Boston College Law Review 

About: Boston College Law Review is an academic journal. The journal publishes majorly in the area(s): Supreme court & Government. It has an ISSN identifier of 0161-6587. Over the lifetime, 971 publications have been published receiving 4402 citations.

Big Data and Due Process: Toward a Framework to Redress Predictive Privacy Harms

Abstract: INTRODUCTIONBig Data analytics have been widely publicized in recent years, with many in the business and science worlds focusing on how large datasets can offer new insights into previously intractable problems.1 At the same time, Big Data poses new challenges for privacy advocates. Unlike previous computational models that exploited known sources of personally identifiable information ("PII") directly, such as behavioral targeting,2 Big Data has radically expanded the range of data that can be personally identifying.3 By primarily analyzing metadata, such as a set of predictive and aggregated findings, or by combining previously discrete data sets, Big Data approaches are not only able to manufacture novel PII, but often do so outside the purview of current privacy protections.4 Existing regulatory schema appear incapable of keeping pace with these advancing business norms and practices.Personal harms emerge from the inappropriate inclusion and predictive analysis of an individual's personal data without their knowledge or express consent. For example, in 2012, a well-publicized New York Times article revealed that the retail chain Target had used data mining techniques to predict which female customers were pregnant, even if they had not yet announced it publicly.5 This activity resulted in the unauthorized disclosure of personal information to marketers.6 In essence, Target's predictive analytics "guessed" that a customer was pregnant and disclosed her name to their marketing department, manufacturing PII about her instead of collecting it directly.7 Although the customers likely knew that Target collected data on their individual purchases, it is doubtful that many considered the risk that Target would use data analytics to create such personal customer models to send advertising material to homes. These types of harms do not necessarily fall within the conventional invasion of privacy boundaries, but such harms are still derived from collecting and using information that centers on an individual's data behaviors. We call these "predictive privacy harms."This Article confronts the tension between the powerful potential benefits of Big Data and the resulting predictive privacy harms. Part I discusses the nature of "Big Data science" and how personal information can be amassed and analyzed.8 It then discusses the nature of predictive privacy harms and why traditional privacy protections are insufficient to address the risks posed by Big Data's use of personal information.9 In Part II, this Article recounts the Anglo-American history of procedural due process and the role it has played in both systems of adjudication and separation of powers.10 Part II then makes the case for why procedural due process principles may be an appropriate source to draw from to address the risks of predictive privacy harms.11 Finally, Part III looks at the procedural due process literature and suggests ways to analogize a similar framework for private sector Big Data systems.12I. PREDICTIVE PRIVACY HARMS AND THE MARGINALIZATION OF TRADITIONAL PRIVACY PROTECTIONSA. What Is Big Data and Why All the Hype?Knowledge is invariably a matter of degree: you cannot put your finger upon even the simplest datum and say "this we know". In the growth and construction of the world we live in, there is no one stage, and no one aspect, which you can take as the foundation.-T.S. Eliot13"Big Data" is a generalized, imprecise term that refers to the use of large data sets in data science and predictive analytics.14 In practice, the term encompasses three aspects of data magnification and manipulation.15 First, it refers to technology that maximizes computational power and algorithmic accuracy.16 Second, it describes types of analyses that draw on a range of tools to clean and compare data.17 Third, it promotes the belief that large data sets generate results with greater truth, objectivity, and accuracy.18 The promise of Big Data's ability to analyze data and provide novel insights has led to profound investment in, consideration of, and excitement about Big Data's power to solve problems in numerous disciplines and business arenas. …

Ten Things to do About Patent Holdup of Standards (and One Not To)

Abstract: A central fact about the information technology sector is the multiplicity of patents that innovators must deal with. Indeed, hundreds of thousands of patents cover semiconductor, software, telecommunications, and Internet inventions. Because of the nature of information technology, innovation often requires the combination of a number of different patents. Currently, various features of the patent system facilitate holdup, particularly in the standard-setting context. These features include insufacient discounting in damages for patent infringement and the resultant inoated demands for royalties, the low standard of proof for willful infringement, which allows patentees to recover treble damages, and the threat of injunctive relief. Frequently, innovators make irreversible investments in their development of new technology, only to have those investments used against them as a bargaining chip by existing patent holders. This Article suggests ave steps that standardsetting organizations may take to reduce the problem of patent holdup and ave ways the law should change to deal with the problem.