TWI629605B - Data extracting method for portable document format file corresponding to credit record of user and personal credit analysis system - Google Patents

Abstract

A portable file format file suitable for a credit record corresponding to a user The method of data acquisition. The method includes performing vector conversion on the portable file format file to convert the portable file format file into a vector file; and using the vector according to multiple layers of the portable file format file The plurality of text content of the file is divided into a first type of text content and a second type of text content; and the vector file is cleaned to convert the vector file into a text file; and from the text file Multiple fields in the target chapters of the chapters draw multiple pieces of data.

Description

Portable file format for corresponding user's credit history File data extraction method and personal credit analysis system

The present invention relates to a data extraction method, and more particularly to a data acquisition method for a portable file format file for a credit record corresponding to a user.

With the advancement of technology and the booming of personal finance, users will begin to efficiently search for personal credit ratings and corresponding information through the Internet (eg, websites or institutions that offer credit ratings). For example, some organizations or websites will ask users to apply for a paper report and use the method of mailing a paper report and manually to conduct a credit rating. After the relevant program is gradually electronicized, the user can download the personal credit report to the association center first, and fill in the corresponding information to the website responsible for the credit rating by the content of the downloaded personal credit report. Or, the user directly uploads the downloaded personal credit report to the website responsible for the rating, and the website enters the credit rating according to the uploaded personal credit report. Related information. In this way, the website can calculate the credit rating of the user based on the information related to the credit rating.

However, the above-mentioned manner of filling in the corresponding information related to the credit rating according to the content of the personal credit report is time consuming and has the possibility of input errors. Therefore, how to efficiently extract relevant information from personal credit reports has become the goal of the development of personnel in this field.

The invention provides a data extraction method for efficiently extracting data from a portable file format file corresponding to a user's credit record.

An embodiment of the present invention provides a data extraction method, which is applicable to a portable file format file corresponding to a user's credit record. The method includes performing vector conversion on the portable file format file to convert the portable file format file into a vector file, wherein the portable file format file has a plurality of layers respectively configured a plurality of pieces of text content, wherein the plurality of pieces of text content of the vector file are the plurality of pieces of text content of the portable file format file disposed in the plurality of layers, and the plurality of pieces of the vector file The text content may be circled; the plurality of text content of the vector file is divided into the first type of text content and the second type of text content according to the plurality of layers of the portable file format file; Performing a cleaning operation on the vector file to convert the vector file into a text file; and extracting a plurality of fields from a target chapter of the plurality of chapters of the text file belonging to the first type of text content Pen data.

Based on the above, the data extraction method provided by an embodiment of the present invention can convert a portable file format file corresponding to a user's credit record into a vector file, and then perform a cleaning operation on the vector file to obtain only A text file containing specific content, which efficiently retrieves data recorded in multiple fields from the target chapter.

The above described features and advantages of the invention will be apparent from the following description.

100‧‧‧personal credit analysis system

110‧‧‧Vector Conversion Module

120‧‧‧cleaning module

130‧‧‧Data Capture Module

140‧‧‧User Profile Database

150‧‧‧Credit rating module

200‧‧‧Personal Credit Report PDF File

S201, S203, S205, S207‧‧‧ Process steps of data acquisition method

S2031, S2033‧‧‧ Process steps of step S203

S2051, S2053, S2055, S2057‧‧ ‧ process steps of step S205

S2071, S2073, S2075‧‧‧ Process steps of step S207

S20731, S20733, S20735, S20737, S20739‧‧‧Step S2073

701‧‧‧ text content at the top of the table

702, 703‧‧‧ text content of the watermark layer

801‧‧‧ Target chapter

810‧‧‧Name of the target chapter, the first line of text of the target chapter

820‧‧‧The second line of text in the target chapter

The third line of text of the target section of 821‧‧‧

The fourth line of text of the target section of 822‧‧‧

1 is a block diagram of a personal credit analysis system in accordance with an embodiment of the present invention.

