module is used to group chapters within a topic cluster. It divides the time interval of a topic cluster into several time-shifting windows with equal time intervals. The time interval of a topic cluster is a range of time defined by the minimum and maximum storage times of the chapters within the cluster. Chapters with storage times within the same time-shifting window form sub-topic clusters. The average number of chapters within each topic cluster is maintained at around 8. The number of time-shifting windows is the integer result of dividing the number of chapters within a topic cluster by 8. The equal time interval is the quotient of the time interval of a topic cluster divided by the number of time-shifting windows.

[0034] The corpus extraction module is used to extract chapter information; the process of extracting sub-cluster keywords by the corpus extraction module includes steps P1 and P2:

[0035] Step P1: The corpus extraction module performs word segmentation on the corpus of chapters within the same subtopic cluster to obtain the segmented word groups;

[0036] Step P2: The corpus extraction module performs word frequency statistics on the segmented word groups. The corpus extraction module sets the word group that appears most frequently in each sub-topic cluster as the sub-cluster topic word of that sub-topic cluster; the topic word is the word group that appears most frequently; the sub-cluster topic word is the word group that appears most frequently in the sub-topic cluster.

[0037] The corpus extraction module's process for extracting chapter research questions and methods includes steps T1-T5:

[0038] Step T1: The corpus extraction module breaks down long sentences in the chapter into short sentences;

[0039] Step T2: The corpus extraction module extracts the three semantic elements (S, R, O) from the short sentences, where S represents the subject, R represents the relation, and O represents the object;

[0040] Step T3: The corpus extraction module filters out the target three elements from the three semantic elements; the target three elements are the three semantic elements that contain the research question and research method;

[0041] Step T4: The corpus extraction module performs part-of-speech tagging on the three target elements;

[0042] Step T5: The corpus extraction module performs word frequency statistics on the words after part-of-speech tagging. The corpus extraction module defines the top three words in the word frequency statistics results for research questions and research methods as chapter research questions and chapter research methods, respectively.

[0043] The corpus extraction module draws a horizontally flowing river to the right for each topic cluster, with each river drawn in a different color. Equidistant vertical lines representing time are drawn along the river; the number of these lines equals the number of sub-topic clusters within the corresponding topic cluster. Timestamps are marked on these vertical lines, increasing sequentially from left to right, with the timestamp representing the midpoint of the time movement window. The timestamps are located above the river. Each equidistant vertical line corresponds one-to-one with the sub-topic clusters within the time window marked on that vertical line. The vertical lines also display information corresponding to the corresponding sub-topic clusters. The keywords for the corresponding subtopic clusters are displayed on the equidistant vertical lines, between the timestamp and the river. The width of the water in the horizontally flowing river to the right varies; the more chapters in the subtopic clusters corresponding to the equidistant vertical lines, the wider the water flow at the intersection with the equidistant vertical lines. Chapter nodes are marked on the equidistant vertical lines, and the number of chapter nodes equals the number of chapters in the subtopic clusters corresponding to the equidistant vertical lines. Each chapter node corresponds one-to-one with a chapter, and the keywords for the corresponding chapter are displayed next to the chapter node. When the user moves the mouse over a chapter node, the chapter research question and chapter research method will be displayed in a pop-up window.

[0044] The import module provides document import and query interfaces; users upload documents through the document import interface; users enter keywords through the query interface to search for chapters in the database; users can view the rivers of each topic cluster drawn by the corpus extraction module through the query interface.

[0045] The text splitting module is used to split user-uploaded documents into several chapters and save them in the database; the text splitting module parses each slide in a PPT document into a single slide document, and extracts the text content of a single slide document as a chapter; the text splitting module splits a Word document into several fragment documents according to the rules of headings, lists and line breaks, and extracts the text content of fragment documents as a chapter.

