Information extraction method and system for technical standard text of power system

Abstract

The invention discloses an information extraction method for a power system technical standard text. The method comprises the steps of obtaining data information of a target power system technical standard text; judging the type of the target power system technical standard text; if the type is the text type, line structure, segment structure, title and relation identification is carried out; if the text type is not the text type, identifying tables, pictures and metadata information; and completing information extraction of the technical standard text of the target power system according to the obtained identification information. The invention also discloses a system for realizing the information extraction method for the technical standard text of the power system. According to the method, the category of the target electric power system technical standard text is identified, and the information is extracted by adopting the corresponding technical scheme according to the corresponding text type, so that the information extraction of the electric power system technical standard text can be realized, the reliability is higher, and the accuracy is better.

Description

Technical Field

[0001] The invention belongs to the field of electrical automation, and in particular relates to an information extraction method and system for a power system technical standard text. Background Art

[0002] With the development of economy and technology and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and life, bringing endless convenience to people's production and life. Therefore, ensuring the stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] At present, the power system has formulated a large number of technical standard texts to ensure its own stable and reliable operation and standardization of operations. These technical standard texts can guide all aspects of the production and operation of the power system, thus greatly improving the safe and stable operation and standardization of operations of the power system. However, these technical standard texts have various formats, some are presented in text form, and some are presented in the form of pictures; currently, if the power system staff want to extract the information of such technical standard texts and conduct subsequent research, they can only rely on general structured tools.

[0004] However, the commonly used general structured tools often use simple image recognition to extract content, but this solution has poor reliability and accuracy and cannot meet the requirements of the power system for extracting technical standard text content. Summary of the invention

[0005] One of the purposes of the present invention is to provide a method for extracting information from a power system technical standard text with high reliability and good accuracy.

[0006] A second object of the present invention is to provide a system for implementing the information extraction method of the power system technical standard text.

[0007] The information extraction method of the power system technical standard text provided by the present invention comprises the following steps:

[0008] S1. Obtain data information of the target power system technical standard text;

[0009] S2. Determine the type of target power system technical standard text according to the data information obtained in step S1;

[0010] S3. If the target power system technical standard text is of text type, then identify the row structure, paragraph structure, title and relationship;

[0011] S4. If the target power system technical standard text is of non-text type, identify the table, picture and metadata information;

[0012] S5. Based on the obtained identification information, complete the information extraction of the target power system technical standard text.

[0013] The step S2 of determining the type of the target power system technical standard text according to the data information obtained in step S1 specifically includes the following steps:

[0014] According to the data information obtained in step S1, the type of the target power system technical standard text is determined; the type of the target power system technical standard text includes a text type and a non-text type.

[0015] If the target power system technical standard text is of text type, the following steps are performed to identify the row structure, segment structure, title and relationship:

[0016] Identify the line structure and paragraph structure of the target power system technical standard text to obtain the paragraph structure information of the target power system technical standard document;

[0017] Perform title recognition on the target power system technical standard text to obtain paragraph title information of the target power system technical standard document;

[0018] According to the obtained paragraph structure information and paragraph title information, the corresponding relationship is identified.

[0019] The above-mentioned step of performing line structure and paragraph structure recognition on the target power system technical standard text to obtain paragraph structure information of the target power system technical standard document specifically includes the following steps:

[0020] The following rules are used to determine the structure of each character group in the target power system technical standard text to obtain the row structure information:

[0021] Rule 1: The coordinate value of the horizontal coordinate where the upper boundary of the text is located and the coordinate value of the horizontal coordinate where the lower boundary of the text is located have an error less than the set value;

[0022] Rule 2: The error between the coordinate value of the ordinate where the upper boundary of the text is located and the coordinate value of the ordinate where the lower boundary of the text is located is less than the set value;

[0023] If both rule 1 and rule 2 are satisfied, the corresponding character group is considered to be a row structure;

[0024] Based on the obtained line structure information, the following rules are used to determine the paragraph structure and obtain the paragraph structure information:

[0025] Rule 1: There is an indent of the set length of characters on the first line;

[0026] Rule 2: Ending with a period or semicolon at the end;

[0027] Rule 3: The right end of the line reaches the edge of the page layout range;

[0028] If rule 1, rule 2 and rule 3 are satisfied at the same time, it is identified as a paragraph structure, and the start line and end line of the paragraph are obtained, thus completing the determination and division of the paragraph.