2 is a flow chart of a data capture method according to an embodiment of the invention.

FIG. 3 is a flow chart of step S203 of FIG. 2 according to an embodiment of the invention.

FIG. 4 is a flow chart of step S205 of FIG. 2 according to an embodiment of the invention.

FIG. 5 is a flow chart of step S207 of FIG. 2 according to an embodiment of the invention.

FIG. 6 is a flow chart of step S2073 of FIG. 5 according to an embodiment of the invention.

7 is a partial personal credit report depicted in accordance with an embodiment of the present invention.

FIG. 8 is a diagram showing a chapter of a text file obtained after cleaning a vector file converted from a personal credit report according to an embodiment of the invention.

For convenience of explanation, assume that the user uploads a file related to the credit history of the user to a website that provides a personal credit rating. The file is, for example, a personal credit report file downloaded by a user from a financial joint credit center of the corporation (or other institution that can provide a user's credit history), wherein the file is a portable document format (Portable Document) Format, PDF). The website receives the file using a personal credit analysis system of a server erected on the website, and can retrieve relevant information from the received file to determine the credit rating of the user.

1 is a block diagram of a personal credit analysis system in accordance with an embodiment of the present invention.

Referring to FIG. 1 , in the embodiment, the personal credit analysis system 100 includes a vector conversion module 110 , a cleaning module 120 , a data capture module 130 , a user profile database 140 , and a credit rating module 150 . In addition, the input module (not shown) of the personal credit analysis system receives the personal credit report portable file format file 200 (hereinafter also referred to as the personal credit report PDF file 200) from the user.

In this embodiment, the personal credit analysis system is converted via a vector An application program of a plurality of code modules, such as a module 110, a cleaning module 120, a data capture module 130, a user profile database 140, and a credit rating module 150, and the personal credit is provided. The processing unit of the server of the rated website can access and execute the personal credit analysis system 100. However, in other embodiments, the personal credit analysis system 100 can also be implemented as an application (APP) that can be installed on a mobile device to execute on the mobile device. Alternatively, the plurality of code modules of the personal credit analysis system 100 can be implemented in a hardware manner. For example, the vector conversion module 110 can be implemented as a vector conversion circuit unit having the function of the vector conversion module 110. By analogy, the cleaning module 120, the data acquisition module 130, and the credit rating module 150 of all the personal credit analysis systems can be implemented as a cleaning circuit unit, a data acquisition circuit unit, and a credit rating circuit unit, respectively, and the user. The archive database 140 can be stored in a storage circuit unit coupled to the data capture circuit unit and the credit rating circuit unit.

The function of each module of each human credit analysis system and the data acquisition method provided by this embodiment will be described below with reference to FIG. 1 and FIG. 2 .

2 is a flow chart of a data capture method according to an embodiment of the invention. Referring to FIG. 1 and FIG. 2 simultaneously, in step S201, the vector conversion module 110 performs vector conversion on the portable file format file to convert the portable file format file into a vector file. The vector file is, for example, a Scalable Vector Graphics (SVG) file. It should be noted that the present invention is not limited to the method/program for performing vector conversion.

More specifically, when the vector conversion module 110 receives the upload by the user The personal credit report PDF file 200, the vector conversion module 110 converts the personal credit report PDF file 200 into a vector file. In general, the personal credit report PDF file 200 will have content (eg, image content or/and text content) that is configured in multiple layers. The plurality of layers are, for example, a body layer, a watermark layer, a header layer, and a tail layer. The main body layer has a plurality of chapters, and the plurality of chapters respectively record various information of the personal credit report of the user, and have a plurality of lines of text content. The watermark layer records a specific pattern or content, for example, a pattern for anti-counterfeiting (such as the pattern 703 shown in FIG. 7), and a code for indicating the identity of the user (eg, as shown in FIG. 7). Code 702), or file number. In addition, the content of the watermark layer mostly overlaps with the text content of the main layer. The header layer (eg, the header layer 701 illustrated in FIG. 7) and the footer layer are used to record other auxiliary information in the personal credit report PDF file 200 that does not affect the credit rating of the user. For example, the page number, the date and time the file was created, the name of the institution that provided the credit history, and so on.

