Personally Identifiable Information

Personally Identifiable Information

Introduction

Information plays an important role in our lives. It drives the way we think, communicate and make decisions. The fast growth and accessibility of data, plus advances in technology, make extraction of information easier every day. Moreover the accessibility of information can play significant roles in research to derive new knowledge and make powerful advancements. However, there is information that is not meant to be public. Under federal regulations, e.g., the Health Insurance Portability and Accountability Act (HIPAA), information like medical records cannot be shared freely unless de-identified due to sensitive or confidential information. This limits their use in research and poses a threat to the solution of problems in the scientific community. Here is where text sanitization or text redaction comes to play. Text sanitization is the process of removing sensitive information from the data, so that it can be available to a larger group than intended for the original data. In general, text sanitization can be divided in to two parts: identifying personally identifiable information and hiding the discovered identifiers. Therefore, all the issues and aspects related to Personally Identifiable Information will be discussed in detail.

Discussion

Complete text redaction may not be necessary, since this can affect the readability and usability of the text. Identifying information, as well as sensitive information, can be quite subtle and still be present in the text even after the removal of obvious identifiers. This tells us that to truly sanitize a text requires hiding information that may go beyond obvious identifiers. The use of bad sanitization practices can result in re-identification of the information. These weaknesses lead to a major problem in text sanitization. Extensive work has been done related to anonymizing structured information with this technique. To this date, most text anonymization work is on finding and removing identifying information concentrated on medical documents. The degree of sensitivity in a dataset varies between applications. That is why the experts need to be careful with what is considered sensitive and how that information is released in order to preserve the privacy of the individuals in the dataset (Anton, Earp, Qingfing, Stufflebeam, Bolchini & Jensen, 2004).

Achieving secure deletion is difficult due to diverse threat models. To narrow down the problem, people focus on dead forensics attacks on local storage, which occur after the computer has been shut down properly. In other words, we protect against attacks that occur after a user cleanly unmounts and shuts down the system after completing secure-deletion operations. If the system crashes before the secure deletion operations finish, our framework must not allow the information that was being securely deleted to leak or be in an inconsistent state after the system is brought back up. This is due in part to the fact that few search websites require a pass worded account to use the service. Most websites making use of personal data, such as the social networks studied in this research, require at least one form of user ...