Sensitive information retrieval method and apparatus

Abstract

This invention provides a method and apparatus for sensitive information retrieval. The method includes: obtaining first information of a user who published a document to be retrieved; obtaining second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated; matching the starting character segments of the first information and the second information with the document to be retrieved; if the starting character segment of the first information or the second information is matched in the document to be retrieved, then reading a string of the length of the first information or the second information starting from the beginning position of the starting character segment in the document to be retrieved, based on the string length of the first information or the second information; matching the string with the first information or the second information; if the match is successful, then using the string as the retrieved sensitive information. This invention enables rapid retrieval of sensitive information.

Description

Technical Field

[0001] This invention relates to the field of information security technology, and in particular to a method and apparatus for retrieving sensitive information. Background Technology

[0002] Currently, complaint ticket information comes from multiple channels and is diverse. The content of these tickets is uniformly filled out and archived by business personnel and contains sensitive information, including but not limited to names, ID numbers, mobile phone numbers, and addresses. When these tickets reach the vendor, they are displayed in plain text, posing a high risk of information leakage.

[0003] During the process, sensitive information such as user privacy data contained in the complaint needs to be anonymized. Before anonymization, sensitive information in the complaint needs to be retrieved. Due to the massive amount of user data related to operator services, existing sensitive word filtering algorithms often ignore the relationships between sensitive words, resulting in duplicate matching and scanning searches.

[0004] Furthermore, DFA (Deterministic Finite Automaton) or Aho-Corasick algorithms require all keywords to be constructed as a trie, i.e., a state transition table, beforehand. When there are a massive number of keywords, the size of the trie will become unacceptably large and cannot be accommodated in memory.

[0005] Regular expression solutions are inherently inefficient and require multiple matching operations, making them unsuitable for scenarios with a massive number of keywords.

[0006] Full-text search algorithms are typically used for a small number of keywords, such as one or a few keywords. They mainly rely on the preprocessing and word segmentation of the original text and cannot be used for efficient retrieval of massive amounts of keywords. Summary of the Invention

[0007] This invention provides a sensitive information retrieval method and apparatus to address the shortcomings of low sensitivity information retrieval efficiency in existing technologies, thereby improving the retrieval efficiency of sensitive information.

[0008] This invention provides a method for retrieving sensitive information, comprising:

[0009] Obtain first information about the user who published the document to be retrieved, and obtain second information about the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated.

[0010] The starting character segments of the first and second information are respectively matched with the document to be searched;

[0011] If the starting character segment of the first information or the second information is matched in the document to be retrieved, then according to the string length of the first information or the second information, a string of the string length is read from the starting position of the starting character segment in the document to be retrieved.

[0012] The string is matched with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information.

[0013] According to a sensitive information retrieval method provided by the present invention, the step of matching the starting character segments of the first information and the second information with the document to be retrieved includes:

[0014] If the first or second information is a phone number, then the first three digits of the first or second information are used as the starting character segment to match the document to be retrieved;

[0015] If the first or second information is not a phone number, then the first character of the first or second information is used as the starting character segment to match the document to be retrieved.

[0016] According to a sensitive information retrieval method provided by the present invention, the step of matching the string with the first information or the second information includes:

[0017] If the first information or the second information is a first preset type, then determine whether the end character of the string is the same as the end character of the first information or the second information;

[0018] If the ending character of the string is the same as the ending character of the first information or the second information, then it is determined whether the string is completely identical to the first information or the second information;

[0019] If the string is exactly the same as the first information or the second information, it is known that the string has successfully matched the first information or the second information.

[0020] According to a sensitive information retrieval method provided by the present invention, the step of matching the string with the first information or the second information includes:

[0021] If the first information or the second information is a second preset type, then the string and the first information or the second information are converted into word vectors;

[0022] Calculate the similarity between the word vector of the string and the word vector of the first information or the second information;

[0023] If the similarity is greater than a preset threshold, it is known that the string successfully matches the first or second information.

[0024] According to a sensitive information retrieval method provided by the present invention, the step of obtaining second information of the user associated with the first information based on the first information includes:

[0025] Construct the information graph relationship of the user based on the user's information table in the CRM system;

[0026] Based on the first information, the information graph relationship is searched to obtain the second information associated with the first information.