The multi-line text content of the converted vector file is the multi-line text content of the personal credit report PDF file 200 disposed in the plurality of layers. The plurality of lines of text content of the vector file may be circled.

The vector module 110 transmits the converted vector file to the cleaning module 120.

In step S203, the cleaning module 120 divides the multi-line text content of the vector file into a first type of text content and a second type of text content according to multiple layers of the portable file format file. The following will be explained in conjunction with Figure 3. Sub-method.

FIG. 3 is a flow chart of step S203 of FIG. 2 according to an embodiment of the invention. Referring to FIG. 3, in step S2031, the cleaning module 120 compares the text content of the watermark layer corresponding to the watermark layer of the multi-line text content of the vector file, the text content of the corresponding header layer, and the text content of the corresponding footer layer. Divided into the second type of text content. Next, in step S2033, the cleaning module 120 divides other text content of the plurality of lines of text content of the vector file that is not the second type of text content into the first type of text content.

Specifically, the cleaning module 120 firstly uses the text content of the text file of the vector file that is not used to determine the user's credit rating (eg, text or image on the floating layer, the header layer, and the foot layer). The content is divided into the second type of text content, and the text content (eg, the text content of the main layer) used to determine the credit rating of the user is divided into the first type of text content. Then, the cleaning module 120 performs further information filtering based on the first type of text content to retrieve data that can be used to determine the user's credit rating.

For example, please return to FIG. 2. In step S205, the cleaning module 120 performs a cleaning operation on the vector file to convert the vector file into a text file. Further details of the cleaning operation will be described in detail below with reference to FIG.

FIG. 4 is a flow chart of step S205 of FIG. 2 according to an embodiment of the invention. Referring to FIG. 4, in step S2051, the cleaning module 120 stores part or all of the plurality of lines of text content of the vector file to the same layer of the text file. In step S2053, the cleaning module 120 will be from the text file. A plurality of chapters are identified in the multi-line text content belonging to the first type of text content. In more detail, the cleaning module 120 will perform subsequent cleaning operations only for the first type of text content. For example, assume that the first type of text content is all of the text content of the body layer of the personal credit report PDF file 200. The cleaning module 120 recognizes a plurality of sections of the first type of text content (ie, all text content of the main layer), and according to the separated horizontal lines of the plurality of chapters (eg, double solid lines, Dash lines, single solid lines) to distinguish each chapter, and identify the name of each chapter. For example, the cleaning module 120 can identify the name of each chapter by using the specific characters "[" and "] in the first line of text content of each chapter. For example, the text content "[Credit Card Information] recorded in one of the sections of the main layer of the personal credit report PDF file 200 can be identified as the "credit card information".

Then, in step S2055, the cleaning module 120 divides the plurality of chapters into first-type chapters having no multiple fields according to the types of the plurality of chapters of the first type of text content, and has The second type of chapter for multiple fields. The first type of chapter and the second type of chapter are used to determine a user's credit rating. Specifically, the cleaning module 120 stores a predetermined rule table, wherein the predetermined rule records all chapter names that may appear in the personal credit report PDF file 200, the total number of fields corresponding to each chapter name, and The character type of each field corresponding to each chapter name. It should be noted that the predetermined rule table is formulated in accordance with the format of the personal credit report PDF file 200.

In another embodiment, the cleaning module 120 can directly record the chapter name. The text content of the next line of the text content is identified to identify the field name, the total number of fields, and the character type of the plurality of fields corresponding to each chapter name.