[0029] The above-mentioned title recognition of the target power system technical standard text to obtain the paragraph title information of the target power system technical standard document specifically includes the following steps:

[0030] Obtain data information of the first-level titles through the catalog page of the target power system technical standard text, and form a first-level title list;

[0031] Compare line by line and item by item within the page number of the target page record;

[0032] Second-level and above titles are judged by the regular expression of the content: if they start with the form of "number, dot, number" and there is no punctuation at the end of the corresponding line, they are judged as second-level and above titles;

[0033] The title's rank is determined by the numerical value.

[0034] The identification of the corresponding relationship based on the obtained paragraph structure information and paragraph title information specifically includes the following steps:

[0035] The paragraph whose vertical coordinate is between the i-th title and the i+1-th title is determined to be the main text content belonging to the i-th title, thereby completing the relationship identification between the paragraph and the paragraph title.

[0036] If the target power system technical standard text is a non-text type, the identification of tables, pictures and metadata information is performed in step S4, which specifically includes the following steps:

[0037] Perform table recognition on the target power system technical standard text to obtain data information of the table header and table content in the target power system technical standard text;

[0038] Performing image recognition on the target power system technical standard text to obtain the image title and image data information of the target power system technical standard text;

[0039] The metadata information of the target power system technical standard text is identified on the cover page to obtain the metadata information of the target power system technical standard text.

[0040] The aforementioned process of performing table recognition on the target power system technical standard text to obtain data information of the table header and table contents in the target power system technical standard text specifically includes the following steps:

[0041] Position the horizontal and vertical straight lines for the target power system technical standard text;

[0042] The horizontal straight lines and vertical straight lines obtained by positioning are used as the boundaries of the table;

[0043] According to the located intersection points of the horizontal straight lines and the vertical straight lines, a set of rectangles with the intersection points as vertices is located, thereby realizing the location of cells in the table;

[0044] Extract the content in each cell;

[0045] For the content extracted from the continuous cells, the row name and column name information are extracted according to the cell position, and the extracted information is stored in a table format;

[0046] Identify table headers: Perform regular expression judgment on the text before and after the table. If the text format is "table", "number", space, or centered, the corresponding content is determined to be a table header.

[0047] The above-mentioned performing image recognition on the target power system technical standard text to obtain the image title and image data information of the target power system technical standard text specifically includes the following steps:

[0048] For each page of the target power system technical standard text, an image extraction algorithm is used to extract the image and obtain the corresponding coordinate information;

[0049] Identify the picture title for the paragraph that is closest to the picture.

[0050] The method of identifying metadata information of the cover of the target power system technical standard text to obtain metadata information of the target power system technical standard text specifically includes the following steps:

[0051] According to the characteristics of metadata information of the target power system technical standard text, metadata information is identified;

[0052] Characteristics of metadata information include:

[0053] If the metadata item is the standard number, the characteristics are: character data, standard code + 1 space character + sequence number + 1 hyphen "-" + 4-digit release year number; the location of the standard number is the first page of the target power system technical standard text;

[0054] If the metadata item is an international standard classification number, the characteristics are: character data; the location of the international standard classification number is the first page of the target power system technical standard text;

[0055] If the metadata item is the Chinese Standard Classification Number, the features are: character data; the location of the Chinese Standard Classification Number is the first page of the target power system technical standard text;

[0056] If the metadata item is at the standard level, the characteristics are: if the value is GB, it is determined as a national standard; if the value is DL, it is determined as a line standard; if the value is T, it is determined as a group standard; if the value is Q, it is determined as an enterprise standard; if the value is DB, it is determined as a local standard; the position of the standard level is the homepage of the target power system technical standard text;

