Protect PII using tokenization and persistence

By managing and replacing personal identity information in the client database, the problem of inefficient protection and deletion of PII in the existing technology is solved, and rapid retrieval and deletion is achieved, resource consumption and financial costs are reduced, and GDPR requirements are met.

CN112262388BActive Publication Date: 2024-11-08MICROSOFT TECHNOLOGY LICENSING LLC
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN201980039002.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2018-06-08
Filing Date
2019-05-31
Publication Date
2024-11-08
Estimated Expiration
2039-05-31

AI Technical Summary

Technical Problem

When protecting personal identity information (PII), the management and deletion process is inefficient, which can easily lead to accidental leakage, and consume a lot of resources, making it difficult to meet the requirements of government regulations such as GDPR.

Method used

Through software applications, store and manage personal identity information in client databases, use authorization levels to detect and replace, generate human-readable views, ensure that only authorized users are visible, and automatically delete PII outside the trust boundary.

Benefits of technology

It realizes rapid retrieval and deletion of personal identity information, reduces resource consumption, ensures information security, meets the requirements of government regulations, and reduces financial costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112262388B_ABST
    Figure CN112262388B_ABST
Patent Text Reader

Abstract

The system includes an electronic processor configured to store a record in a client database. The record includes personally identifiable information associated with an entity and a client identifier reference associated with each individual, and the personally identifiable information can be accessed based on an authorization level associated with the user. The electronic processor is also configured to receive a document containing both non-personally identifiable information and personally identifiable information, parse the document to determine whether personally identifiable information exists in the document, verify access rights of a reader of the document, match the personally identifiable information contained in the document with a record stored in the client database, replace the personally identifiable information contained in the document with a corresponding client identifier reference that references the client database, and generate a human-readable view of the document containing the corresponding client identifier reference.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Embodiments described in this application relate to protecting personally identifiable information associated with digital content. Background Art

[0002] Protection of personally identifiable information (PII) is required by government laws and regulations. Even in cases where protection is not required, organizations may actively protect PII for a variety of reasons, such as building trust with customers / users or minimizing legal risk. In an increasingly collaborative and electronic world, the management of personally identifiable and sensitive information has become particularly important. While the management of such information in a way that prevents unauthorized access is often focused on security, equally important efforts must be made to help prevent users from accidentally leaking such information through the simple exchange of electronic documents. Regardless of the reason, protecting PII is expensive in terms of resources and ultimately imposes a financial cost on the organization. Summary of the invention

[0003] A simplified overview of one or more embodiments of the present disclosure is provided below to provide a basic understanding of such embodiments. This overview is not a broad overview of all contemplated implementations, and is neither intended to identify the key or important elements of all embodiments, nor to describe the scope of any or all embodiments. The sole purpose of this overview is to present some concepts of one or more embodiments of the present disclosure in a simplified form, as a preface to a more detailed description presented later.

[0004] With the implementation of government regulations, such as the General Data Protection Regulation (GDPR), organizations need to be able to efficiently retrieve and (upon request) delete content and documents that contain personally identifiable information. Existing methods for handling personally identifiable information include performing an internal audit, where a customer submits a request to the organization that is associated with that customer's personally identifiable information. Typically, an individual working for the organization performs a bulk query of all customer documents, and then analyzes each document to determine if the document includes personally identifiable information. Redaction of personally identifiable information is performed manually. The entire process can take days, weeks, or months.

[0005] Among other things, the embodiments described in this application include a software application (e.g., it can be a local stand-alone application or a server-based client application) that enables users within an organization to create, save, and share documents. The term "document" is used to refer to a database record, a word document, a spreadsheet, a presentation, a portable document format (PDF), a sound file, an image file, a video, etc. The embodiments described in this application protect personally identifiable information by utilizing the organization's software service that allows the creation and storage of data by creating an account or subscription in the organization using the organization's email service or productivity suite (e.g., a document storage suite such as Sharepoint) or at the operating system level (e.g., using a cloud-based user directory with access control information, such as Azure Active Directory). In some embodiments, among other things, the systems and methods discussed in this application enable instant tagging of named individuals when a document is authored, thereby enabling faster and in some cases almost instant retrieval of all documents related to the individual. The systems and methods discussed also enable a client's deletion request to result in the original document being edited rather than deleted. This ensures that when a document is shared, only authorized recipients can see personally identifiable information associated with the client. Additionally, any documents sent outside of the trust boundary may be automatically deleted or have PII automatically removed.

[0006] Example embodiments include a computer system for protecting information associated with concepts (e.g., topics such as discussions about space or defense programs, etc.) and / or entities including individuals, groups, organizations, etc. The computer system includes an electronic processor configured to store records in one or more client databases, wherein the records include information (e.g., personally identifiable information) associated with entities (such as, but not limited to, individuals, groups, organizations, objects, concepts, etc.), and a client identifier reference associated with specific characteristics associated with each entity (e.g., name, date of birth, place of birth, email address, phone number, fax number, specific content, social network credentials, biometric information, financial account number, organization-issued identification document, government-issued identification document, etc.). The personally identifiable information may be accessed based on an authorization level associated with a user. The electronic processor is also configured to receive a document containing both non-personally identifiable information and personally identifiable information, and parse the document to determine the presence of personally identifiable information in the document. The electronic processor is also configured to match the personally identifiable information included in the document with records stored in one or more client databases, replace the personally identifiable information included in the document with a corresponding client identifier reference stored in the client database, and generate a human-readable view of the document containing the corresponding client identifier reference. In some embodiments, the system can use machine learning to reword the client identifier reference to fit the sentence structure. For example, there are specific words in German to refer to females and males. Saying "eine [client identifier reference]" will immediately identify the person as a female. The system 100 can alternatively automatically edit "eine" from the document so as not to disclose the person's gender.

