Data security identification method and device covering physics, semantics and sensitive points

Abstract

The invention relates to the technical field of data identification, in particular to a data security identification method covering physics, semantics and sensitive points, and the method comprises the following steps: processing an obtained first industry review document set and a second industry review document set to generate an industry semantic feature lexicon; performing semantic analysis on each sensitive point information sentence to generate a sensitive point information phrase sequence, taking the sensitive point information phrase sequence as input, performing training through a professional field document to generate a word vector model, and performing similarity calculation on the word vector model to generate a sensitive feature lexicon; based on the industry semantic feature lexicon and a word segmentation tool, keyword segmentation is carried out on the target document, and a semantic security identifier is generated; and performing word segmentation on a target sentence in the target document based on the inactive word bank and the industry semantic feature word bank, and generating a sensitive security identifier in combination with the sensitive feature word bank. By applying the method, effective identification can still be realized under the condition of data adjustment, and semantic representation of the body is enhanced.

Description

Technical Field

[0001] The present application relates to the field of data identification technology, and in particular to a data security identification method and device covering physical, semantic, and sensitive points. Background Art

[0002] Data security labels are mainly divided into two types: separate labels and binding labels. Binding labels mainly refer to digital watermarks. The origin of digital watermarks can be traced back to the early 1990s, when the technology was mainly used to protect the copyright of audio files. In 1992, the popularity of the Internet led to the gradual increase in the problem of music piracy. At that time, the American Music Industry Association began to look for solutions, and digital watermarks were one of them. Subsequently, researchers applied digital watermark technology to images, videos, texts and other fields, and began to explore the wider application of digital watermarks, such as digital copyright management, anti-counterfeiting and traceability. In August 2002, Rakesh Agrawal and Jerry Kiernan of IBM Almaden Research Center published the paper "Watermarking of Relational Databases" at the 28th VLDB International Academic Conference held in Hong Kong, China. They first proposed and demonstrated database watermark technology and designed a database watermark algorithm based on statistical analysis. Subsequently, database watermark technology has always been a hot topic of research.

[0003] With the development of big data technology, the difficulty of embedding traditional watermarks has increased, while the difficulty of cracking and removing watermarks has decreased. The industry hopes to achieve data fingerprint extraction through separation tags, thereby achieving data ownership identification and responsibility traceability. However, there are few related research results. The mainstream method is to hash the data, generate a data check code, and then encapsulate it through a signature algorithm to achieve data security identification. However, the correlation between the identifier and the data is not great, and if there is a slight local modification of the data, it will cause the identifier to be unrecognizable.

[0004] In view of this, how to provide a data security identification method covering physical, semantic, and sensitive points, which can quickly and accurately compare the identification distance between data, has become a technical problem that urgently needs to be solved. Summary of the invention

[0005] The embodiments of the present application provide a data security identification method covering physics, semantics, and sensitive points, a data security identification device covering physics, semantics, and sensitive points, an electronic device, and a computer storage medium, which are used to solve the problem that mainstream data security identification has little correlation with the data itself and cannot track the semantics of the data based on the data content, and the mainstream data security identification has minor local modifications to the data, resulting in the inability to be identified.

[0006] In a first aspect of an embodiment of the present application, a data security identification method covering physical, semantic, and sensitive points is provided, including:

[0007] Generate an industry semantic feature word library by processing the first industry review document set and the second industry review document set, wherein the first industry review document set is a document set of a specific industry field; the second industry review document set is a document set of other industry fields except the specific industry;

[0008] Perform semantic analysis on each sensitive information sentence to generate a sensitive information phrase sequence, use the sensitive information phrase sequence as input, train through professional field documents to generate a word vector model, and generate a sensitive feature word library by performing similarity calculation on the word vector model;

[0009] Based on the industry semantic feature word library and word segmentation tool, perform keyword segmentation on the target document to generate a semantic security identifier;

[0010] Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines with the sensitive feature word library to generate a sensitive security identifier.