[0057] If the metadata item is the Chinese standard name, the features are: the location of the Chinese standard name is the first page of the target power system technical standard text;

[0058] If the metadata item is the English standard name, the features are: the location of the English standard name is the first page of the target power system technical standard text;

[0059] If the metadata item is the release date, the features are: the numerical format is a format text of YYYY-MM-DD; the location of the release date is the first page of the target power system technical standard text;

[0060] If the metadata item is the implementation or trial date, the features are: the numerical format is a text format of YYYY-MM-DD; the location of the implementation or trial date is the first page of the target power system technical standard text;

[0061] If the metadata item is a publishing organization, the characteristics are: the location of the publishing organization is the first page of the target power system technical standard text;

[0062] If the metadata item is the text language, the features are: the location of the text language is the first page of the target power system technical standard text;

[0063] If the metadata item is a standard category, the characteristics are: the location of the standard category is the first page of the target power system technical standard text; the standard category includes terms, symbols, classifications, tests, specifications, procedures and guidelines;

[0064] If the metadata item is the drafting rule, the characteristics are: the value is the standard number; the location of the standard category is the preface of the target power system technical standard text;

[0065] If the metadata item is a replacement standard, the features are: the value is the replacement standard number and the replacement standard name; the location of the replacement standard is the foreword of the target power system technical standard text;

[0066] If the metadata item is a major revision or change, the characteristics are: the location of the major revision or change is the foreword of the target power system technical standard text;

[0067] If the metadata item is the proposing unit, the features are: the location of the proposing unit is the preface of the target power system technical standard text;

[0068] If the metadata item is the responsible unit, the characteristics are: the location of the responsible unit is the preface of the target power system technical standard text;

[0069] If the metadata item is the drafting unit, the characteristics are: the location of the drafting unit is the foreword of the target power system technical standard text;

[0070] If the metadata item is the implementation unit, the characteristics are: the location of the implementation unit is the foreword of the target power system technical standard text;

[0071] If the metadata item is the drafter, the features are: the drafter’s position is the foreword of the target power system technical standard text;

[0072] If the metadata item is a specified range, the characteristics are: the location of the specified range is the main text of the target power system technical standard text;

[0073] If the metadata item is the scope of application, the characteristics are: the location of the scope of application is the main body of the target power system technical standard text;

[0074] If the metadata item is an inapplicable scope, the features are: the location of the inapplicable scope is the main text of the target power system technical standard text;

[0075] If the metadata item is a normative reference document, the characteristics are: including the standard number and text description; the location of the normative reference document is the main text of the target power system technical standard text;

[0076] If the metadata item is a reference document, the features are: the location of the reference document is the last page of the standard document and the copyright page of the target power system technical standard text;

[0077] If the metadata item is a publishing unit, the features are: the location of the publishing unit is the last page of the standard document and the copyright page of the target power system technical standard text;

[0078] If the metadata item is a unified book number, the feature is: character text; the location of the unified book number is the last page of the standard document and the copyright page of the target power system technical standard text.

[0079] The present invention also provides a system for implementing the information extraction method of the electric power system technical standard text, comprising a data acquisition module, a type determination module, a text extraction module, a non-text extraction module and an information extraction module; the data acquisition module and the type determination module are connected in series; the output end of the type determination module is simultaneously connected to the input end of the text extraction module and the non-text extraction module; the output ends of the text extraction module and the non-text extraction module are simultaneously connected to the information extraction module; the data acquisition module is used to acquire the data information of the target electric power system technical standard text, and upload the data information to the type determination module; the type determination module is used to determine the type of the target electric power system technical standard text according to the received data information and the acquired data information, and upload the data information to the text extraction module and the non-text extraction module; the text extraction module is used to perform line structure, segment structure, title and relationship recognition according to the received data information, if the target electric power system technical standard text is of text type, and upload the data information to the information extraction module; the non-text extraction module is used to perform table, picture and metadata information recognition according to the received data information, if the target electric power system technical standard text is of non-text type, and upload the data information to the information extraction module; the information extraction module is used to complete the information extraction of the target electric power system technical standard text according to the received data information and the obtained recognition information.