[0007] Another example embodiment includes a computer system for protecting personally identifiable information (PII). The computer system includes an electronic processor configured to store an encryption key associated with a user who has access to personally identifiable information in a document. The electronic processor is further configured to receive a document containing unencrypted non-personally identifiable information and encrypted personally identifiable information. The electronic processor is further configured to parse the document to determine the presence of personally identifiable information and detect an encrypted portion in the document. The electronic processor is further configured to retrieve the personally identifiable information based on decrypting the encrypted portion using the encryption key. The electronic processor is further configured to display a human-readable view of the document.

[0008] Another example embodiment includes a method for protecting personally identifiable information of an individual. The method includes storing a record in a client database. The record includes personally identifiable information associated with the individual and a client identifier reference associated with each individual. The personally identifiable information may be accessed based on an authorization level associated with a group to which the user or a particular user belongs. The method also includes receiving a document containing both non-personally identifiable information and personally identifiable information of one or more individuals, and parsing the document to determine the presence of personally identifiable information included in the document. The method also includes matching the personally identifiable information included in the document with a record stored in the client database. The method also includes replacing the personally identifiable information included in the document with a corresponding client identifier reference stored in the client database, and generating a human-readable view of the document containing the corresponding client identifier reference.

[0009] Another example embodiment includes a non-transitory computer-readable medium containing instructions that, when executed by one or more electronic processors, cause the one or more electronic processors to perform a set of actions, including storing a record in a client database, wherein the record includes personally identifiable information associated with individuals and a client identifier reference associated with each individual, and the personally identifiable information can be accessed based on an authorization level associated with a user or a group to which a particular user belongs, receiving a document containing both non-personally identifiable information and personally identifiable information of one or more individuals, parsing the document to determine the presence of personally identifiable information included in the document, matching the personally identifiable information included in the document with a record stored in the client database, replacing the personally identifiable information included in the document with a corresponding client identifier reference stored in the client database, and generating a human-readable view of the document containing the corresponding client identifier reference.

[0010] By using the techniques disclosed in this application, one or more devices can be configured to, among other things, conserve resources in terms of power resources, memory resources, communication bandwidth resources, processing resources, and / or other resources, while providing a mechanism for controlling and deleting personally identifiable information in documents. Technical effects other than those mentioned in this application can also be achieved from implementations of the techniques disclosed in this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The present disclosure will be better understood from the following detailed description taken in conjunction with the accompanying drawings, wherein like reference numerals are used to refer to like parts throughout the description of the drawings.

[0012] Figure 1 An example of a personally identifiable information protection system is described.

[0013] Figure 2 yes Figure 1 A block diagram of an example computing device is shown in FIG.

[0014] Figure 3 According to some embodiments, Figure 2 0 is a block diagram of several software programs stored in the memory of a computing device shown in .

[0015] Figure 4 According to some embodiments Figure 1 A block diagram of the server is shown in FIG.

[0016] Figure 5 is a block diagram of a distributed computing system in which embodiments provided in this application may be practiced.

[0017] Figure 6 and Figure 7 is a flow chart illustrating a method for protecting personally identifiable information, according to some embodiments.

[0018] Figure 8 Example text to be scrubbed to remove PII is shown in accordance with some embodiments.

[0019] Fig. 9 is a flow chart illustrating a method for protecting personally identifiable information using encryption, according to some embodiments. DETAILED DESCRIPTION

[0020] Embodiments of systems and methods for protecting personally identifiable information (PII) are described herein. The personally identifiable information protection system is configured to detect and remove personally identifiable information in documents based on authorization level-based rules provided to the system.

[0021] When creating and editing electronic documents using software applications (e.g., using productivity tools such as word processing applications or email applications), various forms of data (e.g., personally identifiable information) are typically created in the form of content in the document or in the form of metadata. Controlling access to subsequent users and document recipients can protect an individual's personally identifiable information and ensure the privacy of the user. As used in this application, personally identifiable information (PII) includes any information that specifically identifies a specific entity (e.g., a user, organization, or content, etc.). This may mean any data collected about or provided by a user unless they explicitly authorize the organization to share the information. Some common examples of personally identifiable information include name, date of birth, biometric information, government-issued identification numbers (e.g., social security number (SSN), driver's license number, etc.), financial account information (e.g., bank account number, credit card number, etc.), personal telephone number, and mailing address. Other data that can be considered PII include Internet Protocol (IP) addresses, company names, and location information. PII can also be merged with free-form data, such as document names, email titles, or any other aggregated data that can be used to determine detailed information about an entity (e.g., a user, organization, or specific content). In some embodiments, the system is configured to initially retain some PII data in the document, but once a threshold number of items is exceeded, all PII data can be removed. Each item can have a different weight associated with it. For example, when only the name is used and no other information is used - the name can be retained in the document. However, when a date of birth associated with the name is provided in the document to make it easier for the reader to determine the identity of the individual, the system will delete all data related to the individual's PII.