[0027] According to a sensitive information retrieval method provided by the present invention, the step of matching the starting character segments of the first information and the second information with the document to be retrieved includes:

[0028] The document to be retrieved is copied multiple times using the main thread; wherein the number of copies is equal to the total number of the first and second information items.

[0029] Each copy is used as input to the corresponding sub-thread of the main thread, and the starting character segments of the corresponding first and second information are matched with the copy respectively.

[0030] According to a sensitive information retrieval method provided by the present invention, the method further includes, after using the string as the retrieved sensitive information:

[0031] Based on the start and end positions of the sensitive information in the document to be retrieved, one or more of the following methods are applied to the sensitive information in the document to be retrieved: data replacement, invalidation, randomization, offsetting, masking, and encoding, to obtain the desensitization result of the sensitive information.

[0032] The present invention also provides a sensitive information retrieval device, comprising:

[0033] The acquisition module is used to acquire first information of the user who published the document to be retrieved, and acquire second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated.

[0034] The first matching module is used to match the starting character segments of the first information and the second information with the document to be retrieved, respectively.

[0035] The reading module is configured to, if the starting character segment of the first information or the second information is matched in the document to be retrieved, read a string of the length of the first information or the second information from the starting position of the starting character segment in the document to be retrieved, according to the string length of the first information or the second information.

[0036] The second matching module is used to match the string with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information.

[0037] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above-described sensitive information retrieval methods.

[0038] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the above-described sensitive information retrieval methods.

[0039] The sensitive information retrieval method and apparatus provided by this invention, by considering the correlation between sensitive information, first uses the user's first information as the initial keyword for retrieval to obtain the user's main information, namely the second information associated with the first information, which greatly reduces the number of search keywords; then, it matches the starting character segments of the first and second information, and only matches the first and second information if a match is successful, and filters the first and second information that are matched, which greatly reduces the amount of computation for matching, improves the retrieval efficiency of sensitive information in the document to be retrieved, and realizes rapid retrieval of sensitive information. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is one of the flowcharts illustrating the sensitive information retrieval method provided by the present invention;

[0042] Figure 2 This is a schematic diagram of the user information graph relationship structure in the sensitive information retrieval method provided by the present invention;

[0043] Figure 3 This is the second flowchart illustrating the sensitive information retrieval method provided by the present invention;

[0044] Figure 4 This is the third flowchart of the sensitive information retrieval method provided by the present invention;

[0045] Figure 5 This is a schematic diagram of the sensitive information retrieval device provided by the present invention;

[0046] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0048] The following is combined Figure 1 The present invention describes a sensitive information retrieval method, which includes: step 101, obtaining first information of a user who published a document to be retrieved, and obtaining second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated.

[0049] The document to be searched is one containing sensitive information.

[0050] Optionally, the document to be retrieved is a complaint ticket containing user complaint content.

[0051] The first piece of information is user information that is likely to be contained in the documents to be retrieved published by the user.

[0052] Optionally, when the user who published the document to be retrieved is an individual, the first information is the user's phone number, and the second information includes the user's name, ID number, organization name, and address.

[0053] For example, if a mobile phone number is included in a personal complaint from an operator, the user's phone number will be used as the initial keyword to retrieve the user's main information, which is the second information associated with the first information.

[0054] When the user who published the document to be retrieved is an organization, the first information is the organization's name, and the second information includes the organization's address, as well as the names, phone numbers, and ID numbers of the employees within the organization.

[0055] Step 102: Match the starting character segments of the first information and the second information with the document to be retrieved, respectively;

[0056] The first and second pieces of information are user information, which is sensitive information. The search is performed on the first and second pieces of information within the document to be searched.

[0057] During the retrieval process, the starting character segments of the first and second pieces of information are first matched against the document to be retrieved. That is, starting from the beginning of the document, strings of the same length as the starting character segment are sequentially extracted. It is then determined whether the extracted string is identical to the starting character segment. If they are identical, the extracted string is considered a match; otherwise, it is not. This process retrieves all strings in the document that match the starting character segments of the first and second pieces of information.

[0058] Step 103: If the starting character segment of the first information or the second information is matched in the document to be retrieved, then according to the string length of the first information or the second information, a string of the string length is read from the starting position of the starting character segment in the document to be retrieved.

