Owners of software patents took a blow last June with the U.S. Supreme Court’s decision in Alice Corp v. CLS Bank International.[1] The Alice Court held that, to be eligible for patent protection, computer-driven innovations must include “an inventive concept” beyond computer implementation of an abstract idea.

The decision in Alice also holds particular significance for those that have—or want—big data patents. After Alice,the patentability of many big data inventions may now be in question. As a consequence, big data innovators must understand the issues Alice creates if they want a patent that survives its scrutiny. And some big data innovators may find that they can best protect proprietary technologies and systems by remaining open to alternatives outside of patent law when Alice’s requirements prove too difficult to meet.

Big data needs big software systems

The term “big data” already has multiple meanings. Sometimes, the term gets used to refer to the kind of huge data sets that are now constantly under creation. Other times, the term describes the real-time, data-based analytics, data management, or data mining   functions current parallel processing capabilities power.

Despite these different usages, big data’s true importance comes from the exciting new business solutions it enables. Big data lets organizations extract a whole new level of data-driven insights, value, and actionable strategies. That extraction, however, relies heavily on software and proprietary algorithms along every “V” in the big data formula of volume, velocity, and variety.[2]

As a result, big data will continue to need software-based innovations. Possibilities include software systems that:

• Improve upon data collection methods
• Offer advanced methods for incorporating a constant stream of new data
• Account for and analyze bad, uncertain, or unreliable data
• Enable the aggregation of data that includes protected personal information while preventing violations of consumer protection laws
• Allow for new ways to represent, transmit, and/or access insight gains

After Alice, big data must ask whether the decision will allow those who create these solutions to gain patent protection for their innovations.

Alice’s stance on software patentability

In Alice v. CLS, the existential question of the software’s eligibility for patenting seemed at stake. In Alice, the plaintiff-patent holder sought to enforce a patent covering a computer-implemented process that lessens settlement risk for financial instrument trades. The Supreme Court granted certiorari in the case after a deeply divided en banc Federal Circuit issued a seven-opinion per curiam decision rejecting the patent holder’s claims.

The Supreme Court’s decision did not go so far as to decide the ultimate question of software patentability. It did, however, unanimously reject the validity of the patent claims at issue. In its decision, the Court turned to the framework for assessing the patentability of potentially abstract ideas it has established in recent decisions outside of software.[3] As in those cases, the Alice decision highlights the concerns the Court has about letting a patent preempt nature, natural phenomena, and abstract ideas.

The Court also turned to the call for an “inventive concept” as laid out in Mayo Collaborative Services v. Prometheus Laboratories, Inc. [4] Under this analysis, the Court found the claimed invention lacked the necessary inventive concept needed to bestow patentability. It further concluded that the claims involved the abstract idea of intermediated settlement. In this case, the use of a computer in a “particular technological environment” was not enough to transform an otherwise unpatentable abstract idea into something patentable.

The Court also cited with approval Gottschalk v. Benson and its treatment of the patentability of computer-based algorithms. Benson said:  

The mathematical formula involved here has no substantial practical application except in connection with a digital computer, which means that if the judgment below is affirmed, the patent would wholly pre-empt the mathematical formula and in practical effect would be a patent on the algorithm itself. [5]

The claims at issue in Benson involved a simple algorithm and not any particular use or machine. Nonetheless, the Court’s emphasis on Benson gives pause, given big data’s heavy use of algorithms. But as one commentator aptly noted, the Court “misrepresents the nature of algorithms (which simply do not grow on trees)” by effectively stating that algorithms are species of abstract ideas. The result? “An entire shelf full of discredited cases on the metaphysics of what is and is not an algorithm must now be dusted off.”[6]

Big data patent battles is one potential place that such dusting off will occur. If so, software patent holders and seekers alike must become aware of the protection options available in the wake of Alice.

Protecting big data after Alice

As of now, software appears to remain patentable after Alice. But just what type of software remains patent-eligible is a question that will likely become the subject of significant future litigation.

Certainly, some classes of software now look unlikely for patent protection. This is especially true for those that do nothing more than use a computer to handle pre-existing general business processes. But complex solutions that analyze, manipulate, or store big data may be less vulnerable to the attack that cost Alice its patent. And at the very least, courts will again face the question of what is and is not an algorithm.

It is clear that big data patent holders will face challenges if their rights primarily rely on computer execution of nothing more than routine algorithms.[7] What, then, will be enough to make big data patentable? And what should holders do if patenting is not a reliable option?

Option one: Follow Diehr

One source of guidance on the patentability of big data comes from Diamond v. Diehr.[8] This is another case that the Alice Court cited with approval.