[0046] The mapping module is used to search for chapters associated with user-input keywords. It measures the similarity between user-input keywords and sub-cluster keywords by calculating cosine similarity. The mapping module sends the sub-cluster information of the sub-cluster with the highest similarity to the user-input keywords to the import module. Sub-cluster information includes sub-cluster keywords, chapter IDs of all chapters within the sub-cluster, chapter keywords, time window, chapter research questions, and chapter research methods. The chapter ID is the chapter's identifier in the database; it is a unique identifier for each chapter.

[0047] The import module provides feedback on sub-cluster information to the user. The user can select several chapters to reference or discard sub-cluster information through the interface provided by the import module. If the user selects several chapters to reference by clicking, the import module uses the obtained chapter ID of the selected chapter to query the database for all document information of the selected chapter and inserts all the retrieved document information into the document position specified by the user.

[0048] The beneficial results of this invention are as follows: This invention provides a method for quickly referencing PPT and Word documents, including an associated layout module, a chronological reorganization module, a corpus extraction module, a text decomposition module, a writing and import module, and a mapping construction module. For files that have been decomposed and saved to the server in chapter format, quickly referencing the technical solutions provided in a specific chapter requires only a global search and a click on the reference operation to insert the desired chapter content into the document, saving users time and improving document writing efficiency.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the claims. Furthermore, the above description should be understood and implemented by those skilled in the art; therefore, any equivalent modifications made based on the disclosure of this invention should be included within the scope of these claims.