[0059] If the starting character segment of the first information and / or the second information exists in the document to be retrieved, it indicates that the first information and / or the second information is very likely to exist in the document to be retrieved.

[0060] If the document to be retrieved contains a starting character segment of the first information, then a string containing the starting character segment of the first information is extracted from the document to be retrieved, and the length of the string is the same as the length of the first information string.

[0061] If the document to be retrieved contains a starting character segment of the second information, then a string containing the starting character segment of the second information is extracted from the document to be retrieved, with a length equal to the length of the second information string.

[0062] Step 104: Match the string with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information.

[0063] The string containing the first information is matched with the first information, that is, it is determined whether the two are the same or similar. If the match is successful, it is known that the first information exists in the document to be searched.

[0064] The string containing the starting segment of the second information is matched with the second information to determine whether they are the same or similar. If the match is successful, it is known that the second information exists in the document to be searched.

[0065] The first and / or second information retrieved from the document to be searched will be considered as sensitive information.

[0066] This embodiment considers the relationships between sensitive information. It first uses the user's first information as the initial keyword for retrieval to obtain the user's main information, namely the second information associated with the first information, which greatly reduces the number of search keywords. Then, it matches the starting character segments of the first and second information. Only when a match is successful will the first and second information be matched. The matched first and second information are then filtered, which greatly reduces the amount of computation for matching and improves the retrieval efficiency of sensitive information in the document to be retrieved, thus achieving rapid retrieval of sensitive information.

[0067] Based on the above embodiments, the method of matching the starting character segments of the first information and the second information with the document to be retrieved in this embodiment includes: if the first information or the second information is a telephone number, then the first three digits of the first information or the second information are used as the starting character segment to match the document to be retrieved;

[0068] Because the first three characters of a phone number are specific and not arbitrarily combined, to improve the efficiency of phone number retrieval, the first three characters of the phone number are matched against every three adjacent characters in the document to be retrieved. If they are different, the phone number does not exist. If they are the same, the phone number may exist.

[0069] If the first or second information is not a phone number, then the first character of the first or second information is used as the starting character segment to match the document to be retrieved.

[0070] For sensitive information that is not a phone number, the first character of the sensitive information is directly matched against each character of the document to be searched. If they are different, the sensitive information does not exist. If they are the same, the sensitive information may exist.

[0071] This embodiment first matches the starting character segments of the first and second information, and only matches the first and second information if the match is successful. This greatly reduces the amount of computation for matching and improves the retrieval efficiency of sensitive information.

[0072] Based on the above embodiments, the step of matching the string with the first information or the second information in this embodiment includes: if the first information or the second information is a first preset type, then determining whether the end character of the string is the same as the end character of the first information or the second information;

[0073] The first preset type is a simple type, such as name, phone number, and ID number, which are unique user information. This embodiment performs precise matching for this type of sensitive information.

[0074] For sensitive information of the first preset type, if the starting character segment of the sensitive information is successfully matched, the ending character of the sensitive information will be matched further.

[0075] When matching the ending character of sensitive information, the starting position of the sensitive information's starting character segment is added to the length of the sensitive information string to obtain the ending character position of the matching string of equal length in the document to be searched. Then, it is determined whether the ending character at that position in the document to be searched is the same as the ending character of the sensitive information.

[0076] If the ending character of the string is the same as the ending character of the first information or the second information, then it is determined whether the string is completely identical to the first information or the second information;

[0077] If the ending character at the end position in the document to be searched is different from the ending character of the sensitive information, then the sensitive information does not exist. If they are the same, the sensitive information may exist. If they are the same, further investigation is conducted to determine whether a string of equal length in the document to be searched that matches both the starting and ending characters of the sensitive information is completely identical to the sensitive information.

[0078] If the string is exactly the same as the first information or the second information, it is known that the string has successfully matched the first information or the second information.

[0079] For example, if a user's phone number is 13500000001 and the starting character field is 135, if 135 exists in the document to be searched, then a string of 11 characters, starting with 135 and including 135, is extracted from the document. The string is then checked to see if the ending character is 1. If it is, it is further checked to see if the string is exactly the same as the user's phone number. If they are exactly the same, it means that the user's phone number exists in the document to be searched. The start and end positions of the phone number in the document are recorded for subsequent anonymization during the work order process.