[0011] In a second aspect of the embodiment of the present application, a data security identification device covering physical, semantic, and sensitive points is provided, including:

[0012] A first generating module is configured to generate an industry semantic feature word library by processing the acquired first industry review document set and second industry review document set, wherein the first industry review document set is a document set of a specific industry field; the second industry review document set is a document set of other industry fields except the specific industry;

[0013] The second generation module is configured to perform semantic analysis on each sensitive point information sentence to generate a sensitive point information phrase sequence, take the sensitive point information phrase sequence as input, train it through professional field documents, generate a word vector model, and generate a sensitive feature word library by performing similarity calculation on the word vector model;

[0014] A third generation module is configured to perform keyword segmentation on the target document based on the industry semantic feature word library and the word segmentation tool to generate a semantic security identifier;

[0015] The fourth generation module is configured to segment the target sentence in the target document based on the stop word library and the industry semantic feature word library, and generate a sensitive security identifier in combination with the sensitive feature word library.

[0016] In a third aspect of an embodiment of the present application, a computing device is provided, including:

[0017] Memory and processor;

[0018] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the above-mentioned data security identification method covering physical, semantic, and sensitive points are implemented.

[0019] According to the fourth aspect of the embodiment of the present application, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of the above-mentioned data security identification method covering physical, semantic, and sensitive points are implemented.

[0020] The present application provides a data security identification method covering physical, semantic and sensitive points, including: first, by processing the acquired first industry review document set and second industry review document set, an industry semantic feature word library is generated, wherein the first industry review document set is a specific industry field document set; the second industry review document set is a document set of other industry fields except the specific industry; then, semantic analysis is performed on each sensitive point information sentence to generate a sensitive point information phrase sequence, the sensitive point information phrase sequence is used as input, and a word vector model is generated through training through professional field documents, and a sensitive feature word library is generated by performing similarity calculation on the word vector model; secondly, based on the industry semantic feature word library and the word segmentation tool, keyword segmentation is performed on the target document to generate a semantic security identification; finally, based on the stop word library, the industry semantic feature word library is used to segment the target sentence in the target document, and combined with the sensitive feature word library, a sensitive security identification is generated.

[0021] The application of the data security identification method covering physical, semantic and sensitive points provided in the embodiment of the present application has the following beneficial effects: on the one hand, it avoids the problem of being unable to be identified due to minor local adjustments to the data in the mainstream data security identification, and can still achieve effective identification when the data is adjusted; on the other hand, it strengthens the weakness that the mainstream data security identification has little correlation with the data itself and cannot achieve the semantic representation of the data itself.

[0022] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0024] Figure 1 A flowchart of a data security identification method covering physical, semantic and sensitive points provided in an embodiment of the present application;

[0025] Figure 2 A schematic diagram of the construction process of an industry semantic feature word library in a data security identification method covering physical, semantic, and sensitive points provided in an embodiment of the present application;

[0026] Figure 3 A schematic diagram of the construction process of a sensitive feature word library in a data security identification method covering physical, semantic and sensitive points provided in an embodiment of the present application;

[0027] Figure 4 A schematic diagram of the structure of a data security identification device covering physical, semantic and sensitive points provided in an embodiment of the present application;

[0028] Figure 5 A structural block diagram of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0030] See also Figure 1 , Figure 1 A flow chart of a data security identification method covering physical, semantic and sensitive points provided in an embodiment of the present application. Figure 1 As shown, the specific steps include the following steps.

[0031] Step S102: Generate an industry semantic feature word library by processing the first industry review document set and the second industry review document set, wherein the first industry review document set is a document set in a specific industry field; the second industry review document set is a document set in other industry fields except the specific industry;

[0032] Step S104: semantic analysis is performed on each sensitive information sentence to generate a sensitive information phrase sequence, the sensitive information phrase sequence is used as input, training is performed through professional field documents to generate a word vector model, and similarity calculation is performed on the word vector model to generate a sensitive feature word library;

[0033] Step S106: Based on the industry semantic feature word library and word segmentation tool, perform keyword segmentation on the target document to generate a semantic security identifier;

[0034] Step S108: Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines the sensitive feature word library to generate a sensitive security label.

[0035] First, the industry semantic feature vocabulary is modeled, see Figure 2 , Figure 2 A schematic diagram of the construction process of an industry semantic feature vocabulary in a data security identification method covering physical, semantic, and sensitive points provided in an embodiment of the present application.