Next, in step S2057, the cleaning module 120 correspondingly marks the first type chapter and the second type chapter according to the names of the first type chapter and the plurality of chapters of the second type chapter. a plurality of lines of text content of a portion of the plurality of chapters, wherein the first type of chapters and the plurality of lines of the plurality of chapters of the second type of chapters that are not marked are separated by a plurality of lines of text content line. Specifically, step S2057 can be regarded as the cleaning module 120 only marking the first type and the second type of the text belonging to the first type of text content used to determine the user's credit rating, and not marking the separation thereof. Horizontal lines (or do not mark other text content that is not used to judge the user's credit rating). All of the marked text content will be selected later, and the corresponding data will be extracted. In other words, it can be seen that the cleaning operation performed by the cleaning module 120 can (through a non-marking manner) let the text content of the floating layer, the header layer, and the tail layer be compared with other credits that are not used to judge the user. The text content (eg, the horizontal line) will not be captured.

In this embodiment, the plurality of lines of text content of the text file may be sequentially arranged according to the plurality of layers of the personal credit report PDF file 200. For example, the multi-line text content of the text file sequentially records the multi-line text content of the header layer, the foot layer, the floating layer, and the main layer of the personal credit report PDF file 200 in the vector file. And the multi-line text content of the text file is configured on the same layer (does not overlap each other). It should be noted that the order in which the plurality of lines of text content corresponding to the plurality of layers in the text file are not limited to the embodiment. For example, in another embodiment, the plurality of lines of text content of the text file may also sequentially record multiple rows of the floating layer, the header layer, the footer layer, and the main layer of the personal credit report PDF file 200. The text is in a text file.

Referring back to FIG. 2, in step S207, the data retrieval module 130 may extract the plurality of fields from the target chapters of the plurality of chapters of the first type of text content of the text file. Multiple pieces of data recorded in multiple fields. The details of step S207 will be described in detail below with reference to FIG.

FIG. 5 is a flow chart of step S207 of FIG. 2 according to an embodiment of the invention. Referring to FIG. 5, in step S2071, the data capture module 130 selects the marked plurality of lines of text content of the target chapter from the text file. In other words, the data capture module 130 only selects the text content from the previously marked text content. In this way, you can avoid using useless information. The target chapter is used to indicate the chapter that is currently being retrieved. The manner of selecting the target chapter may be selected according to the order of the plurality of chapters, or may be selected according to a specific rule of the corresponding credit rating module 150 or the user profile database 140, and the present invention is not limited thereto.

After selecting the marked multi-line text content of the target chapter of the data to be retrieved, the data capture module 130 performs the subsequent data capture operation according to the target chapter belonging to the first type chapter or the second type chapter. . In other words, the data capture module 130 may use different operations for capturing data according to whether the target chapter has multiple fields.

For example, in step S2073, if the target chapter belongs to the second type chapter The data capture module 130 separately divides each line of text content of the selected plurality of lines of text content into a plurality of fields according to the selected plurality of lines of text content of the target chapter, Extracting the plurality of pieces of data recorded by the plurality of fields from the plurality of fields. The text content recorded in the first row of the plurality of rows belonging to the same field is the column name of the same field, and the text content recorded by the other rows in the rows belonging to the same field is the same The name of the field name for the field.

For another example, in step S2075, if the target chapter belongs to the first type chapter, the data capture module 130 identifies the name of the target chapter from the selected plurality of lines of text content, which will be the target chapter. The text content of the name is identified as a chapter name field, and the other text content of the selected multi-line text content is recognized as a chapter content material corresponding to the chapter name field, and the chapter content material is retrieved. For example, in the "[Bank Loan Information] check database in Figure 7, there is no Taiwanese end of the end of April, 105, the balance of loans in various financial institutions in the country" is the text of the target chapter. The data capture module 130 determines that the target chapter belongs to the first type of chapter, and the data capture module 130 does not distinguish between multiple fields of the text content of the target chapter. The data capture module 130 will identify the chapter name field as "[bank loan information]", and retrieve the chapter content data corresponding to the chapter name field. "The database is not in the database at the end of April, 105. Balance".

The details of step S2073 will be described in detail below with reference to FIGS. 6, 7, and 8.