[0080] This embodiment first matches the starting character segment of sensitive information of the first preset type. If the starting character segment is matched, the ending character is matched. If the ending character is matched, the entire sensitive information is matched. This reduces the amount of computation for matching and improves the retrieval efficiency. On the other hand, it achieves accurate matching of sensitive information of the first preset type.

[0081] Based on the above embodiments, the step of matching the string with the first information or the second information in this embodiment includes: if the first information or the second information is a second preset type, then convert the string and the first information or the second information into word vectors; calculate the similarity between the word vector of the string and the word vector of the first information or the second information; if the similarity is greater than a preset threshold, then it is known that the string has successfully matched the first information or the second information.

[0082] The second preset type is a complex type, such as address and company name, which lacks a unified filling rule and may be filled in incorrectly, thus constituting non-unique user information. This embodiment uses fuzzy matching for this type of sensitive information.

[0083] For sensitive information of the second preset type, if the starting character segment of the sensitive information is successfully matched, the ending character of the sensitive information is not matched. Instead, the extracted string is directly matched with the sensitive information.

[0084] During matching, semantic analysis algorithms such as word2vec are used to convert the truncated string and sensitive information into word vectors. Then, based on similarity calculation methods, such as cosine similarity, the similarity between the word vectors of the two is calculated. If the similarity is greater than a preset threshold, such as 0.8, it is determined that the truncated string and the sensitive information have matched successfully.

[0085] This embodiment first performs matching on the starting character segment of sensitive information of the first preset type, and then performs matching on the entire sensitive information if the starting character segment is matched. On the one hand, it reduces the amount of matching calculation and improves the retrieval efficiency, and on the other hand, it enables fuzzy matching on sensitive information of the second preset type.

[0086] Based on the above embodiments, the step of obtaining the second information of the user associated with the first information in this embodiment includes: constructing the information graph relationship of the user based on the information table of the user in the CRM (Customer Relationship Management) system; searching the information graph relationship according to the first information to obtain the second information associated with the first information.

[0087] When a user opens an account through the operator's CRM system, the user's basic information, such as name, ID number, phone number, home address, and company name, will be stored in the CRM user information table.

[0088] This basic information is considered sensitive during the complaint ticket process. When it reaches the maintenance vendor, the relevant content in the ticket needs to be masked. This embodiment constructs an information graph relationship based on the CRM user information table and stores it in a graph database. This facilitates subsequent retrieval and identification of associated sensitive information.

[0089] When the first information of the user who published the document to be retrieved is obtained, such as the phone number used by the user when making a complaint, the first information is searched in the constructed information graph relationship based on the first information. If the first information exists in the information graph relationship, the second information associated with the first information is obtained in the information graph relationship.

[0090] Optionally, when constructing the information graph relationships, in addition to constructing the relationships between entities such as user name, ID number, phone number, and address, it is also necessary to construct the starting character range and string length information of each entity object description to facilitate subsequent rapid matching and retrieval of sensitive information, such as... Figure 2 As shown.

[0091] This embodiment constructs an information graph relationship based on the CRM user information table, and applies graph technology to the retrieval process of multiple sensitive information with related relationships. It obtains keywords related to the document to be retrieved from a massive amount of user sensitive information keywords, and realizes the rapid retrieval of the document to be retrieved, which greatly improves the efficiency of sensitive word retrieval.

[0092] Based on the above embodiments, the method of matching the starting character segments of the first information and the second information with the document to be retrieved in this embodiment includes: copying the document to be retrieved into multiple copies based on the main thread; wherein the number of copies is equal to the total number of the first information and the second information; using each copy as the input of a corresponding sub-thread of the main thread, and matching the starting character segments of the corresponding first information and the second information with the copy respectively.

[0093] like Figure 3 As shown, for example, when the first piece of information is a mobile phone number, the steps for retrieving sensitive user information associated with the keyword "mobile phone number" are as follows:

[0094] 1. Use the graph database query syntax to quickly retrieve related sensitive information in the graph database by mobile phone number, such as user name, ID number, address, and subgraph data such as the start character, string length, and end character of the sensitive information.

[0095] 2. Employ multi-threading technology for rapid retrieval of associated sensitive information. Based on the number of sensitive information items retrieved, create multiple sub-threads. The main thread copies the document to be retrieved into multiple copies, which serve as inputs for each sub-thread, and positions the thread at the beginning of each copy.