[0036] like Figure 2 As shown, the industry semantic feature word library is generated by processing the acquired first industry review document set and the second industry review document set, including:

[0037] Based on the first industry review document set and the second industry review document set, a stop word library is expanded to generate a stop word library;

[0038] Based on the first industry review document and the stop word library, perform document segmentation to generate a first segmentation set, perform keyword screening based on the segmentation set to generate segmentation combinations, and sort the segmentation combinations according to word frequency to generate a second segmentation set;

[0039] Based on the first industry summary document and the second word segmentation set, performing word importance statistics to generate statistical results;

[0040] Based on the statistical results, an industry semantic feature vocabulary is generated.

[0041] Specifically, the stop word library is expanded based on the first industry review document set and the second industry review document set to generate a stop word library, including:

[0042] Screening the first industry review documents to generate a first industry review document set, screening the second industry review documents to generate a second industry review document set;

[0043] Perform word frequency statistics on the word segmentation of the first industry review document set and the second industry review document respectively, calculate the intersection of the word segmentation results with a word frequency greater than a corresponding threshold, and extract high-frequency words within a preset range;

[0044] Merging the high-frequency words and stop words in the open source analysis library to generate a stop word set;

[0045] Based on the stop word set, a stop word library is expanded to generate a stop word library.

[0046] Specifically, the word importance statistics are performed based on the first industry review document and the second word segmentation set to generate statistical results, including:

[0047] Generate a word segmentation thesaurus based on the second word segmentation set, segment the first industry review document based on the word segmentation thesaurus to obtain a word segmentation result, and merge the word segmentation results to generate a third word segmentation set;

[0048] Intersecting the second word segmentation set and the third word segmentation set to generate a fourth word segmentation set;

[0049] Based on the fourth word segmentation set, word importance statistics are performed to generate statistical results.

[0050] More specifically, the word importance statistics are performed based on the fourth word segmentation set to generate statistical results, including:

[0051] Calculate the number of documents in which the target segmentation word appears in the first industry summary document;

[0052] Calculating the word frequency of the target word in the first industry review document, and calculating the word importance of the target word in the first industry review document based on the word frequency;

[0053] Based on the importance of the segmented words, the segmented words in the fourth segmented word set are sorted to generate statistical results.

[0054] Specifically, generating an industry semantic feature word library based on the statistical results includes:

[0055] The statistical results are used to segment the target document, and the industry semantic feature word library is generated by combining the word segmentation importance weights.

[0056] In practical applications, the modeling process of the industry semantic feature vocabulary includes the following steps.

[0057] (1) Corpus preparation: Filter a document set D in a certain industry field = {d 01 ,d 02 ,...,d 0i ,...,d0n} and other industry document collections D1 = {d 11 ,d 12 ,...,d 1i ,...,d 1n}.

[0058] (2) Cross-screening: The document sets D and D1 in different fields are segmented uniformly, as shown in formula ①. The word frequency is counted, and the intersection of the segmentation results with word frequencies greater than the corresponding threshold is calculated. High-frequency words within a certain range are extracted, and the stop words in the open source segmentation library are merged to form a common descriptive stop word set stop_words in the industry field.

[0059] Formula ①: Word frequency statistics Among them, num_t is the number of word t that appears in the document, and num_d is the number of word segments in the entire document.

[0060] (3) Text segmentation: Use stop_words to segment the D text set separately, and find the union of the segmentation sets of all documents to form a segmentation set SD = {sd1, sd2, ..., sd i ,...,sd n}.

[0061] (4) Keyword screening: Use sliding window ② to traverse the SD word set to obtain all possible word combinations in the document, and then sort the word combination set according to word frequency according to formula ① to form a set KD1 = {kd 11 ,kd 12 ,...,kd 1i ,...,kd 1n};

[0062] Formula ②: {[sd1,sd2,sd3],sd4,...,sd i ,...,sd n}→{sd1,[sd2,sd3,sd4],...,sd i ,...,sd n The sliding window size can be adjusted according to the actual situation. The window size selected in the example is 3, and the corresponding word segmentation [sd1,sd2,sd3] includes two forms: [sd1,sd2,sd3] and [sd1,sd2].