[0022] Figure 1 An example of a personally identifiable information protection system 100 is illustrated. In the example shown, the system 100 includes client devices 120a, 120b, 120c accessing a server 104 via a network 103. The server 104 is connected to a client database 106, a shared storage 108, a directory service 110 (e.g., a database or application that provides documents and associated authorization levels for viewing personally identifiable information in the documents), a rules and context database 112, and an encryption key database 114. The directory service 110 provides a data structure to manage several categories of users 102, including managing access credentials for each user 102 to view personally identifiable information. The system 100 can operate locally or in a distributed environment, such as in a client-server or cloud service architecture. Any distributed components of the system 100 can communicate with each other or can be accessed through at least one computer network, such as, but not limited to, the Internet, an intranet, a local area network, or a wide area network.

[0023] The terms "client," "consumer," "customer," or "subscriber" may be used interchangeably herein to refer to users 102 that utilize services provided by system 100. One or more users 102 may subscribe, or otherwise register, to access one or more software applications as a "user" of the software application. In some embodiments, users may include a single user 102 or a group of multiple users (e.g., Figure 1 ). User group A may include a group of users who can access, view, or create documents whose personally identifiable information is fully protected. User group B may include a group of users who can access, view, or create documents whose certain personally identifiable information is protected but other personally identifiable information may be exposed. User group C may include a group of users who can access, view, or create documents whose personally identifiable information is fully exposed.

[0024] The client computing device 120 (sometimes referred to herein as "client device 120") may be implemented as any number of computing devices, including, but not limited to, a personal computer, a laptop computer, a desktop computer, a portable communication device (e.g., a mobile phone or tablet), or other electronic device that can send and / or receive data over the network 103. The network 103 represents many different types of networks, and includes wired and / or wireless networks that enable communications between various entities in the system 100. In some configurations, the network 103 may include the Internet, a local area network (LAN), a wide area network (WAN), a mobile telephone network (MTN), and other types of networks (possibly used in conjunction with one another) to facilitate communications between the server 104 and the client device 120. Although some configurations are described in the context of a network-based system, other types of client / server-based communications and associated application logic may be used.

[0025] Figure 2 yes Figure 1 An example block diagram of a computing device 120 is shown in . The computing device 120 includes a microphone 121, a speaker 122, a display 123, a communication interface 124, a user interface 125, an electronic processor 126, and a memory 128 coupled to a data bus 129.

[0026] The communication interface 124 provides a communication gateway to the computing device 120 with an external network (e.g., a wireless network, the Internet, etc.). The communication interface 124 may include, for example, an Ethernet card or adapter or a wireless local area network (WLAN) card or adapter (e.g., IEEE Standard 802.11a / b / g / n). The communication interface 124 may include address, control, and / or data connections to enable appropriate communications on the external network.

[0027] The user interface 125 provides an input mechanism such as a keyboard, which may be a physical keyboard or a "soft" keyboard generated on a touch screen display. The user interface 125 is configured to receive user input to access and interact with one or more software applications stored in the memory 128.

[0028] The electronic processor 126 may include at least one processor or microprocessor that interprets and executes a set of instructions stored in the memory 128. The memory 128 includes an operating system 130 and software applications 132. In the example shown, the operating system 130 includes a graphical user interface generator 131 that creates a graphical user interface on the display 123. The software application 132 is configured to implement the methods described in this application, among other things. The memory 128 includes volatile storage elements (e.g., random access memory (RAM)), non-volatile storage elements (e.g., ROM), and combinations thereof. In some embodiments, the memory 128 has a distributed architecture in which various components are placed away from each other but can be accessed by the electronic processor 126.

[0029] The data bus 129 or other component interconnections allow for communication between components of the computing device 120. The data bus 129 may have other elements omitted for simplicity, such as controllers, buffers (e.g., caches), drivers, repeaters and receivers or other similar components to enable communication. The data bus 129 may also include address, control, data connections, or a combination of the foregoing to enable appropriate communication between the foregoing components. In some embodiments, the electronic processor 126 is configured to encrypt data associated with the PII and store the encrypted data in the memory 128.

[0030] Figure 3 According to some embodiments, the storage Figure 2 100. In some embodiments, the software applications 132 include a word processing application 310, a spreadsheet application 420, and other applications 330. The software applications also include a save agent 340 and a purge agent 350. The document save agent 340 and the purge agent 350 are run using the electronic processor 126 in the computing device 120. In one embodiment, the document save agent 340 is responsible for providing the client with the ability to access and save documents using the system 100. In some embodiments, the document save agent 340 also saves a log associated with the client's success (or failure) event when accessing the document. In some embodiments, the directory service 110 is responsible for providing the client with appropriate access to the components of the system 100.

[0031] In some embodiments, the document preservation agent 340 processes data containing personally identifiable information collected from the user 102 as part of data routinely recorded for business purposes. Examples of the type of data that may be preserved include, but are not limited to, word sets, formatting, any embedded images, charts within documents, security-related events (e.g., login attempts and user authorizations), input / output events (e.g., network or file system transactions, including viewing, editing, and saving files), and the like.

[0032] The clearing agent 350 processes the saved data based on the rules stored in the rule database 112. The clearing agent 350 can be a stand-alone application or a component of another application. Although illustrated and described as running in a single computing device, the document preservation agent 340 and the clearing agent 350 can be distributed and / or run across many computing devices. In some embodiments, the clearing agent 350 and / or the clearing application 420 include different engines that perform different types of actions on the document. In various embodiments, the clearing agent 350 and / or the clearing application 420 include a parsing engine (e.g., to identify possible PII) and a processing engine (e.g., to convert the PII stored in plain text into a reference using the rules from the database 112). The parsing engine performs processing actions responsible for the following tasks, including but not limited to parsing the fields containing personal identity information in the document or forwarding the identified fields to a sub-parser, which can decompose the field into smaller components (e.g., sub-fields or name / value pairs). In some embodiments, a filtering engine that performs a filtering action is provided, which determines whether a field in a document needs to be processed (e.g., determining whether an audio or video component in a document needs to be processed).