[0096] 3. Starting from the beginning position, read a character X from the copy. Use character X as the key character to search in the subgraph associated with the keyword. If the character X does not exist, read the next character of character X again and search in the subgraph associated with the keyword again until the corresponding character is found in the subgraph. Then proceed to step 4.

[0097] 4. Take the string length L and the ending character Y corresponding to the sensitive information entity. If the sensitive entity is of the first preset type, then based on L and Y, determine whether the remaining length of the copy is greater than or equal to L, and whether the character at the current position + L is equal to Y. If they do not match, return to step 3. If they match, proceed to step 5. If the sensitive entity is of the second preset type, proceed directly to step 5.

[0098] 5. Take a substring S of length L from the current position of the copy and perform a complete match with the sensitive information entity; if no match is found in the keyword association subgraph, return to step 3; if a match is found in the keyword association subgraph, proceed to step 6.

[0099] 6. Record the start and end positions of the matched sensitive information in the document to be retrieved, and skip the length L directly, continue to step 3, until the scanning of the document to be retrieved is completed.

[0100] Based on the above embodiments, this embodiment further includes, after using the string as the retrieved sensitive information, performing one or more of the following on the sensitive information in the document to be retrieved: data replacement, invalidation, randomization, offsetting, masking, and encoding, according to the start and end positions of the sensitive information in the document to be retrieved, to obtain the desensitization result of the sensitive information.

[0101] Specifically, in this embodiment, after identifying sensitive information in the document to be retrieved, the start and end positions of the identified sensitive information in the document are recorded. The sensitive information in the document to be retrieved is then de-identified based on its start and end positions.

[0102] Data anonymization refers to the transformation of certain sensitive information through anonymization rules to reliably protect sensitive and private data, prevent the leakage of sensitive information, and improve information security. The sensitive information anonymization methods in this embodiment include, but are not limited to, data replacement, invalidation, randomization, offsetting, masking, and flexible encoding.

[0103] Figure 4This is a schematic diagram illustrating the complete process of sensitive information retrieval provided in this embodiment. First, a user information graph relationship is constructed based on CRM user information; then, initial keywords are obtained, and the user information graph relationship is retrieved based on the initial keywords to obtain associated sensitive information. The keywords and associated sensitive information are then searched in the document to be retrieved to obtain the sensitive information in the document to be retrieved; finally, the sensitive information in the document to be retrieved is anonymized.

[0104] The sensitive information retrieval device provided by the present invention is described below. The sensitive information retrieval device described below can be referred to in correspondence with the sensitive information retrieval method described above.

[0105] like Figure 5 As shown, the sensitive information retrieval device provided in this embodiment includes an acquisition module 501, a first matching module 502, a reading module 503, and a second matching module 504, wherein:

[0106] The acquisition module 501 is used to acquire first information of the user who published the document to be retrieved, and acquire second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated.

[0107] The document to be searched is one containing sensitive information.

[0108] Optionally, the document to be retrieved is a complaint ticket containing user complaint content.

[0109] The first piece of information is user information that is likely to be contained in the documents to be retrieved published by the user.

[0110] Optionally, when the user who published the document to be retrieved is an individual, the first information is the user's phone number, and the second information includes the user's name, ID number, organization name, and address.

[0111] When the user who published the document to be retrieved is an organization, the first information is the organization's name, and the second information includes the organization's address, as well as the names, phone numbers, and ID numbers of the employees within the organization.

[0112] The first matching module 502 is used to match the starting character segments of the first information and the second information with the document to be retrieved, respectively.

[0113] The first and second pieces of information are user information, which is sensitive information. The search is performed on the first and second pieces of information within the document to be searched.

[0114] During the retrieval process, the starting character segments of the first and second pieces of information are first matched against the document to be retrieved. That is, starting from the beginning of the document, strings of the same length as the starting character segment are sequentially extracted. It is then determined whether the extracted string is identical to the starting character segment. If they are identical, the extracted string is considered a match; otherwise, it is not. This process retrieves all strings in the document that match the starting character segments of the first and second pieces of information.

[0115] The reading module 503 is used to read a string of the length of the first information or the second information from the beginning position of the starting character segment in the document to be retrieved if the starting character segment of the first information or the second information is matched from the document to be retrieved.