[0063] (5) Word importance statistics: Use KD1 to form a word segmentation word library, segment the D text set, and merge the segmentation results to obtain a set KD2 = {kd 21 ,kd 22 ,...,kd 2i ,...,kd 2n}, find the intersection of KD2 and KD1 to generate KD3 = {kd 31 ,kd 32 ,...,kd 3i ,...,kd 3n}, use KD3 to segment the entire document and sort it according to the following conditions:

[0064] 1) Calculate DF = df(kdi, D), which refers to the number of documents in which the word kdi appears in the document set D;

[0065] 2) Calculate the word frequency TFi of kdi in set D according to formula ③;

[0066] Formula ③: tfij is the number of times word kdi appears in document D.

[0067] 3) Calculate the importance of the word segmentation as IFi = log(df(kdi,D)×TFi), sort the word segmentation KD3 according to IFi and form a set C1 for output.

[0068] It should be noted that this step reduces the impact of word frequency in the document on the results, and increases the impact of the word appearing in more documents on the importance of word segmentation.

[0069] (6) Word segmentation relationship statistics: Use C1 to segment the document, and combine the word segmentation importance weight IFi to form the document's semantic feature word library O = {(o1,w1),(o2,w2),...,(oi,wi),...(on,wn)}, where num_i is the number of times word i appears in the article, and num is the total number of words in the article.

[0070] (7) Semantic security identifier generation: Given a document, the document is segmented by keywords based on the semantic feature vocabulary O and the word segmentation tool, and weight information is attached to calculate the Simhash of the keyword sequence. The Simhash is the semantic security identifier of the document.

[0071] Secondly, build a sensitive feature word library, see Figure 3 , Figure 3 A schematic diagram of the construction process of a sensitive feature vocabulary in a data security identification method covering physical, semantic, and sensitive points provided in an embodiment of the present application.

[0072] In the embodiment of the present application, the semantic analysis of each sensitive point information sentence is performed to generate a sensitive point information phrase sequence, and the sensitive point information phrase sequence is used as input, and training is performed through professional field documents to generate a word vector model, including:

[0073] Use the statistical results generated by ranking each segmentation word based on the importance of the segmentation word to segment the sensitive information, and perform combination transformation on the segmentation results to generate a phrase set;

[0074] A word vector model is constructed using the first industry review document, the stop word library and the fourth word segmentation set.

[0075] In the embodiment of the present application, the sensitive feature word library is generated by performing similarity calculation on the word vector model, including:

[0076] Using the constructed word vector model and combining it with the importance of word segmentation, similar words are calculated for the phrase set to generate calculation results, and based on the calculation results, a sensitive feature word library is generated.

[0077] In the embodiment of the present application, the stop word library and the industry semantic feature word library are used to segment the target sentence in the target document, and combined with the sensitive feature word library to generate a sensitive security identifier, including:

[0078] Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines with the sensitive feature word library to determine whether the target word segmentation belongs to the key or value of the extension library in the sensitive feature word library;

[0079] In response to the target word belonging to a key or value of an extension library in the sensitive feature word library, the target word is saved and a sensitive security identifier is generated.

[0080] In practical applications, the process of building a sensitive feature vocabulary includes the following steps.

[0081] (1) Vocabulary transformation: Use the C1 domain vocabulary library to segment sensitive information, and combine the segmented results to form a phrase set Q = {Q1, Q2, ..., Qi, ..., Qn} and a list segmentation item L = {L1: [l 11 ,l 12 ,...,l 1z ],...,L i :[l i1 ,l i2 ,...,l iz ],...,L n :[l n1 ,l n2 ,...,l nz ]}.

[0082] It should be noted that L is a list of sensitive information, where L i For one of the sensitive list items, Q i Included in L.

[0083] (2) Word vector construction: Use the domain document set D = {d01, d02, ... d0i, ..., d0n}, the stop word set stop_words, and the domain word segmentation set KD3 = {kd31, kd32, ..., kd3i, ..., kd3n} to build a local word2vec word vector model.

[0084] (3) Construction of similar sensitive word segmentation word library: Use the constructed local word vector model combined with the importance of word segmentation IFi to calculate similar words on the word set Q and output them to form an extended library of sensitive word groups SQ = {Q1: [q 11 ,q 12 ,...,q 1z ],...,Q i :[q i1 ,q i2 ,...,q im ],...,Q n :[q n1 ,q n2 ,...,q np ]}.