[0033] The word processing application 310 provides a platform for the user 102 to compose, edit, format, and print documents, including documents containing personally identifiable information. In some embodiments, the word processing application 310 includes an extension or feature that allows the word processing application 310 to access the client database 106 to retrieve a client identifier reference associated with personally identifiable information in a document accessed by the user 102. In some embodiments, the system 100 is designed to identify possible PII (e.g., a common name identifying a person or organization). In some embodiments, when a network connection is not available or in an architecture where no central database 106 is available, the personally identifiable information in a document can be encrypted and saved with the document, rather than storing the encrypted personally identifiable information in a centralized client database.

[0034] Figure 4 According to some embodiments Figure 1. The server 104 includes an electronic processor 410 and a memory 415 including one or more software applications. The computing device 120 accesses the software applications in the memory 415 of the server 104 through the network 103. The software applications may include, but are not limited to, image and video processing applications 430, audio processing applications 440, text processing applications 450, cleaning applications 420, and other programs 470. In some embodiments, the image and video processing application 430 includes a software program capable of processing still or video images and determining whether there is an image in the document that contains personally identifiable information associated with the customer.

[0035] In some embodiments, the text processing application 450 includes a software program capable of processing text data and determining whether there is any text information in the document including personally identifiable information associated with a client. In some embodiments, the cleaning application 420 uses the rules and context information stored in the rules and context database 112 to locate and replace personally identifiable information in the document. In some embodiments, the cleaning application 420 itself is a unique application that can open a document file and enable a user to browse the document content by locating and replacing specific text to clean the information (e.g., PII) in the document. When the parsing process in the cleaning application 420 identifies PII, a post-processing operation is performed to prepare a message for storage in a data store (e.g., shared storage 108). In some embodiments, the post-processing operation includes an annotation operation that annotates the document with additional information required to properly interpret the data in the document when the document is added to the data store (e.g., in the case where the document has not been cleared before). In some embodiments, when a document is received by an application that does not have an annotation operation, the application uses a previous annotation (e.g., stored in a database / memory) and applies it to these documents.

[0036] In some embodiments, the client identifier includes multiple segments to indicate several pieces of information. For example, the first segment may be associated with an individual, while the second segment may be associated with an organization. The third segment may be associated with a group or company (with some kind of hierarchical structure), or the fourth segment may be associated with a client database query system (e.g., checking a passport number).

[0037] Other information that can be added to the document includes information such as the document source (e.g., a server identifier). In some embodiments, a protected record storage operation is performed that sends the cleaned document with PII to a data repository (e.g., shared storage 108) in a protected format. Once in the data repository, the cleaned document is available to various consumers of the data.

[0038] Figure 51 is a block diagram of a distributed computing system in which embodiments provided in the present application can be practiced. Content developed, interacted with, or edited in association with software application 132 can be stored and retrieved from different servers, communication channels, or other storage types. For example, directory service 110, web portal 510, mailbox service 520, or social network service 530 can be used to store various documents. Software application 132 can use any of these types of systems, etc. to implement data utilization as described in the present application. Server 104 can also provide software application 132 to the client over the network via network 103. For example, the client computing device can be implemented as computing device 120 and embodied in personal computing device 120-1, tablet computer 120-2, and / or mobile computing device (e.g., smart phone) 120-3. Any of these embodiments of the client device can obtain content from shared memory 108.

[0039] Figure 6 and Figure 7 6 is a flow chart illustrating a method for protecting personally identifiable information according to some embodiments. At block 610, the electronic processor 410 in the server 104 stores a record (e.g., personally identifiable information associated with an individual and a client identifier) ​​in the client database 106. Each record in the client database 106 contains a name or other information (e.g., identifying an entity (e.g., an individual, group, or organization, etc.)), but may also contain one or more of the client's title, first name, middle name, and last name, and nickname / abbreviation. Additional information may be stored to help systems and organizations disambiguate between individuals and other people. This may include an email address, home address, phone number, the name of the organization they own or are employed by, individuals they work with in the organization (e.g., a specific teacher, doctor), and the names of children. The scenarios provided here are examples only and may vary depending on the nature of the organization storing data about the client.

[0040] At block 620, the electronic processor 410 receives a document containing both non-personally identifiable information and personally identifiable information of one or more individuals from the shared memory 108. At block 630, the electronic processor 410 parses the document to determine the presence of personally identifiable information included in the document. In some embodiments, parsing the document includes analyzing text strings in the document and converting the text into logical syntactic components that can be used to determine the presence of personally identifiable information.

[0041] At block 640, the electronic processor 410 matches the personally identifiable information with records stored in the client database 106. In some embodiments, the electronic processor 410 compares various portions of the document (e.g., words, sentences, lines, etc.) with entries in the client database 106 to verify whether those portions include any personally identifiable information. In some embodiments, the electronic processor 410 determines whether the personally identifiable information in the document completely matches a record (at block 710). When the personally identifiable information completely matches a record, the personally identifiable information is removed from the document (at block 720) and replaced with the corresponding client identifier (at block 730). In some embodiments, the electronic processor 410 determines (at block 740) whether the personally identifiable information matches more than one record in the client database 106. When the personally identifiable information matches more than one record in the client database 106, the system 100 prompts (at block 750) the user 102 with a prioritized list of potential candidates. In this case, the user 102 selects the most appropriate match based on viewing the matching selections from the client database 106 displayed to the user 102. In some embodiments, the system 100 uses the information in parallel (e.g., date of birth) to automatically infer the user. In other embodiments, if the system 100 does not need to be strictly accurate, a best guess is made. In some embodiments, the electronic processor 410 determines whether the personally identifiable information does not match any of the records in the client database 106. In this case, the client database 106 is updated with the personally identifiable information that did not previously have a corresponding match. In some cases, new ways of personally identifying a person or entity can also be added to the client database 106. In some embodiments, when PII and / or a person or entity is identified, the author can be prompted to provide more information about the person or entity to confirm what type of PII has been discovered.