[0116] If the starting character segment of the first information and / or the second information exists in the document to be retrieved, it indicates that the first information and / or the second information is very likely to exist in the document to be retrieved.

[0117] If the document to be retrieved contains a starting character segment of the first information, then a string containing the starting character segment of the first information is extracted from the document to be retrieved, and the length of the string is the same as the length of the first information string.

[0118] If the document to be retrieved contains a starting character segment of the second information, then a string containing the starting character segment of the second information is extracted from the document to be retrieved, with a length equal to the length of the second information string.

[0119] The second matching module 504 is used to match the string with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information.

[0120] The string containing the first information is matched with the first information, that is, it is determined whether the two are the same or similar. If the match is successful, it is known that the first information exists in the document to be searched.

[0121] The string containing the starting segment of the second information is matched with the second information to determine whether they are the same or similar. If the match is successful, it is known that the second information exists in the document to be searched.

[0122] The first and / or second information retrieved from the document to be searched will be considered as sensitive information.

[0123] This embodiment considers the relationships between sensitive information. It first uses the user's first information as the initial keyword for retrieval to obtain the user's main information, namely the second information associated with the first information, which greatly reduces the number of search keywords. Then, it matches the starting character segments of the first and second information. Only when a match is successful will the first and second information be matched. The matched first and second information are then filtered, which greatly reduces the amount of computation for matching and improves the retrieval efficiency of sensitive information in the document to be retrieved, thus achieving rapid retrieval of sensitive information.

[0124] Based on the above embodiments, in this embodiment, the first matching module is used to: if the first information or the second information is a phone number, then use the first three digits of the first information or the second information as the starting character segment to match the document to be retrieved; if the first information or the second information is not a phone number, then use the first character of the first information or the second information as the starting character segment to match the document to be retrieved.

[0125] Based on the above embodiments, in this embodiment, the second matching module is used to: if the first information or the second information is a first preset type, determine whether the end character of the string is the same as the end character of the first information or the second information; if the end character of the string is the same as the end character of the first information or the second information, determine whether the string is completely identical to the first information or the second information; if the string is completely identical to the first information or the second information, it is known that the string has successfully matched the first information or the second information.

[0126] Based on the above embodiments, in this embodiment, the second matching module is used to: convert the string and the first information or the second information into word vectors if the first information or the second information is a second preset type; calculate the similarity between the word vector of the string and the word vector of the first information or the second information; if the similarity is greater than a preset threshold, it is known that the string and the first information or the second information are successfully matched.

[0127] Based on the above embodiments, the acquisition module in this embodiment is used to: construct the information graph relationship of the user based on the user's information table in the CRM system; search the information graph relationship according to the first information, and obtain the second information associated with the first information.

[0128] Based on the above embodiments, in this embodiment, the first matching module is used to: copy the document to be retrieved into multiple copies based on the main thread; wherein the number of copies is equal to the total number of the first information and the second information; and use each copy as the input of the corresponding sub-thread of the main thread, and match the starting character segments of the corresponding first information and the second information with the copy respectively.

[0129] Based on the above embodiments, this embodiment also includes a desensitization module, which is used to perform one or more of the following on the sensitive information in the document to be retrieved, namely, data replacement, invalidation, randomization, offset, masking and encoding, according to the start position and end position of the sensitive information in the document to be retrieved, so as to obtain the desensitization result of the sensitive information.

[0130] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640, wherein the processor 610, communications interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute a sensitive information retrieval method. This method includes: obtaining first information about a user who published a document to be retrieved; obtaining second information about the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated; matching the starting character segments of the first information and the second information with the document to be retrieved; if the starting character segment of the first information or the second information is matched in the document to be retrieved, then reading a string of the length of the first information or the second information starting from the beginning position of the starting character segment in the document to be retrieved, based on the string length of the first information or the second information; matching the string with the first information or the second information; if the match is successful, then using the string as the retrieved sensitive information.

[0131] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0132] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the sensitive information retrieval method provided by the above methods, the method comprising: obtaining first information of a user who published a document to be retrieved; obtaining second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated; matching the starting character segments of the first information and the second information with the document to be retrieved respectively; if the starting character segment of the first information or the second information is matched in the document to be retrieved, then reading a string of the length of the first information or the second information starting from the starting position of the starting character segment in the document to be retrieved according to the string length of the first information or the second information; matching the string with the first information or the second information; if the match is successful, then using the string as the retrieved sensitive information.