FIG. 6 is a diagram S2073 of FIG. 5 according to an embodiment of the invention. Flow chart. 7 is a partial personal credit report depicted in accordance with an embodiment of the present invention. FIG. 8 is a diagram showing a chapter of a text file obtained after cleaning a vector file converted from a personal credit report according to an embodiment of the invention.

Please refer to FIG. 6, FIG. 7, and FIG. 8 at the same time. In the present embodiment, it is assumed that the target chapter 801 in the plurality of chapters of the converted text file is the chapter "credit card information". As can be seen from FIG. 8, the target chapter 801 has four lines of text content, which are "[credit card information]" of the first line 810, and "the card issuer card name quota card issue date deactivation date use status" of the second line 820, In the third line 821, "Yuanda Bank VISA Puka 64 105/02/15 112/02/15 in use" and the fourth line 822 "Yuanda Bank VISA Platinum Card 100 106/01/03 115/01/03 Using". There is a space between each line of the content.

In step S20731, the data capture module 130 identifies the name of the target chapter based on the first line of text content of the selected plurality of lines of text content. As described above, the data capture module 130 can utilize a particular character (eg, "[" and "]") to identify the chapter name in the first line of text content of the multi-line text content of the target chapter.

In step S20733, the data capture module 130 queries the predetermined rule table for the total number of fields of the target chapter according to the name of the target chapter, or according to the selected plurality of lines of text content. Two lines of text content to identify the total number of the fields, wherein the second line of text content according to the selected plurality of lines of text content is used to represent a plurality of field names of the target chapter, wherein each of the plurality of fields One There is a separator character between the field names. Specifically, the data capture module 130, as described above, can recognize that the name of the target chapter 801 is “credit card information”, and is queried from the predetermined rule table according to “credit card information”, corresponding to “credit card information”. The total number of columns in the chapter is "6". That is, the data capture module 130 can recognize that the text content of the line other than the first line of the target chapter can be divided into six fields (ie, the card issuing institution, the card name, the credit amount, the card issuing date, and the expiration date). There are 6 fields in the use status). However, in another embodiment, the data capture module 130 may not query the predetermined rule table, but the data capture module 130 directly recognizes the text content of the next line of the chapter name (eg, the second line of text content). A field name for multiple fields, and each of the multiple field names has a (eg, space) separator character. For example, the second line of text of the target chapter is “issuing institution, card name, credit, date of issuance, date of expiration, status of use”, where “issuing agency”, “card name”, “credit”, “issuing card” Date, Deactivation Date, Usage Status are field names and are separated by a space. The data capture module 130 determines that the target chapter has six fields based on the number of the field names. That is, the data capture module 130 determines that the total number of fields in the target chapter is "6". In an embodiment, the data capture module 130 may add 1 to the total number of separator characters of the second line of text content as the total number of fields of the target chapter.

In step S20735, the data capture module 130 identifies the character type corresponding to each of the plurality of fields according to each of the plurality of field names. For example, the data capture module 130 identifies, according to the predetermined rule table, a field name corresponding to the “issuing institution”, and the character type of the data recorded in the field is a Chinese character; the corresponding is the “card name”. The name of the field, the character type of the data recorded in this field is Chinese, English character or space; corresponding to the name of the "number" field, the character type of the data recorded in this field is a numeric character; the field name corresponding to the "issuing date", recorded in this field The character type of the data is a numeric character or a slash character; the field name corresponding to the "deactivation date", and the character type of the data recorded in the field is a numeric character or a slash character; For the field name of "Usage Status", the character type of the data recorded in this field is Chinese character.

It is worth mentioning that in this embodiment, the data corresponding to the partial fields may be allowed to be blank data, that is, some fields do not record any data. The blank data will also be recorded to the user profile database.

Next, in step S20737, the data capture module 130 divides the third line of text content of the selected plurality of lines of text content into each line of the text content of the last line of text content into a plurality of columns according to the separation character. a bit such that the number of the plurality of fields distinguished by the line of text is equal to the total number of the fields of the target chapter, and the plurality of lines of the text content are differentiated The data of the fields is in accordance with the character type of the field to which the data belongs.