[0080] The method and system for extracting information from the electric power system technical standard text provided by the present invention identify the category of the target electric power system technical standard text and extract information using the corresponding technical solution according to the corresponding text type. Therefore, the present invention can not only realize the information extraction of the electric power system technical standard text, but also has higher reliability and better accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 The figure is a schematic diagram of the method flow of the present invention.

[0082] Figure 2 Schematic diagram of the functional modules of the system of the present invention. DETAILED DESCRIPTION

[0083] like Figure 1 The method flow chart of the method of the present invention is shown as follows: The information extraction method of the power system technical standard text disclosed in the present invention comprises the following steps:

[0084] S1. Obtain data information of the target power system technical standard text;

[0085] S2. According to the data information obtained in step S1, determine the type of the target power system technical standard text; specifically comprising the following steps:

[0086] According to the data information obtained in step S1, the type of the target power system technical standard text is determined; the type of the target power system technical standard text includes a text type and a non-text type;

[0087] S3. If the target power system technical standard text is of text type, then identify the row structure, segment structure, title and relationship; including the following steps:

[0088] The line structure and paragraph structure of the target power system technical standard text are identified to obtain the paragraph structure information of the target power system technical standard file; the specific steps include:

[0089] The following rules are used to determine the structure of each character group in the target power system technical standard text to obtain the row structure information:

[0090] Rule 1: The coordinate value of the horizontal coordinate where the upper boundary of the text is located and the coordinate value of the horizontal coordinate where the lower boundary of the text is located have an error less than the set value;

[0091] Rule 2: The error between the coordinate value of the ordinate where the upper boundary of the text is located and the coordinate value of the ordinate where the lower boundary of the text is located is less than the set value;

[0092] If both rule 1 and rule 2 are satisfied, the corresponding character group is considered to be a row structure;

[0093] Based on the obtained line structure information, the following rules are used to determine the paragraph structure and obtain the paragraph structure information:

[0094] Rule 1: There is an indent of the set length of characters on the first line;

[0095] Rule 2: Ending with a period or semicolon at the end;

[0096] Rule 3: The right end of the line reaches the edge of the page layout range;

[0097] If rules 1, 2, and 3 are met at the same time, it is identified as a paragraph structure, and the start line and end line of the paragraph are obtained, completing the determination and division of the paragraph;

[0098] The title recognition is performed on the target power system technical standard text to obtain the paragraph title information of the target power system technical standard document; the specific steps include:

[0099] Obtain data information of the first-level titles through the catalog page of the target power system technical standard text, and form a first-level title list;

[0100] Compare line by line and item by item within the page number of the target page record;

[0101] Second-level and above titles are judged by the regular expression of the content: if they start with the form of "number, dot, number" and there is no punctuation at the end of the corresponding line, they are judged as second-level and above titles;

[0102] Determine the level of the title by the numerical value;

[0103] According to the obtained paragraph structure information and paragraph title information, the corresponding relationship is identified; specifically, the following steps are included:

[0104] The paragraph whose ordinate is between the i-th title and the i+1-th title is determined to be the text content belonging to the i-th title, thereby completing the relationship identification between the paragraph and the paragraph title;

[0105] S4. If the target power system technical standard text is of non-text type, identify the table, picture and metadata information; specifically, the steps include:

[0106] Performing table recognition on the target power system technical standard text to obtain data information of the table header and table content in the target power system technical standard text; specifically comprising the following steps:

[0107] Position the horizontal and vertical straight lines for the target power system technical standard text;

[0108] The horizontal straight lines and vertical straight lines obtained by positioning are used as the boundaries of the table;

[0109] According to the located intersection points of the horizontal straight lines and the vertical straight lines, a set of rectangles with the intersection points as vertices is located, thereby realizing the location of cells in the table;

[0110] Extract the content in each cell;

[0111] For the content extracted from the continuous cells, the row name and column name information are extracted according to the cell position, and the extracted information is stored in a table format;

[0112] Identify table headers: Perform regular expression judgment on the text before and after the table. If the text format is "table", "number", space, or centered, the corresponding content is determined to be a table header.