Diehr involved a computer-implemented process for curing rubber. The purpose of the invention was to provide precise, real-time recalculations of the remaining cure time. The patented process achieved this by using a well-known mathematical equation in combination with a recording device constantly measured temperatures inside a rubber mold. According to the Court in Alice, the invention in Diehr was patentable because it "transformed the process into an inventive application of the formula.[9]

Under Alice’s reading of Diehr, the effect that will have the greatest impact on patentability is a software solution’s ability to improve upon existing processes. Patent holders can look to Diehr to help them establish the requisite inventive concept as “sufficient” or “enough” to establish patentable subject matter. According to one commentator, when it comes to computers and uses of mathematical algorithms, “enough” may mean “action and demonstrating function beyond merely informing.”[10]

One way of doing this is to argue just how much the software actively “does something.” The good news for big data patent holders is that big data is complex. Processing, analyzing, manipulating, and storing big data is not a simple matter.[11] It involves complex methods and systems for quickly analyzing and storing mass quantities of structured, semi-structured, and unstructured data. So patent seekers or holders will only need to capture this complexity to survive judicial review under Alice.

Unfortunately, that very complexity will make big data patent disputes expert-intense and expensive. Parties will have to continue to battle out patentable subject matter issues. They will also have to address other patent validity standards, such as defining what constitutes ordinary skill of art when it comes to big data. One way to avoid such expense is to consider going an alternate route.

Option two: Turn to trade secret protection

Given the current patent protection environment, big data software innovators may consider looking at trade secret protection for their proprietary inventions. On the up side, trade secret law:

• Protects a wide range of processes and formulas
• Can keep confidential revolutionary or new algorithms, data structures, or methods for delivering content benefits
• Offers protections against employee or competitor misappropriation [12]
• Avoids subjecting trade secret damages to the same scrutiny that patent awards have recently received

As with any legal alternative, this approach does come with its own disadvantages. For instance, trade secret law:

• Does not preclude reverse engineering
• Does not protect its holder from later allegations of infringement in a patent proceeding
• Requires that the holder take active steps to keep the information secret (and have the ability to demonstrate those steps should a dispute occur)

Whether the benefits of this approach outweigh its drawbacks will depend on the specifics of the software involved. It will also heavily hinge on the IP owner’s appetite for risk.

Conclusion

To increase the odds of surviving a Section 101 patentability challenge, big data software patent holders will need to follow the groundwork laid out in Alice. Patent seekers or holders who are not sure their systems will stand up to the scrutiny may choose trade secret protection instead. In either case, all IP owners should continue to watch legal developments that rely on Alice. This is not the end of this discussion, but instead the beginning of a potentially drawn-out process to fully determine the patentability of big data software.

Alice Corp v CLS Bank Intl, No 13–298, 2014 WL 2765283 (US 19 June 2014).
Data volume continues to increase at an unprecedented speed. This requires continual innovation into scalable data storage options.  Data variety covers the multiple types of data and data source, including structured data that resides in a fixed field within a record or file and unstructured data from sources like videos, social media, and RSS.

The Court established this framework in cases such as Association for Molecular Pathology v. Myriad Genetics, Inc.; 569 U. S. ___, ___ (2013) and  Bilski v. Kappos, 561 U. S. 593 (2010).

First, we determine whether the claims at issue are directed to one of those patent-ineligible concepts. If so, we then ask, “[w]hat else is there in the claims before us?” To answer that question, we consider the elements of each claim both individually and “as an ordered combination” to determine whether the additional elements “transform the nature of the claim” into a patent-eligible application. We have described step two of this analysis as a search for an “‘inventive concept’”—i.e., an element or combination of elements that is “sufficient to ensure that the patent in practice amounts to significantly more than a patent upon the [ineligible concept] itself.” Alice Corp v CLS Bank Intl, supra, citing, Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. ___ (2012) (internal citations omitted).

Gottschalk v. Benson, 409 U.S. 63, 71-72 (1972).
Rob Merges, Symposium: Go ask Alice — what can you patent after Alice v. CLS Bank? SCOTUSblog (June 20, 2014), http://www.scotusblog.com/2014/06/symposium-go-ask-alice-what-can-you-patent-after-alice-v-cls-bank/
For example, patent claims that recite well known data processing algorithms, including linear algebra and basic statistic methods, are vulnerable to attack under an Alice-approved reading of Benson.

Diamond v. Diehr, 450 U. S. 175, 187 (1981).

Alice Corp v CLS Bank Intl, supra.
Emily Michiko Morris, Alice, Artifice and Action—and Ultramercial, Patently-O (July 8, 2014 ) http://patentlyo.com/patent/2014/07/artifice-action-ultramercial.html; See also Emily Michiko Morris, What is Technology? V. 20.1 B.U. J. Sci. & Tech. L. (Winter 2014).
For example, big data systems are often more complex than converting binary coded decimals in to binary numbers—as was the case in Benson.

Note that trade secret laws differ from state to state on this issue.