Specifically, in this embodiment, the data capture module 130 also uses a separator character (eg, a space) to distinguish multiple fields of other rows of the target chapter 801 (eg, the third row to the last row). And the differentiated field is checked by the total number of identified fields and the character type corresponding to each field to ensure whether the differentiated field meets the total number of fields in the target chapter or corresponds to The character type of each field in the target section.

For example, referring to FIG. 8, the data capture module 130 will be according to a predetermined rule. Then, the table identifies that the target chapter 801 has multiple fields (ie, the second type of chapter). Then, the data capture module 130 performs a field operation on the text content of other lines (eg, the second line 820, the third line 821, and the fourth line 822) other than the first line of the target chapter. The data capture module 130 can recognize that the second row 820 is a plurality of field names, and the third row 821 and the fourth row 822 are data corresponding to the plurality of fields.

For example, via the separator character that is a space, the data capture module 130 distinguishes six fields for the text content of the second line 820, "card issuer card name credit card issue date deactivation date usage status." The data capture module 130 compares the total number of fields in the target chapter with the number of divided fields. In this example, since the total number of fields of the target chapter (ie, 6) coincides with the number of differentiated fields (ie, 6), the data capture module 130 determines the text content for the second row 820. The operation to distinguish the fields is successful. In addition, the data capture module 130 determines that the field names of the six fields of the target chapter are "issuing institution", "card name", "amount", "issuing date", "deactivation date", and " The usage status", and, as described above, the data retrieval module 130 can obtain the corresponding "issuing institution", "card name", "credit", "issuing date", "deactivation date", respectively, via the predetermined rule table. The character type with the "Use Status" field.

Then, the data capture module 130 distinguishes 7 texts of the third line 820 by the text content "Yuanda Bank VISA Puka 64 105/02/15 112/02/15 in use". In the field, the text contents of the seven fields are “Yuanda Bank”, “VISA”, “Puka”, “64”, “105/02/15”, “112/02/15”, and "Using". In this example, due to the total number of fields in the target chapter The number (ie, 6) is inconsistent with the number of differentiated fields (ie, 7), and the data capture module 130 determines that the operation of the distinguishing field for the text content of the second line 820 fails, and will go according to each The character type corresponding to a field is used to determine how to divide 7 fields into 6 fields.

For example, the third field in the second row 820 is "amount" and the corresponding character type is a numeric character. However, the third field distinguished in the third row 821 is the Chinese character "Puka", which does not match the character type of the field to which it belongs (the third field, "the amount") ( The numeric character), but conforms to the character type of the second column of the second row 820 (ie, the Chinese character). Also, the fourth field distinguished in the third row 821 is a numeric character "64" that conforms to the character type of the numeric character. Based on the above comparison, the data capture module 130 will attempt to merge the data "Puka" in the third column of the third row 821 into the data "VISA" in the second column, and continue to compare the subsequent rules through the predetermined rule table. Whether the character type of the data recorded in the field conforms to the corresponding character type.

In this example, since the "Puka" to "VISA" is merged to become the same field (ie, the data in the second column is "VISA Puka"), the third line 821 has the fields " Yuanda Bank, "VISA Puka", "64", "105/02/15", "112/02/15", and "in use" have 6 fields, which are consistent with the total number of fields. And the character type of the data recorded in each field of the third row 821 also conforms to the character type corresponding to the field to which it belongs (for example, the data "64" in the third column matches the field to which it belongs. The character type of the "number" - the numeric character). Therefore, the data capture module 130 determines that the operation of the distinguishing field for the third row 821 is successful. Then, the data capture module 130 continues to distinguish the fields from the text content of the fourth line 822. The method is as described above and will not be described again.