[0113] Perform image recognition on the target power system technical standard text to obtain the image title and image data information of the target power system technical standard text; specifically, the steps include:

[0114] For each page of the target power system technical standard text, an image extraction algorithm is used to extract the image and obtain the corresponding coordinate information;

[0115] Identify the picture title for the paragraph closest to the picture;

[0116] The metadata information of the cover of the target power system technical standard text is identified to obtain the metadata information of the target power system technical standard text; the specific steps include:

[0117] According to the characteristics of metadata information of the target power system technical standard text, metadata information is identified;

[0118] Characteristics of metadata information include:

[0119] If the metadata item is the standard number, the characteristics are: character data, standard code + 1 space character + sequence number + 1 hyphen "-" + 4-digit release year number; the location of the standard number is the first page of the target power system technical standard text;

[0120] If the metadata item is an international standard classification number, the characteristics are: character data; the location of the international standard classification number is the first page of the target power system technical standard text;

[0121] If the metadata item is the Chinese Standard Classification Number, the features are: character data; the location of the Chinese Standard Classification Number is the first page of the target power system technical standard text;

[0122] If the metadata item is at the standard level, the characteristics are: if the value is GB, it is determined as a national standard; if the value is DL, it is determined as a line standard; if the value is T, it is determined as a group standard; if the value is Q, it is determined as an enterprise standard; if the value is DB, it is determined as a local standard; the position of the standard level is the homepage of the target power system technical standard text;

[0123] If the metadata item is the Chinese standard name, the feature is: free text data, and the location of the Chinese standard name is the first page of the target power system technical standard text;

[0124] If the metadata item is the English standard name, the features are: free text data, the location of the English standard name is the first page of the target power system technical standard text;

[0125] If the metadata item is the release date, the features are: the numerical format is a format text of YYYY-MM-DD; the location of the release date is the first page of the target power system technical standard text;

[0126] If the metadata item is the implementation or trial date, the features are: the numerical format is a text format of YYYY-MM-DD; the location of the implementation or trial date is the first page of the target power system technical standard text;

[0127] If the metadata item is the issuing agency, the characteristics are: free text data, and the location of the issuing agency is the first page of the target power system technical standard text;

[0128] If the metadata item is the text language, the features are: language type such as Chinese / English, and the location of the text language is the homepage of the target power system technical standard text;

[0129] If the metadata item is a standard category, the characteristics are: the location of the standard category is the first page of the target power system technical standard text; the standard category includes terms, symbols, classifications, tests, specifications, procedures and guidelines;

[0130] If the metadata item is the drafting rule, the characteristics are: the value is the standard number; the location of the standard category is the preface of the target power system technical standard text;

[0131] If the metadata item is a replacement standard, the features are: the value is the replacement standard number and the replacement standard name; the location of the replacement standard is the foreword of the target power system technical standard text;

[0132] If the metadata item is a major revision or change, the features are: free text data, with ";" separating different entries, and the location of the major revision or change is the preface of the target power system technical standard text;

[0133] If the metadata item is the proposed unit, the characteristics are: free text data, different items are separated by ";", and the location of the proposed unit is the preface of the target power system technical standard text;

[0134] If the metadata item is a centralized unit, the characteristics are: free text data, different items are separated by ";", and the location of the centralized unit is the preface of the target power system technical standard text;

[0135] If the metadata item is the drafting unit, the characteristics are: free text data, different items are separated by ";", and the location of the drafting unit is the foreword of the target power system technical standard text;

[0136] If the metadata item is an implementation unit, the characteristics are: free text data, different items are separated by ";", and the location of the implementation unit is the preface of the target power system technical standard text;

[0137] If the metadata item is the drafter, the features are: free text data, the position of the drafter between the names of the people is the foreword of the target power system technical standard text;

[0138] If the metadata item is a specified range, the feature is: free text data, the location of the specified range is the main text of the target power system technical standard text;

[0139] If the metadata item is scope of application, the features are: free text data, the location of the scope of application is the main body of the target power system technical standard text;