[0042] At block 640, electronic processor 410 replaces the personally identifiable information included in the document with the corresponding client identifier stored in client database 106. At block 660, electronic processor 410 generates a human-readable view of the document that includes the corresponding client identifier.

[0043] In some embodiments, when a user 102 (e.g., John Smith with an email identifier of "john.smith@contoso.com") uses an organization's email service and / or productivity tool to retrieve or delete information stored about them, various software tools associated with the email service and / or productivity tool may be used to retrieve the information stored about the user with the email identifier of "john.smith@contoso.com". john.smith@contoso.comHowever, in many cases, the information is stored in a way that is more difficult to retrieve. For example, if an employee of a government agency (e.g., healthcare, law enforcement, or education) is writing an article about a particular person, the author of the document is not the same person as the person being referenced. Additionally, in some scenarios, the document may mention multiple individuals, for example, a business such as a bank may cite the behavior and / or history of various individual customers to change internal policies, or a sales team may have a marketing document that has "customer evidence" that references various customers. In some embodiments, the document includes nicknames (e.g., Dave / David), or only uses first names, etc., which can lead to confusion because multiple individuals may have the same name. Therefore, when a customer requests instant retrieval and deletion of all documents about them, the request may require further identification of the appropriate individual. In some embodiments, characteristics associated with the document (e.g., when the document was created, edited, or deleted) can be used to determine the identity of the appropriate individual from multiple selections available in the system.

[0044] Some embodiments provided in the present application use client database 106 to enumerate a complete list of clients (e.g., users, individuals, customers, businesses, etc.) that can be referenced in documents. Each record refers to a separate client. In one embodiment, a record can facilitate online interaction and therefore contain a username (which can be an email address) and a password in some form (e.g., typed characters, phone-based authentication, fingerprints, etc.). Other implementations may be directed to situations where a client encounters an agent of an organization (e.g., a clerk at a bank counter), so the clerk can show physical proof (e.g., a government ID) to confirm his identity to the agent.

[0045] In some embodiments, when the author of the document wishes to mention a client, they can use an explicit "client selector" to pull the client's name from the database and insert it into the document. However, more frequently, the user 102 can simply type the client's name into the document in a sentence. By this implicit input of the client's name, the software will use a background process to detect whether someone's name has been entered. This can be achieved by querying each word typed by the user 102 in the client database. In the case of Latin letters, capitalization is a strong reason to enter proper nouns in the document. Other signals can be used, such as titles (e.g., "Mr." or "Ms." or "Dr."), common names or prepositions (such as "Name:"), or the author has typed something in the field, which is known to the system as often containing the client's name (e.g., "Customer" or "Patient"). Those skilled in the art know that there are many established methods in computational linguistics and computer science to detect names in documents.

[0046] In some embodiments, the electronic processor 410 is configured to provide implicit client identification by prompting the user 102 to confirm the identity with the name or other identifying information (such as an address) in the document. In some cases, this may not be the case. For this case, a prompt is provided to the author to explicitly select the person the author envisions from a set of potential matches. In some embodiments, potential matches are sorted based on relevance to the user. In other cases, for example, if it is determined that the author has a low level of access rights, all potential matches of the individual can be stored and not displayed to the user. In addition, when a document is deleted, in those cases, each potential match must be asked whether the user referred to is the actual user. In some embodiments, there may be additional data records about the user in the database between when the document is entered into the system and when the document is deleted. In this case, the system 100 is configured to determine the identity of the individual based on the information in the document. In some embodiments, the deletion time of the document can be used to determine the identity of the individual envisioned by the user. In some embodiments, when there are two potential individual matches, one of which has requested that all its data be deleted from the system, the system can determine the identity of the individual based on this information.

[0047] In some cases, for data protection reasons, an author may not have immediate access to see a specific individual. For example, an author might type "Mary Brown" when they actually meant "Mary Jane Brown". However, there may be two records in the client database ("Mary Jane Brown" and "Mary Hanna Brown"). Therefore, when an author wants to refer to the existence of "Mary Jane Brown", they may have to seek approval (e.g., from a manager or internal management body) to write about "Mary Jane Brown" before their document can be saved, shared, published (e.g., to a staging environment for testing features, to generate a newspaper article and then send to a production / release environment where others can view the information), captured (e.g., using screenshots), printed or emailed, etc.

[0048] In other cases, the author may be writing information about a customer that is not yet recorded in the client database, and therefore may need to create this record. This may also trigger other workflows, such as approval that a new customer can be created, or notification to the customer, etc.