After the fields of all the target chapters are distinguished, in step S20739, the data capturing module 130 extracts the plurality of pieces of data recorded by the plurality of fields from the plurality of divided fields. For example, the data capture module 130 reads a plurality of fields corresponding to the second row 820, "issuing institution", "card name", "amount", "issuing date", "deactivation date", and The "Usage Status" of the multiple records in the third line 821 "Yuanda Bank", "VISA Puka", "64", "105/02/15", "112/02/15", and " Using". In addition, the data capture module 130 may record the captured multiple data to the corresponding user in the user profile database 140 according to the plurality of field names to which the plurality of data belong respectively. In the file, the credit rating module 150 can calculate the credit rating of the user based on the user profile in which the plurality of data has been stored.

It should be noted that the above method according to the present invention may be implemented in a hardware, a firmware, or may be implemented as a software that can be stored in a recording medium such as a CD ROM, a RAM, a floppy disk, a hard disk, or a magneto-optical disk or Computer code or software or computer code downloaded over a network and stored on a non-transitory machine readable medium, such that the methods described herein can be implemented in such software using a general purpose computer or in a dedicated processor or programmable or dedicated In hardware (such as ASIC or FPGA). It should be understood by those of ordinary skill in the art that a computer, processor, microprocessor controller or programmable hardware includes storage elements (eg, RAM, ROM, flash memory, etc.) that can store or receive software or computer code. The processor or hardware implements the processing methods described herein when the software or computer code is accessed and executed by a computer. In addition, it should be noted that When a general purpose computer accesses code for implementing the processing described herein, execution of the code transforms the general purpose computer into a dedicated computer for performing the processing described herein.

In summary, the data extraction method provided by an embodiment of the present invention can convert a portable file format file corresponding to a user's credit record into a vector file, and then perform a cleaning operation on the vector file to Obtain a text file containing only specific content, and efficiently retrieve the data recorded in multiple fields from the target chapter.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and any one of ordinary skill in the art can make some changes and refinements without departing from the spirit and scope of the present invention. The scope of the invention is defined by the scope of the appended claims.

Claims (9)