Claims

1. A method for quickly referencing PPT and Word documents, characterized in that, include: The module includes: related layout module, temporal reorganization module, corpus extraction module, text decomposition module, writing and import module, and mapping construction module. The association layout module categorizes the associations between different chapters into three types of community relationships: direct reference association, co-reference association, and co-cited association. These community relationships are used to divide the document network. A direct reference association indicates that two chapters share the same keywords. A co-reference association indicates that two chapters share the same keywords, but the keywords are not exactly the same. A co-cited association indicates that two chapters share the same keywords with a third chapter, but the two chapters do not share any keywords. The association layout module is used to classify the chapters; the association layout module classifies the chapters using the community relationships between chapters as features; the association layout module uses a clustering algorithm to divide the chapters into several topic clusters; the clustering algorithm uses aggregation degree to measure the quality of community division; the clustering algorithm is unsupervised learning; The topic clusters are the different classes in the clustering analysis results; the aggregation degree is an indicator that measures the classification effect of the clustering algorithm, and the calculation method of the aggregation degree is shown in Formula 1: Where D is the topic cluster; j is the topic cluster number; D j It is the topic cluster numbered j; P is the aggregation degree, the value of which ranges from (0, 1), and the larger the aggregation degree, the better the classification effect of the clustering algorithm; P(D j ) represents the aggregation degree of the topic cluster numbered j; N represents the number of phrases in the chapter; i represents the chapter number; N ji It is the number of phrases in chapter i within topic cluster j; n is the number of chapters in topic cluster j; n j is the number of chapters in topic cluster j; w is the phrase in topic cluster j; k is the phrase number; m is the number of phrases in topic cluster j; m j It is the number of phrases in the topic cluster numbered j; w k It is the phrase numbered k; p is the probability of the phrase appearing; p ji p is the probability of a phrase appearing in chapter i within topic cluster j; ji (wk) is the probability of the phrase numbered k appearing in the chapter numbered i within the topic cluster numbered j; The time-series reorganization module is used to group the chapters within the topic cluster; the time-series reorganization module divides the time interval of the topic cluster into several time-shifting windows according to equal time intervals; the time interval of the topic cluster is a time range formed by the minimum and maximum storage times of the chapters within the topic cluster; chapters with storage times within the same time-shifting window form a sub-topic cluster; the average number of chapters in each topic cluster is maintained at 8; the number of time-shifting windows is the integer result of dividing the number of chapters within the topic cluster by 8; the equal time interval is the quotient of the time interval of the topic cluster divided by the number of time-shifting windows. The corpus extraction module is used to extract chapter information; the process of extracting sub-cluster topic words by the corpus extraction module includes steps P1 and P2: Step P1: The corpus extraction module performs word segmentation on the corpus of chapters within the same subtopic cluster to obtain segmented word groups; Step P2: The corpus extraction module performs word frequency statistics on the segmented word groups, and sets the word group that appears most frequently in each sub-topic cluster as the sub-cluster topic word of that sub-topic cluster; The corpus extraction module's process for extracting chapter research questions and methods includes steps T1-T5: Step T1: The corpus extraction module breaks down long sentences in the chapter into short sentences; Step T2: The corpus extraction module extracts the three semantic elements (S, R, O) from the short sentence, where S represents the subject, R represents the relation, and O represents the object; Step T3: The corpus extraction module filters out the target three elements from the semantic three elements; the target three elements are semantic three elements that include the research question and research method; Step T4: The corpus extraction module performs part-of-speech tagging on the target three elements; Step T5: The corpus extraction module performs word frequency statistics on the words after part-of-speech restoration. The corpus extraction module defines the top three words in the word frequency statistics results for research questions and research methods as chapter research questions and chapter research methods, respectively. The corpus extraction module draws a horizontally flowing river to the right for each topic cluster, with each river drawn in a different color. Equidistant vertical lines representing time are drawn on the river; the number of these lines equals the number of sub-topic clusters within the corresponding topic cluster. Timestamps are marked on the vertical lines, increasing sequentially from left to right, and the timestamps represent the midpoint of the time movement window. These timestamps are located above the river. Each equidistant vertical line corresponds one-to-one with the sub-topic clusters corresponding to the time windows marked on those vertical lines. The equidistant vertical lines also display information related to the time movement window. The keywords of the sub-topic clusters corresponding to the lines, and the keywords displayed on the equidistant vertical lines are between the timestamp and the river; the width of the water flow in the horizontally flowing river to the right varies, and the more chapters in the sub-topic clusters corresponding to the equidistant vertical lines, the wider the water flow at the intersection with the equidistant vertical lines; chapter nodes are marked on the equidistant vertical lines, and the number of chapter nodes is equal to the number of chapters in the sub-topic clusters corresponding to the equidistant vertical lines; each chapter node corresponds one-to-one with a chapter, and the keywords of the corresponding chapter are displayed next to the chapter node; when the user moves the mouse over the chapter node, the chapter research question and chapter research method will be displayed in a pop-up window; The document import module provides a document import interface and a query interface; users upload documents through the document import interface; users enter keywords through the query interface to query chapters in the database; users can view the rivers of each topic cluster drawn by the corpus extraction module through the query interface. The text decomposition module is used to split user-uploaded documents into several chapters and save them in the database; the text decomposition module parses each slide in a PPT document into a single slide document, and extracts the text content of the single slide document as a chapter; the text decomposition module splits a Word document into several fragment documents according to the rules of headings, lists, and line breaks, and extracts the text content of the fragment documents as a chapter. The mapping construction module is used to search for chapters associated with keywords input by the user; the mapping construction module measures the similarity between the keywords input by the user and the sub-cluster keywords by calculating cosine similarity; the mapping construction module sends the sub-cluster information of the sub-topic cluster with the highest similarity to the keywords input by the user to the writing and importing module; the sub-cluster information includes sub-cluster keywords, chapter IDs of all chapters in the sub-cluster, chapter keywords, time windows, chapter research questions, and chapter research methods; the chapter ID is the number of the chapter in the database, and the chapter ID is a unique identifier for the chapter; The writing and importing module feeds back the sub-cluster information to the user; the user can select several chapters to reference or discard the sub-cluster information through the interface provided by the writing and importing module; if the user selects several chapters to reference by clicking, the writing and importing module uses the obtained chapter ID of the selected chapter to query all document information of the selected chapter in the database and inserts all the queried document information into the document position specified by the user.

Images