[0049] In some embodiments, a document is marked as containing a potential name. Documents so marked may be subject to further manual or automatic review. For example, the word "Sam" may refer to a specified individual or Samantha or the Singapore Art Museum. Since many name analysis techniques have algorithms that assign a probability possibility to a given entity as a name, those skilled in the art will recognize that the list of documents for review will be given an assigned grade. Other techniques can be used to detect that a particular document template will always contain references to customers when completed, but no customers have yet been identified in the document. In addition, in a document, a full reference to an entity is often provided first, and then a subsequent reference to the entity is made using a shorter reference, such as only a title and a surname together (e.g., Ms. Brown), or a given name in an informal document (e.g., Mary). Therefore, the most common situation is that a document has multiple references to at least one person. In this case, pronouns (e.g., using "he" or "she") or abstract representations (e.g., "the patient") are often used to further refer to the individual's activities. In some cases, the editing of "his" or "her" may be skipped. For example, in an environment where all people are male (such as a military document from 75 years ago), redaction of "he" can be skipped. In some embodiments, if "he" and "she" are scattered throughout the document, it is difficult to infer which person the pronoun refers to. In this case, the cleaning application 420 can use knowledge about the grammar of the user-specified language, or knowledge inferred from the document content, to analyze the sentence and determine the entity (e.g., person, subject, group, organization, etc.).

[0050] Through the above processing, the software executed by the electronic processor 410 determines that the document contains certain specified clients at certain points in the document. Once the author is ready to save or share the document, the software executed by the electronic processor 410 allows the author to view the client references. In some embodiments, the document template contains metadata indicating that the author must perform a review for one document type (while for other document types, the review may not be mandatory).

[0051] In some embodiments, when saving a document with a client name stored as text, the electronic processor 410 uses tags to expand the client name. For example, tags can be placed around the text (e.g., "Patient<CLIENT ID=12381289> Mary Brown presents herself to me”). This will record in the document that a specific highly identifiable client has been mentioned. In some embodiments, the document is stripped of all personally identifiable information, leaving only the reference. For example, a document may contain “Patient<CLIENT ID=12381289TEXT=FIRST_NAME / >< / CLIENT ID=12381289TEXT=LAST_NAME / > Present herself to me". In this embodiment, the personally identifiable information is located in a centralized service of the organization rather than stored as plain text. The electronic processor 410 performs further processing to present the document. It can call to retrieve the fields "First Name" and "Last Name" for the customer. To this end, authentication and authorization are used. In some embodiments, a directory service (e.g., Microsoft Azure Active Directory) is used to confirm the electronic identity of the employee. After the customer's identity is confirmed, the identity and the requested list of personally identifiable information will be passed to an intermediary service (e.g., a PII retrieval service), which determines whether the employee has the right to access the requested customer's personally identifiable information.

[0052] In some cases, all employees in an organization may have access to all customers' PII. In other cases, access may be more granular, with different viewing / editing permissions assigned (or blocked) even within an organization. For example, when electronic processor 410 receives a request from Bob York@contoso.com to access<CLIENT ID=12381289TEXT=FIRST_Name / > and<CLIENT ID=12381289TEXT=LAST_NAME / > , the PII retrieval service may confirm that Bob has viewing privileges to Mary's data and therefore retrieve "Mary" and / or "Brown" through the software, depending on whether the user has access to first name only, last name only, or first and last name. At this point, the electronic processor 410 converts the tags back into text, enabling it to render "Patient Mary Brown presented herself to me." When another employee, Joe Schmidt, attempts to view the same document, he may see the same document, but with portions blacked out (redacted).

[0053] In some embodiments, when the author of the original document indicates that a portion of the text includes personally identifiable information (e.g., such as "Because the client I represent lives in a house near the general manager of FabrikamSoft and...") but its type is not part of the regular fields in the client database, then additional data storage can be used, or this new data can be written to the client database as part of a new field. A record can be created for the selected personally identifiable information (e.g., "Because the general manager of FabrikamSoft is a neighbor" and it can be given a new identifier, such as GUID=18923101218209812). Alternatively, the identifier can be a document ID and a personally identifiable information instance ID, etc. Thus, the document may end up containing "Because the client I represent lives in a house near the general manager of FabrikamSoft and..."<CLIENT ID=78910121ADHOCTEXTIP=18923101218209812> and…” Again, many different implementations and choices of where to store the data are possible.

[0054] In some embodiments, the client can use an encryption key associated with the client, and all text is encrypted and stored in the document itself, rather than in a centralized data store. In operation, when the user's client application (used to manage PII embedded in the document) is first opened, it detects that no keys are stored locally and sends a request to the encryption key database 114 to obtain local encryption keys. The server 104 sends back all the keys that the client is authorized to use based on the user's authorization level. Once the client detects that the key has expired (although only a single key needs to be retrieved in this case), it determines to use the stored metadata as part of the key, then the same process is performed by querying the encryption key database 114 to obtain the key and recording these details locally or by a similar method.

[0055] In some embodiments, the user's client application may also have an initial key embedded in the application. This allows the user to launch the application without the network requesting the key. In some embodiments, the server delivering the client application prepares the application for a specific user, in which case the user will receive a key that they are authorized to use.

[0056] While offline, the client application can encrypt and decrypt metadata in the document using the retained encryption key. The metadata can contain information about the key required for decryption directly in the document or via metadata provided by an online service, which includes a table mapping identifiers to encryption keys. In either case, when the user does not have the necessary key, the information will be displayed to the user in a redacted form.

[0057] While online, the client application can continue to use the retained keys, or can request new keys from the encryption key database 114. Performance gains can be achieved by using the retained keys, and new keys can be requested asynchronously as a background task to ensure that the keys are kept up to date. In some embodiments, when operating in online mode, it is also possible to revert to performing decryption on the server 104 (or some other endpoint).