[0133] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned sensitive information retrieval methods. The method includes: obtaining first information of a user who published a document to be retrieved; obtaining second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated; matching the starting character segments of the first information and the second information with the document to be retrieved, respectively; if the starting character segment of the first information or the second information is matched in the document to be retrieved, then reading a string of the length of the first information or the second information from the beginning position of the starting character segment in the document to be retrieved, based on the string length of the first information or the second information; matching the string with the first information or the second information; and if the match is successful, using the string as retrieved sensitive information.

[0134] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0135] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for retrieving sensitive information, characterized in that, include: Obtain first information about the user who published the document to be retrieved, and obtain second information about the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated. The starting character segments of the first and second information are respectively matched with the document to be searched; If the starting character segment of the first information or the second information is matched in the document to be retrieved, then according to the string length of the first information or the second information, a string of the string length is read from the starting position of the starting character segment in the document to be retrieved. The string is matched with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information. Based on the start and end positions of the sensitive information in the document to be retrieved, one or more of the following methods are applied to the sensitive information in the document to be retrieved: data replacement, invalidation, randomization, offsetting, masking, and encoding, to obtain the desensitization result of the sensitive information.

2. The sensitive information retrieval method according to claim 1, characterized in that, The step of matching the starting character segments of the first and second information with the document to be searched includes: If the first or second information is a phone number, then the first three digits of the first or second information are used as the starting character segment to match the document to be retrieved; If the first or second information is not a phone number, then the first character of the first or second information is used as the starting character segment to match the document to be retrieved.

3. The sensitive information retrieval method according to claim 1, characterized in that, The step of matching the string with the first information or the second information includes: If the first information or the second information is a first preset type, then determine whether the end character of the string is the same as the end character of the first information or the second information; If the ending character of the string is the same as the ending character of the first information or the second information, then it is determined whether the string is completely identical to the first information or the second information; If the string is exactly the same as the first information or the second information, it is known that the string has successfully matched the first information or the second information.

4. The sensitive information retrieval method according to claim 1, characterized in that, The step of matching the string with the first information or the second information includes: If the first information or the second information is a second preset type, then the string and the first information or the second information are converted into word vectors; Calculate the similarity between the word vector of the string and the word vector of the first information or the second information; If the similarity is greater than a preset threshold, it is known that the string successfully matches the first or second information.

5. The sensitive information retrieval method according to any one of claims 1-4, characterized in that, The step of obtaining the second information of the user associated with the first information based on the first information includes: Construct the information graph relationship of the user based on the user's information table in the CRM system; Based on the first information, the information graph relationship is searched to obtain the second information associated with the first information.

6. The sensitive information retrieval method according to any one of claims 1-4, characterized in that, The step of matching the starting character segments of the first and second information with the document to be searched includes: The document to be retrieved is copied multiple times using the main thread; wherein the number of copies is equal to the total number of the first and second information items. Each copy is used as input to the corresponding sub-thread of the main thread, and the starting character segments of the corresponding first and second information are matched with the copy respectively.

7. A sensitive information retrieval device, characterized in that, include: The acquisition module is used to acquire first information of the user who published the document to be retrieved, and acquire second information of the user associated with the first information based on the first information; wherein the first information and the second information are pre-associated. The first matching module is used to match the starting character segments of the first information and the second information with the document to be retrieved, respectively. The reading module is configured to, if the starting character segment of the first information or the second information is matched in the document to be retrieved, read a string of the length of the first information or the second information from the starting position of the starting character segment in the document to be retrieved, according to the string length of the first information or the second information. The second matching module is used to match the string with the first information or the second information. If the match is successful, the string is used as the retrieved sensitive information. The desensitization module is used to perform one or more of the following on the sensitive information in the document to be retrieved, namely, data replacement, invalidation, randomization, offset, masking and encoding, based on the start and end positions of the sensitive information in the document to be retrieved, so as to obtain the desensitization result of the sensitive information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the sensitive information retrieval method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the sensitive information retrieval method as described in any one of claims 1 to 6.

Images