[0140] If the metadata item is a non-applicable scope, the feature is: free text data, the location of the non-applicable scope is the main text of the target power system technical standard text;

[0141] If the metadata item is a normative reference document, the characteristics are: free text data, including the standard number and text description; the location of the normative reference document is the main text of the target power system technical standard text;

[0142] If the metadata item is a reference document, the features are: free text data, which may include the standard number and standard name, and the items are separated by ";"; the location of the reference document is the last page of the standard document and the copyright page of the target power system technical standard text;

[0143] If the metadata item is the publishing unit, the features are: free text data, the location of the publishing unit is the last page of the standard document and the copyright page of the target power system technical standard text;

[0144] If the metadata item is a unified book number, the features are: character text; the location of the unified book number is the last page of the standard document and the copyright page of the target power system technical standard text;

[0145] S5. Based on the obtained identification information, complete the information extraction of the target power system technical standard text.

[0146] like Figure 2 The functional module schematic diagram of the system of the present invention is shown as follows: the system disclosed in the present invention for realizing the information extraction method of the electric power system technical standard text comprises a data acquisition module, a type determination module, a text extraction module, a non-text extraction module and an information extraction module; the data acquisition module and the type determination module are connected in series; the output end of the type determination module is simultaneously connected to the input end of the text extraction module and the non-text extraction module; the output ends of the text extraction module and the non-text extraction module are simultaneously connected to the information extraction module; the data acquisition module is used to acquire the data information of the target electric power system technical standard text, and upload the data information to the type determination module; the type determination module is used to acquire the data information according to the acquired data information according to the received data information Data information, determine the type of the target power system technical standard text, and upload the data information to the text extraction module and the non-text extraction module; the text extraction module is used to identify the row structure, segment structure, title and relationship according to the received data information, if the target power system technical standard text is of text type, and upload the data information to the information extraction module; the non-text extraction module is used to identify the table, picture and metadata information according to the received data information, if the target power system technical standard text is of non-text type, and upload the data information to the information extraction module; the information extraction module is used to complete the information extraction of the target power system technical standard text according to the received data information and the obtained identification information.

Claims

1. A method for extracting information from a power system technical standard text, comprising the following steps: S1. Obtain data information of the target power system technical standard text; S2. Determine the type of target power system technical standard text according to the data information obtained in step S1; S3. If the target power system technical standard text is of text type, then identify the line structure, paragraph structure, title and relationship; S4. If the target power system technical standard text is of non-text type, identify the table, picture and metadata information; S5. Based on the obtained identification information, complete the information extraction of the target power system technical standard text.

2. The information extraction method of the power system technical standard text according to claim 1 is characterized in that The step S2 of determining the type of the target power system technical standard text according to the data information obtained in step S1 specifically includes the following steps: According to the data information obtained in step S1, the type of the target power system technical standard text is determined; the type of the target power system technical standard text includes a text type and a non-text type; If the target power system technical standard text is of text type, the following steps are performed to identify the row structure, segment structure, title and relationship: Identify the line structure and paragraph structure of the target power system technical standard text to obtain the paragraph structure information of the target power system technical standard document; Perform title recognition on the target power system technical standard text to obtain paragraph title information of the target power system technical standard document; According to the obtained paragraph structure information and paragraph title information, the corresponding relationship is identified.

3. The information extraction method of the power system technical standard text according to claim 2 is characterized in that The above-mentioned step of performing line structure and paragraph structure recognition on the target power system technical standard text to obtain paragraph structure information of the target power system technical standard document specifically includes the following steps: The following rules are used to determine the structure of each character group in the target power system technical standard text to obtain the row structure information: Rule 1: The coordinate value of the horizontal coordinate where the upper boundary of the text is located and the coordinate value of the horizontal coordinate where the lower boundary of the text is located have an error less than the set value; Rule 2: The error between the coordinate value of the ordinate where the upper boundary of the text is located and the coordinate value of the ordinate where the lower boundary of the text is located is less than the set value; If both rule 1 and rule 2 are satisfied, the corresponding character group is considered to be a row structure; Based on the obtained line structure information, the following rules are used to determine the paragraph structure and obtain the paragraph structure information: Rule 1: There is an indent of the set length of characters on the first line; Rule 2: Ending with a period or semicolon at the end; Rule 3: The right end of the line reaches the edge of the page layout range; If rule 1, rule 2 and rule 3 are satisfied at the same time, it is identified as a paragraph structure, and the start line and end line of the paragraph are obtained, thus completing the determination and division of the paragraph.