[0058] When an encryption key is compromised (intentionally or accidentally), it will result in a security incident as PII can be decrypted. In this case, the next time a client application connects to the database, new keys will be delivered to replace the old affected keys. These will be marked as preferred so that the client application will immediately use the old keys to decrypt all documents containing PII encrypted with these old keys before re-encrypting the data with the new keys. At this point, the old keys will be deleted. This task will also be performed during normal scheduled key rotation, which can be performed in the event that a key leak has occurred but has not been detected.

[0059] In the above situation, clients that have deleted copies of the old key cannot read other clients that send documents encrypted with the old key. To achieve more seamless operation, the client can retain the old invalid key for a fixed period of time, such as 30 days. In some embodiments, a policy can be established that requires clients to connect to the database at regular, defined intervals to ensure that new keys are retrieved at regular intervals. This can quickly remove damaged keys from all client applications. In some embodiments, the expiration date of the key can be set to a shorter duration than usual in order to regularly change the key and mitigate the impact of any unknown leaks. This alleviates the problems associated with distributing keys to multiple entities, which may not be as easy to keep secure as a centralized system (or a system that is only distributed to a small number of endpoints).

[0060] In some embodiments, the document may be viewed by other software applications that have limited compatibility with the personally identifiable information retrieval service. Such an application may only present the plain text of the document, but may not be able to call the personally identifiable information retrieval service, and therefore cannot display or decrypt personally identifiable information. In some embodiments, when a document is sent outside the organizational boundaries (e.g., from one law firm to another), the receiving attorney may not have login access rights to be able to access the client database used to create the reference. Therefore, they see the edited text. In other examples, the receiving attorney can request access rights from the sending attorney. In some embodiments, this can grant access rights by creating a new user identity in the directory service and associating the relevant client records with the user identity and sharing the new user identity with the receiving attorney.

[0061] In some embodiments, the personally identifiable information that has been cleared may be displayed on a dashboard that allows the user 102 to view or delete documents that reference the user or other customers. Figure 8 An example of text in an original document that has been cleaned to remove PII is shown according to some embodiments. The document was sent by John Wash (lawyer) to Hilda Higgins (lawyer). The original document includes the following text (as shown in block 802):

[0062] I spoke to Tom Smith yesterday about this case. Tom (and his daughter Maria Smith) were surprised by the news.

[0063] The above text is a portion of the original document after cleaning using the cleaning application 420. The output cleaned document includes the following text (as shown in block 804):

[0064] "Yesterday I talked to [Ref ID=1321 / 1Val= Given Name ][Ref ID=1321 / 1Val= Last Name ] talked about the case. [Ref ID=1321 / 1Val=given name](and his daughter [Ref ID=1321 / 1Val= Daughter name ]) were surprised by the news.

[0065] As shown above, the PII is extracted and replaced with a reference (e.g., a client identifier reference). Figure 8 Also shown are the portions of text that various users see based on their user access privileges. For example, John Wash (lawyer) can see the following (as shown in block 806):

[0066] "I spoke to Tom Smith yesterday about the case. Tom (and his daughter Maria Smith) were surprised by the news."

[0067] Hilda Higgins (the attorney), who receives the document from John Wash, sees the following (as shown in block 808):

[0068] Yesterday I Talked about the case. and his daughter Surprised by the news.

[0069] The above example can be implemented by utilizing a search service that scans the various data repositories used by each organization associated with John Wash and Hilda Higgins. For example, a user can log in with their electronic identity for authentication. Since this electronic identity is bound to (or is the same as) their client identifier, the search will retrieve all documents containing their client identifier. Then, the user can view all documents containing references to it. If the document they read also contains references to other clients, the personal identity information related to other clients may be edited out. In some embodiments, since further workflows may be triggered, the user 102 may not see all of this immediately on the display 123, which may include dynamically verifying policies (e.g., according to rules provided in a rule database) or laws allowing a given document to be issued to a client, or not subject to legal proceedings. In other examples, the workflow may include a final manual review of the information retrieved by the personal identity information retrieval service. In some embodiments, the system 100 provided in the present application stores all documents containing a specific client identifier. When a user requests, this helps to provide all data associated with the user. If the server 104 does not know the existence of the document, this also prevents the document with a specific PII tag from obtaining the PII from the server 104, thereby preventing malicious users from inserting random client identifiers into documents in an attempt to obtain large-scale data access. In some embodiments, the user 102 (e.g., a staff member) can work with the client in person to give them a printout of all documents related to the client and all PII of other clients edited out from the document. In some embodiments, the client can request to delete the client's PII information in all documents associated with the organization or institution. In this case, the personal identity information protection system can ensure that the remaining documents remain intact while only editing the PII from each document. Further annotations (e.g., at the document template level or manually by the document creator) can determine whether to delete the complete document or simply edit it. In addition, at the client database level or through the operating system grouping policy, the settings for protecting PII in the document can be pushed to one or more authors' applications (e.g., word processing applications).

[0070] Fig. 91 is a flow chart illustrating a method for protecting personally identifiable information using encryption according to some embodiments. At block 910, the electronic processor 126 is configured to store an encryption key associated with a user who can access personally identifiable information in a document. At block 920, the electronic processor 126 is further configured to receive a document containing unencrypted non-personally identifiable information and encrypted personally identifiable information. At block 930, the electronic processor 126 is further configured to parse the document to determine the presence of personally identifiable information and detect encrypted portions in the document. At block 940, the electronic processor 126 is further configured to retrieve the personally identifiable information based on decrypting the encrypted portion using the encryption key. At block 950, the electronic processor 126 is configured to display a human-readable view of the document.