[0085] (4) Generation of sensitive security labels: Given a sentence or text to be judged, the sentence is segmented based on the stop word set stop_word and the semantic feature library O = {(o1, w1), (o2, w2), ..., (oi, wi), ... (on, wn)}. Combined with the extended library SQ of sensitive phrases, each segmented content is searched in SQ to see if the content belongs to the key or value of the extended library. If it does, the key record is saved. Finally, Simhash is evaluated on the saved result to obtain the sensitive security label.

[0086] The application of the data security identification method covering physical, semantic and sensitive points provided in the embodiment of the present application has the following beneficial effects: on the one hand, it avoids the problem of being unable to be identified due to minor local adjustments to the data in the mainstream data security identification, and can still achieve effective identification when the data is adjusted; on the other hand, it strengthens the weakness that the mainstream data security identification has little correlation with the data itself and cannot achieve the semantic representation of the data itself.

[0087] Corresponding to the above method embodiment, this specification also provides a data security identification device embodiment covering physical, semantic, and sensitive points. Figure 4 This is a schematic diagram of a data security identification device that covers physical, semantic, and sensitive points provided in an embodiment of the present application. Figure 4 As shown, the device comprises:

[0088] The first generating module 402 is configured to generate an industry semantic feature word library by processing the acquired first industry review document set and second industry review document set, wherein the first industry review document set is a document set of a specific industry field; the second industry review document set is a document set of other industry fields except the specific industry;

[0089] The second generation module 404 is configured to perform semantic analysis on each sensitive point information sentence to generate a sensitive point information phrase sequence, take the sensitive point information phrase sequence as input, train it through professional field documents, generate a word vector model, and generate a sensitive feature word library by performing similarity calculation on the word vector model;

[0090] The third generating module 406 is configured to perform keyword segmentation on the target document based on the industry semantic feature word library and the word segmentation tool to generate a semantic security identifier;

[0091] The fourth generating module 408 is configured to segment the target sentence in the target document based on the stop word library and the industry semantic feature word library, and generate a sensitive security identifier in combination with the sensitive feature word library.

[0092] In an optional embodiment, the first generating module 402 is further configured to:

[0093] Based on the first industry review document set and the second industry review document set, a stop word library is expanded to generate a stop word library;

[0094] Based on the first industry review document and the stop word library, perform document segmentation to generate a first segmentation set, perform keyword screening based on the segmentation set to generate segmentation combinations, and sort the segmentation combinations according to word frequency to generate a second segmentation set;

[0095] Based on the first industry summary document and the second word segmentation set, performing word importance statistics to generate statistical results;

[0096] Based on the statistical results, an industry semantic feature vocabulary is generated.

[0097] In an optional embodiment, the first generating module 402 is further configured to:

[0098] Screening the first industry review documents to generate a first industry review document set, screening the second industry review documents to generate a second industry review document set;

[0099] Perform word frequency statistics on the word segmentation of the first industry review document set and the second industry review document respectively, calculate the intersection of the word segmentation results with a word frequency greater than a corresponding threshold, and extract high-frequency words within a preset range;

[0100] Merging the high-frequency words and stop words in the open source analysis library to generate a stop word set;

[0101] Based on the stop word set, a stop word library is expanded to generate a stop word library.

[0102] In an optional embodiment, the first generating module 402 is further configured to:

[0103] Generate a word segmentation thesaurus based on the second word segmentation set, segment the first industry review document based on the word segmentation thesaurus to obtain a word segmentation result, and merge the word segmentation results to generate a third word segmentation set;

[0104] Intersecting the second word segmentation set and the third word segmentation set to generate a fourth word segmentation set;

[0105] Based on the fourth word segmentation set, word importance statistics are performed to generate statistical results.

[0106] In an optional embodiment, the first generating module 402 is further configured to:

[0107] Calculate the number of documents in which the target segmentation word appears in the first industry summary document;

[0108] Calculating the word frequency of the target word in the first industry review document, and calculating the word importance of the target word in the first industry review document based on the word frequency;

[0109] Based on the importance of the segmented words, the segmented words in the fourth segmented word set are sorted to generate statistical results.

[0110] In an optional embodiment, the first generating module 402 is further configured to:

[0111] The statistical results are used to segment the target document, and the industry semantic feature word library is generated by combining the word segmentation importance weights.