4. The information extraction method of the power system technical standard text according to claim 3 is characterized in that The above-mentioned title recognition of the target power system technical standard text to obtain the paragraph title information of the target power system technical standard document specifically includes the following steps: Obtain data information of the first-level titles through the catalog page of the target power system technical standard text, and form a first-level title list; Compare line by line and item by item within the page number of the target page record; Second-level and above titles are judged by the regular expression of the content: if they start with the form of "number, dot, number" and there is no punctuation at the end of the corresponding line, they are judged as second-level and above titles; The title's rank is determined by the numerical value.

5. The information extraction method of the power system technical standard text according to claim 4 is characterized in that The identification of the corresponding relationship based on the obtained paragraph structure information and paragraph title information specifically includes the following steps: The paragraph whose vertical coordinate is between the i-th title and the i+1-th title is determined to be the main text content belonging to the i-th title, thereby completing the relationship identification between the paragraph and the paragraph title.

6. The information extraction method of the power system technical standard text according to claim 5 is characterized in that If the target power system technical standard text is a non-text type, the identification of tables, pictures and metadata information is performed in step S4, which specifically includes the following steps: Perform table recognition on the target power system technical standard text to obtain data information of the table header and table content in the target power system technical standard text; Performing image recognition on the target power system technical standard text to obtain the image title and image data information of the target power system technical standard text; The metadata information of the target power system technical standard text is identified on the cover page to obtain the metadata information of the target power system technical standard text.

7. The information extraction method of the power system technical standard text according to claim 6 is characterized in that The aforementioned process of performing table recognition on the target power system technical standard text to obtain data information of the table header and table contents in the target power system technical standard text specifically includes the following steps: Position the horizontal and vertical straight lines for the target power system technical standard text; The horizontal straight lines and vertical straight lines obtained by positioning are used as the boundaries of the table; According to the located intersection points of the horizontal straight lines and the vertical straight lines, a set of rectangles with the intersection points as vertices is located, thereby realizing the location of cells in the table; Extract the content in each cell; For the content extracted from the continuous cells, the row name and column name information are extracted according to the cell position, and the extracted information is stored in a table format; Identify table headers: Perform regular expression judgment on the text before and after the table. If the text format is "table", "number", space, or centered, the corresponding content is determined to be a table header.

8. The information extraction method of the power system technical standard text according to claim 7 is characterized in that The above-mentioned performing image recognition on the target power system technical standard text to obtain the image title and image data information of the target power system technical standard text specifically includes the following steps: For each page of the target power system technical standard text, an image extraction algorithm is used to extract the image and obtain the corresponding coordinate information; Identify the picture title for the paragraph that is closest to the picture.