[0071] The embodiments described and illustrated in the above description and the accompanying drawings are not limited to the specific details provided in the present application, but may be modified in various ways. In addition, there may be other embodiments not described in the present application. Moreover, the functions described as being performed by a component in the present application may be performed in a distributed manner by multiple components. Similarly, the functions performed by multiple components may be merged and performed by a single component.

[0072] In the embodiments provided, components described as performing specific functions may also perform additional functions not described in the present application. For example, a device or structure that is "configured" in a certain manner is configured in at least this manner, but may also be configured in an unlisted manner. In addition, some embodiments described in the present application may include one or more electronic processors configured to perform the described functions by executing instructions stored in a non-transitory computer-readable medium. Similarly, the embodiments described in the present application may be implemented as a non-transitory computer-readable medium storing instructions that can be executed by one or more electronic processors to perform the described functions. As used in the present application, "non-transitory computer-readable medium" includes all computer-readable media, but does not include temporary propagation signals. Therefore, non-transitory computer-readable media may include, for example, a hard disk, a CD-ROM, an optical storage device, a magnetic storage device, a ROM (read-only memory), a RAM (random access memory), a register memory, a processor cache, or any combination thereof.

[0073] In addition, the wording and terminology used in this application are for the purpose of description and should not be considered as restrictive. For example, the use of "including", "comprising", "consisting of", "having" and its variations in this application means to cover the items listed thereafter and their equivalents and additional items. The terms "connect" and "couple" are widely used, and cover direct and indirect connections and couplings. In addition, "connect" and "couple" are not limited to physical or mechanical connections or couplings, but may include direct or indirect electrical connections or couplings. In addition, wired connections, wireless connections or their combinations may be used to perform electronic communications and notifications, and electronic communications and notifications may be sent directly or through one or more intermediate devices on various types of networks, communication channels and connections. In addition, relational terms such as first and second, top and bottom, etc. may be used only in this application to distinguish one entity or action from another entity or action, without requiring or implying any actual such relationship or order between entities or actions.

Claims

1. A computer system for protecting information associated with an entity, the computer system comprising: An electronic processor configured to: storing a record in a client database, the record comprising personally identifiable information associated with entities and a client identifier reference associated with each of the entities, the personally identifiable information being accessible based on an authorization level associated with the user; receive documents containing both non-personally identifiable information and personally identifiable information; parsing the document to determine the presence of personally identifiable information in the document; matching the personally identifiable information included in the document with records stored in the client database; replacing the personally identifiable information included in the document with a corresponding client identifier reference stored in the client database; as well as generating a human-readable view of said document including said corresponding client identifier reference; In response to first personal identification information matching only one record in the client database, removing the first personal identification information from the document and replacing the first personal identification information with a first client identifier reference; In response to the first personal identification information matching a plurality of records in the client database, providing for selection of a matching record in the client database; as well as In response to the first personal identification information not matching any of the records in the client database, the client database is updated with a new record and a new client identifier reference associated with the first personal identification information is generated.

2. The computer system according to claim 1, wherein: The entity is an item selected from the group consisting of a person, a group, an organization, and content.

3. The computer system according to claim 1, wherein: The electronic processor is further configured to: An encryption key associated with the user who has access to personally identifiable information is stored in the document.

4. The computer system according to claim 1, wherein: The electronic processor is further configured to: The human-readable view of the document is generated with a portion of the document edited based on access permissions to the personally identifiable information of an entity included in the document.

5. The computer system according to claim 1, wherein: The personally identifiable information includes items selected from the group consisting of: name, date of birth, place of birth, email address, telephone number, fax number, social network credentials, biometric information, financial account numbers, organization-issued identification, and government-issued identification.

6. The computer system according to claim 1, wherein: The electronic processor is further configured to: Delete all documents that contain personally identifiable information associated with an individual.

7. A method for protecting personally identifiable information of an entity, the method comprising: storing a record in a client database, the record comprising personally identifiable information associated with entities and a client identifier reference associated with each of the entities, the personally identifiable information being accessible based on an authorization level associated with the user; receiving a document containing both non-personally identifiable information and personally identifiable information of one or more entities; parsing the document to determine the presence of personally identifiable information included in the document; matching the personally identifiable information included in the document with records stored in the client database; replacing the personally identifiable information included in the document with a corresponding client identifier reference stored in the client database; generating a human-readable view of said document including said corresponding client identifier reference; In response to first personal identification information matching only one record in the client database, removing the first personal identification information from the document and replacing the first personal identification information with a first client identifier reference; In response to the first personal identification information matching a plurality of records in the client database, providing for selection of a matching record in the client database; as well as In response to the first personal identification information not matching any of the records in the client database, updating the client database with a new record and generating a new client identifier reference associated with the first personal identification information.

8. The method according to claim 7, further comprising: Retrieve all documents that contain personally identifiable information associated with an individual.

9. The method according to claim 7, further comprising: Delete all documents containing personally identifiable information associated with the individual.

10. The method according to claim 7, further comprising: An encryption key associated with the user who has access to personally identifiable information is stored in the document.

11. A computer-readable storage medium comprising instructions which, when executed by one or more electronic processors, cause the one or more electronic processors to perform the method of any one of claims 7-10.

Citation Information

Patent Citations

  • Method and system for identity exchange and recognition

    CN101084642A

  • System and method for protecting identification information

    CN1725226A

  • Data security system and method with editor

    US20050132070A1

  • Scrubber to Remove Personally Identifiable Information

    US20150213288A1

  • Document redaction with data retention

    US20160239668A1