A data capture method is applicable to a portable file format (PDF) file corresponding to a credit record of a user, the method comprising: performing a vector conversion on the portable file format file to enable the The portable file format file is converted into a vector file, wherein the portable file format file has multiple lines of text content respectively configured in multiple layers, wherein the multi-line text content of the vector file is the portable file format file Configuring the plurality of lines of text content of the layers, and the plurality of lines of text content of the vector file may be circled; according to the layers of the portable file format file, the vector file is The multi-line text content is divided into a first type of text content and a second type of text content; a cleaning operation is performed on the vector file to convert the vector file into a text file; and the first type from the text file belongs to the first type a plurality of fields of a target chapter of the plurality of chapters of the text content capture a plurality of data recorded by the fields, wherein the cleaning operation is performed on the vector file The step of converting the vector file into the text file comprises: storing part or all of the plurality of lines of text content of the vector file to a same layer of the text file; from the text file belonging to the first type Identifying a plurality of chapters in a plurality of lines of text content; dividing the plurality of chapters into a first type chapter having no plurality of fields And a second type of chapter having a plurality of fields; and correspondingly marking the first type of chapters and the second type of chapters according to the names of the plurality of sections of the first type of chapters and the second type of chapters A multi-line text content of a portion of the chapter, wherein the first type of chapter is separated from the multi-line text content of the other parts of the second type of chapters that are not marked. The data extraction method of claim 1, wherein the vector file is a scalable vector graphics (SVG) file. The data extraction method of claim 1, wherein the layers of the portable file format file comprise: a main layer; a watermark layer, wherein the text content of the watermark layer and the main layer The text content overlaps each other; a header layer; and a tail layer. The data extraction method of claim 3, wherein the plurality of lines of text content of the vector file are divided into the first type of text content according to the layers of the portable file format file. The step of the second type of text content includes: dividing the text content corresponding to the watermark layer in the multi-line text content of the vector file, the text content corresponding to the header layer, and the text content corresponding to the footer layer into The second type of text content; and the plurality of lines of text content of the vector file are not the second type of text content The other text content is divided into the first type of text content. The method for extracting data according to claim 1, wherein the fields are extracted from the fields of the target chapters of the text files belonging to the first type of text content. The step of recording the plurality of pieces of data includes: selecting the marked plurality of lines of text content of the target chapter from the text file; and if the target chapter belongs to the first type of chapter, identifying from the selected plurality of lines of text content The name of the target chapter identifies the text content of the name of the target chapter as a chapter name field, and identifies the other text content of the selected multi-line text content as one corresponding to the chapter name field. a chapter content material, and extracting the chapter content material; and if the target chapter belongs to the second type chapter, respectively selecting the plurality of lines of text according to the selected plurality of lines of text content of the target chapter Each line of text content of the content is divided into a plurality of fields, to extract the plurality of pieces of data recorded by the fields from the fields, wherein the plurality of lines belonging to the same field are The text content recorded in one line is the field name of the same field, and the text content recorded in the other lines in the same field is the data corresponding to the field name of the same field. The data extraction method of claim 5, wherein the selected plurality of lines of text content according to the selected target chapter respectively distinguish each line of text content of the selected plurality of lines of text content into The fields, wherein the steps of extracting the plurality of pieces of data recorded by the fields from the fields include: Identifying the name of the target chapter according to the first line of text content of the selected plurality of lines of text content; querying the total number of fields of the target chapter from a predetermined rule table according to the name of the target chapter, or according to Selecting a second line of text content of the plurality of lines of text content to identify a total number of the fields, wherein the selected second line of text content of the plurality of lines of text content is used to represent a plurality of fields of the target chapter a name, wherein each of the field names has a separator character; each of the field names identifies a character type corresponding to each of the fields; and the selected character is selected according to the separator character Each line of text content of the plurality of lines of text content to the last line of text content is respectively divided into a plurality of fields, such that the number of the plurality of fields of the line of text is distinguished. Corresponding to the total number of the fields of the target chapter, and making the data of the divided plurality of fields of the line of text content conform to the character type of the field to which the data belongs; distinguish These are the fields that capture the multi-pen information recorded in these fields. The method of extracting data as described in claim 6 wherein the separator character comprises a space. The method for extracting data according to claim 1, wherein the portable file format file related to the credit record of the user includes a credit report of the user. A personal credit analysis system includes: a vector conversion module that receives a portable file format (PDF) file of a credit record of a user, wherein the vector conversion module performs a portable file format file Converting the portable file format file into a vector file, wherein the portable file format file has a plurality of lines of text content respectively configured in a plurality of layers, wherein the plurality of lines of text content of the vector file are The multi-line text content of the portable file format file is configured in the layer, and the multi-line text content of the vector file can be circled; a cleaning module according to the portable file format file The plurality of lines of text of the vector file are divided into a first type of text content and a second type of text content, and a cleaning operation is performed on the vector file to convert the vector file into a text. a file capture module that retrieves the columns from a plurality of fields of a target chapter of the plurality of chapters of the first type of text content of the text file a plurality of records recorded; a user profile database; and a credit rating module, wherein the data capture module further captures the field names of the fields to which the plurality of data belong respectively The plurality of pieces of data are stored in a user file corresponding to the user in the user profile database, wherein the credit rating module calculates a credit rating of the user according to the user profile. In the operation of performing the cleaning operation on the vector file to convert the vector file into the text file, the cleaning module stores part or all of the multi-line text content of the vector file to the text. The same layer of the file; the cleaning module identifies a plurality of chapters from the plurality of lines of text content belonging to the first type of text content of the text file; the cleaning module divides the plurality of chapters into having no plurality of fields a first type chapter and a second type chapter having a plurality of fields; and the cleaning module correspondingly marks the first type according to the names of the first type chapter and the plurality of chapters of the second type chapter a plurality of lines of text content of the chapters and portions of the chapters of the second type of chapters, wherein the first type of chapters and the plurality of lines of text of the other parts of the second type of chapters that are not marked are Separate the horizontal lines.