9. The method for extracting information from the power system technical standard text according to claim 8, characterized in that The method of identifying metadata information of the cover of the target power system technical standard text to obtain metadata information of the target power system technical standard text specifically includes the following steps: According to the characteristics of metadata information of the target power system technical standard text, metadata information is identified; Characteristics of metadata information include: If the metadata item is the standard number, the characteristics are: character data, standard code + 1 space character + sequence number + 1 hyphen "-" + 4-digit release year number; the location of the standard number is the first page of the target power system technical standard text; If the metadata item is an international standard classification number, the characteristics are: character data; the location of the international standard classification number is the first page of the target power system technical standard text; If the metadata item is the Chinese Standard Classification Number, the features are: character data; the location of the Chinese Standard Classification Number is the first page of the target power system technical standard text; If the metadata item is at the standard level, the characteristics are: if the value is GB, it is determined as a national standard; if the value is DL, it is determined as a line standard; if the value is T, it is determined as a group standard; if the value is Q, it is determined as an enterprise standard; if the value is DB, it is determined as a local standard; the position of the standard level is the homepage of the target power system technical standard text; If the metadata item is the Chinese standard name, the features are: the location of the Chinese standard name is the first page of the target power system technical standard text; If the metadata item is the English standard name, the features are: the location of the English standard name is the first page of the target power system technical standard text; If the metadata item is the release date, the features are: the numerical format is a format text of YYYY-MM-DD; the location of the release date is the first page of the target power system technical standard text; If the metadata item is the implementation or trial date, the features are: the numerical format is a text format of YYYY-MM-DD; the location of the implementation or trial date is the first page of the target power system technical standard text; If the metadata item is a publishing organization, the characteristics are: the location of the publishing organization is the first page of the target power system technical standard text; If the metadata item is the text language, the features are: the location of the text language is the first page of the target power system technical standard text; If the metadata item is a standard category, the characteristics are: the location of the standard category is the first page of the target power system technical standard text; the standard category includes terms, symbols, classifications, tests, specifications, procedures and guidelines; If the metadata item is the drafting rule, the characteristics are: the value is the standard number; the location of the standard category is the preface of the target power system technical standard text; If the metadata item is a replacement standard, the features are: the value is the replacement standard number and the replacement standard name; the location of the replacement standard is the foreword of the target power system technical standard text; If the metadata item is a major revision or change, the characteristics are: the location of the major revision or change is the foreword of the target power system technical standard text; If the metadata item is the proposing unit, the features are: the location of the proposing unit is the preface of the target power system technical standard text; If the metadata item is the responsible unit, the characteristics are: the location of the responsible unit is the preface of the target power system technical standard text; If the metadata item is the drafting unit, the features are: the location of the drafting unit is the foreword of the target power system technical standard text; If the metadata item is the implementation unit, the characteristics are: the location of the implementation unit is the foreword of the target power system technical standard text; If the metadata item is the drafter, the features are: the drafter’s position is the foreword of the target power system technical standard text; If the metadata item is a specified range, the characteristics are: the location of the specified range is the main text of the target power system technical standard text; If the metadata item is the scope of application, the characteristics are: the location of the scope of application is the main body of the target power system technical standard text; If the metadata item is an inapplicable scope, the features are: the location of the inapplicable scope is the main text of the target power system technical standard text; If the metadata item is a normative reference document, the characteristics are: including the standard number and text description; the location of the normative reference document is the main text of the target power system technical standard text; If the metadata item is a reference document, the features are: the location of the reference document is the last page of the standard document and the copyright page of the target power system technical standard text; If the metadata item is a publishing unit, the features are: the location of the publishing unit is the last page of the standard document and the copyright page of the target power system technical standard text; If the metadata item is a unified book number, the features are: character text; the location of the unified book number is the last page of the standard document and the copyright page of the target power system technical standard text.

10. A system for implementing the information extraction method of the power system technical standard text according to any one of claims 1 to 9, characterized in that It includes a data acquisition module, a type determination module, a text extraction module, a non-text extraction module and an information extraction module; the data acquisition module and the type determination module are connected in series; the output end of the type determination module is simultaneously connected to the input end of the text extraction module and the non-text extraction module; the output end of the text extraction module and the non-text extraction module is simultaneously connected to the information extraction module; the data acquisition module is used to obtain the data information of the target power system technical standard text, and upload the data information to the type determination module; The type determination module is used to determine the type of the target power system technical standard text based on the received data information and the acquired data information, and upload the data information to the text extraction module and the non-text extraction module; the text extraction module is used to identify the row structure, segment structure, title and relationship based on the received data information if the target power system technical standard text is of text type, and upload the data information to the information extraction module; the non-text extraction module is used to identify the table, picture and metadata information based on the received data information if the target power system technical standard text is of non-text type, and upload the data information to the information extraction module; the information extraction module is used to complete the information extraction of the target power system technical standard text based on the received data information and the obtained identification information.

Images