[0112] In an optional embodiment, the second generating module 404 is further configured to:

[0113] Use the statistical results generated by ranking each word based on word importance to segment the sensitive information, and perform combination transformation on the segmented results to generate a phrase set;

[0114] A word vector model is constructed using the first industry review document, the stop word library and the fourth word segmentation set.

[0115] In an optional embodiment, the second generating module 404 is further configured to:

[0116] Using the constructed word vector model and combining it with the importance of word segmentation, similar words are calculated for the phrase set to generate calculation results, and based on the calculation results, a sensitive feature word library is generated.

[0117] In an optional embodiment, the fourth generating module 408 is further configured to:

[0118] Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines with the sensitive feature word library to determine whether the target word segmentation belongs to the key or value of the extension library in the sensitive feature word library;

[0119] In response to the target word belonging to a key or value of an extension library in the sensitive feature word library, the target word is saved and a sensitive security identifier is generated.

[0120] The application of the data security identification device covering physical, semantic and sensitive points provided by the embodiment of the present application has the following beneficial effects: on the one hand, the problem of being unable to be identified due to minor local adjustments to the data in the mainstream data security identification is avoided, and effective identification can still be achieved when the data is adjusted; on the other hand, the weakness of the mainstream data security identification that has little correlation with the data itself and cannot achieve semantic representation of the data itself is strengthened.

[0121] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the data security identification device covering physics, semantics, and sensitive points, since it is basically similar to the data security identification method embodiment covering physics, semantics, and sensitive points, the description is relatively simple, and the relevant parts can be referred to the partial description of the data security identification method embodiment covering physics, semantics, and sensitive points.

[0122] Figure 5 This is a block diagram of a computing device provided in an embodiment of the present application. The components of the computing device 500 include but are not limited to a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and the database 550 is used to store data.

[0123] The computing device 500 also includes an access device 540, which enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 540 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0124] In one embodiment of the present specification, the above components of the computing device 500 and Figure 5 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 5 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0125] The computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 500 may also be a mobile or stationary server.

[0126] Among them, the processor 520 is used to execute the following computer executable instructions, which, when executed by the processor, implement the steps of the above-mentioned data security identification method covering physical, semantic, and sensitive points.

[0127] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to the data security identification method embodiment covering physics, semantics, and sensitive points, the description is relatively simple, and the relevant parts can be referred to the partial description of the data security identification method embodiment covering physics, semantics, and sensitive points.

[0128] An embodiment of the present specification also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the above-mentioned data security identification method covering physical, semantic, and sensitive points.

[0129] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to the data security identification method embodiment covering physics, semantics, and sensitive points, the description is relatively simple, and the relevant parts can be referred to the partial description of the data security identification method embodiment covering physics, semantics, and sensitive points.

[0130] An embodiment of the present specification also provides a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned data security identification method covering physical, semantic, and sensitive points.

[0131] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer program embodiment, since it is basically similar to the data security identification method embodiment covering physics, semantics, and sensitive points, the description is relatively simple, and the relevant parts can be referred to the partial description of the data security identification method embodiment covering physics, semantics, and sensitive points.

[0132] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0133] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0134] It should be noted that the above is a description of a specific embodiment of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of the present specification.

[0135] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0136] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A data security identification method covering physical, semantic and sensitive points, characterized in that: include: Generate an industry semantic feature word library by processing the first industry review document set and the second industry review document set, wherein the first industry review document set is a document set of a specific industry field; the second industry review document set is a document set of other industry fields except the specific industry; Perform semantic analysis on each sensitive information sentence to generate a sensitive information phrase sequence, use the sensitive information phrase sequence as input, train through professional field documents to generate a word vector model, and generate a sensitive feature word library by performing similarity calculation on the word vector model; Based on the industry semantic feature word library and word segmentation tool, perform keyword segmentation on the target document to generate a semantic security identifier; Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines with the sensitive feature word library to generate a sensitive security identifier.

