In our last Big Data blog posting, we cautioned that the protection of the intellectual property rights (IP) in Big Data may warrant its own focus. While there are legitimate concerns about finding IP in data, because data may be an inert lump of code, bits, or pieces of information, it is worthwhile to think about the different kinds of IP that arise in conjunction with and in the context of Big Data. This blog entry focuses on the IP opportunities ‘in relation to’ Big Data.
The IP asset candidates to consider for Big Data are in four possible categories: utility patents, trademarks, copyrights and trade secrets. Each of those four categories represents an opportunity to put a fence around and own some aspect of the Big Data project; however, a key point in understanding the IP opportunities, is understanding that each of the four categories is best PROACTIVELY claimed. Waiting is the way to lose in the Big Data IP ownership game. The biggest challenge to extracting the IP assets from Big Data may well be whether, and how successfully, the would-be IP owner is in getting its team to proactively take the steps to protect the IP in Big Data (including, big surprise, working with its lawyers).
What follows is an over simplified category by category map of where the IP assets may be found in relation to Big Data:
Big Data may be a literary work either by forming original expression fixed in tangible medium or by having been structured, sequenced or arranged in an original expression. When copyright can be claimed, the key to taking full advantage of the possible opportunities is getting a copyright registration, such as, from the United States Copyright Office and also from any other relevant jurisdiction. There are substantive benefits to getting a copyright registration earlier rather than later.
The second two IP categories hinge upon algorithms that Big Data employs. It is through algorithms that Big Data speaks to us. The algorithm is the process for calculation, recombining different sources and extracting meaning. Algorithms are how the ‘small’ data is turned into Big Data and into intelligence.
B. Trade Secret
A trade secret can be any data that is not commonly known to others, that gives the owner a competitive advantage and that is properly protected under trade secret rules. Data is often a trade secret. So is Big Data often a trade secret. Big Data typically has three characteristics:
- Volume – the daunting huge size of the amount of data in Big Data
- Velocity – the unprecedented processing speed necessary to access and process in the algorithm
- Variety – the widely disparate sources and formats from which the data is extracted and which change constantly.
To protect the trade secrets in Big Data, the value of the trade secret will likely lie in the potential for re-use and the recombination of the particular data that forms the Big Data flow. Affirmative and documented processes for keeping the Big Data proactively protected from disclosure will define the trade secret status.
An effective trade secret program is dynamic and will impose strong confidentiality on the Big Data and the processes that allow that Big Data to effectively and regularly block access by others to it.
C. Utility Patent
Patent strategies differ from one industry to another and might be the entire subject of a longer blog. Briefly, an issued patent articulates exclusive rights granted on novel and nonobvious technical inventions. A gating event to the granting of a patent, is that the patent office will know existing technology as of the time of filing the patent application and will refuse the application if the technology described is not both ‘novel’ and ‘nonobvious’.
The patent application must include the explicit disclosure of how the novel, nonobvious invention is rendered. This disclosure aspect of a patent application is one of the points making the utility application, perhaps, a suboptimal IP protection category for Big Data. For ongoing effectiveness, Big Data algorithms must be adapted continuously to have relevance and business value. Big Data, by definition, is data that continuously changes its volume, sources, velocity and behavior. While utility patents may apply, the dynamic development cycles of Big Data analytics and algorithms may not be optimally protected with patents.
In addition to these three IP asset categories, ownership in Big Data may be protected by contract. Financial Services is a market sector that has long protected ownership of data and Big Data under increasingly sophisticated contracts. Other industries that have protected their Big Data under contract include air transport industry (fares data), healthcare (clinical outcome data), and recorded music (track based metadata).
As noted above, this is a too-brief overview of the spectrum of IP opportunities to protect Big Data. However, benefits may flow to the business that proactively becomes knowledgeable and selective in considering IP protection for Big Data.