2. The method according to claim 1, characterized in that The process of generating an industry semantic feature word library by processing the first industry review document set and the second industry review document set obtained includes: Based on the first industry review document set and the second industry review document set, a stop word library is expanded to generate a stop word library; Based on the first industry review document and the stop word library, perform document segmentation to generate a first segmentation set, perform keyword screening based on the segmentation set to generate segmentation combinations, and sort the segmentation combinations according to word frequency to generate a second segmentation set; Based on the first industry summary document and the second word segmentation set, performing word importance statistics to generate statistical results; Based on the statistical results, an industry semantic feature vocabulary is generated.

3. The method according to claim 2, characterized in that The method of expanding the stop word library based on the first industry review document set and the second industry review document set to generate the stop word library includes: Screening the first industry review documents to generate a first industry review document set, screening the second industry review documents to generate a second industry review document set; Perform word frequency statistics on the word segmentation of the first industry review document set and the second industry review document respectively, calculate the intersection of the word segmentation results with a word frequency greater than a corresponding threshold, and extract high-frequency words within a preset range; Merging the high-frequency words and stop words in the open source analysis library to generate a stop word set; Based on the stop word set, a stop word library is expanded to generate a stop word library.

4. The method according to claim 2, characterized in that: The step of performing word importance statistics based on the first industry review document and the second word segmentation set to generate statistical results includes: Generate a word segmentation thesaurus based on the second word segmentation set, segment the first industry review document based on the word segmentation thesaurus to obtain a word segmentation result, and merge the word segmentation results to generate a third word segmentation set; Intersecting the second word segmentation set and the third word segmentation set to generate a fourth word segmentation set; Based on the fourth word segmentation set, word importance statistics are performed to generate statistical results.

5. The method according to claim 4, characterized in that The step of performing word importance statistics based on the fourth word segmentation set and generating statistical results includes: Calculate the number of documents in which the target segmentation word appears in the first industry summary document; Calculating the word frequency of the target word in the first industry review document, and calculating the word importance of the target word in the first industry review document based on the word frequency; Based on the importance of the segmented words, the segmented words in the fourth segmented word set are sorted to generate statistical results.

6. The method according to claim 2, characterized in that The generating of the industry semantic feature word library based on the statistical results includes: The statistical results are used to segment the target document, and the industry semantic feature word library is generated by combining the word segmentation importance weights.

7. The method according to claim 1, characterized in that The semantic analysis of each sensitive point information sentence is performed to generate a sensitive point information phrase sequence, and the sensitive point information phrase sequence is used as input to train through professional field documents to generate a word vector model, including: Use the statistical results generated by ranking each word based on word importance to segment the sensitive information, and perform combination transformation on the segmented results to generate a phrase set; A word vector model is constructed using the first industry review document, the stop word library and the fourth word segmentation set.

8. The method according to claim 1, characterized in that The method of generating a sensitive feature word library by performing similarity calculation on the word vector model includes: Using the constructed word vector model and combining it with the importance of word segmentation, similar words are calculated for the phrase set to generate calculation results, and based on the calculation results, a sensitive feature word library is generated.

9. The method according to claim 1, characterized in that: The stop word library and the industry semantic feature word library are used to segment the target sentence in the target document, and combined with the sensitive feature word library to generate a sensitive security identifier, including: Based on the stop word library, the industry semantic feature word library performs word segmentation on the target sentence in the target document, and combines with the sensitive feature word library to determine whether the target word segmentation belongs to the key or value of the extension library in the sensitive feature word library; In response to the target word belonging to a key or value of an extension library in the sensitive feature word library, the target word is saved and a sensitive security identifier is generated.

10. A data security identification device covering physical, semantic and sensitive points, characterized in that: include: A first generating module is configured to generate an industry semantic feature word library by processing the acquired first industry review document set and second industry review document set, wherein the first industry review document set is a document set of a specific industry field; the second industry review document set is a document set of other industry fields except the specific industry; The second generation module is configured to perform semantic analysis on each sensitive point information sentence to generate a sensitive point information phrase sequence, take the sensitive point information phrase sequence as input, train it through professional field documents, generate a word vector model, and generate a sensitive feature word library by performing similarity calculation on the word vector model; A third generation module is configured to perform keyword segmentation on the target document based on the industry semantic feature word library and the word segmentation tool to generate a semantic security identifier; The fourth generation module is configured to segment the target sentence in the target document based on the stop word library and the industry semantic feature word library, and generate a sensitive security identifier in combination with the sensitive feature word library.

Images