Our Online Dialogue colleague and analyst Arend Zwaneveld is for Webanalists.com started a series of articles. He looks at several real-world cases of Big Data examples and what we can learn from them. Today his fourth article: Making universities make money from their Intellectual Property

University patents: often gold in the basement ... with the basement locked!

Patents can make a lot of money ... at least, if those who use your protected idea pay you a license for the use of your patents. Patent users do not always come forward of their own accord: often, as a patent owner, you have to track down these parties yourself and point them out to them. intentional-or-not patent infringement.

Situation: thousands of patents untapped

Patents are like petroleum: underground they are of no use!

Universities generally hold masses of patents. Potentially these are worth a lot of money: the value of patents generated in 2007 by the top 100 American universities was estimated at USD 1.4 billion... That is, IF the universities were to monetize them!

University Patents = Big Bucks

However, universities have limited funds, and thus in practice often do not have the manpower To actually monetize their patents! Also North Carolina State University's Office of Technology Transfer struggled with this problem. ^[1]

Task: turning patents into cash

Labor-intensive, specialized human work

Making money with patents is hard work, you have to:

1. find a user of your patent
2. demonstrate that it uses your patent
3. letting the user use your patent for a financial fee (or a patent exchange)

Until recently, this was largely human work and therefore quite labor intensive. Nor is it easy!

North Carolina State University's Office of Technology Transfer manages 176,000 patents that are often only on detail level differ from one another. No man can keep track of so many patents, let alone know their contents and tell them apart: often only one on the specific sub-area specialized patent expert determine which of 2 patents (or both) applies to a product or technology....

Action: Big Data as a replacement for a small army of highly trained Internet-surfing student assistants

A Big Data solution from IBM consisting of LanguageWare™, Content Analytics and BigSheets (a kind of super-spreadsheet) helped NC State: based on a list of (often technical) search terms, it scoured (much of) the Internet, looking for matches to its own patent database. ^[2]

This Big Data system can read, scan, summarize documents - and make connections to other files - as well as an entire army of interns!

Input 1: Long-list URLS of potential Patent users.

Input 2: Database of 176000 patents from NC State University

Input 3: Patent related Keywords

Result: call list for patent experts

Output of this Big Data system is a shortlist of URLs/web pages that the system suspects are using patents from North Carolina State University's patent pool. Moreover, the Big Data Content analytics algorithm was able to immediately add the corresponding contact information for each potential patent user.

This Big Data solution thus saves NC State University a lot of detective work and immediately provides a call-list of patent users to whom licensing rights can be granted - for a fee.

Output: a call list of organizations to whom licensing rights can be sold

With repetitive patent-related sleuthing and research automated, NC State's patent experts - short-list in hand - can immediately pick up the phone and contact the tracked patent users.

More money, more patents ... more money!

North Carolina State University is finally reaping the rewards of its Intellectual Property (IP): ideas that it has taken great pains to protect - for a time - with a patent. And the proceeds? They can be used for better education, more research and ... more patents!

Additional benefit: information about the licenses paid for patent use can in turn be used to determine what type of research/patents a university can earn the most from!

Conclusions

1. Big data can make you money
2. Big Data systems can perform tasks that until recently could only be performed by humans
3. Big Data systems can, with the help of natural language processing Making unstructured data “somewhere in documents somewhere on a hard drive” usable again *
4. Big Data systems can extract information from data that has no value to the owner itself (gmail inbox, FourSquare check-ins, etc...).

Big Data Natural language processing and privacy

This clever piece of ‘human work’ shows not only the opportunities but also the privacy threats of Big Data: I sometimes catch myself thinking “there is so much on my hard disk, I don't even know what it all is anymore”. An (unwanted?) Big Data crawler has no problem with this: as a gmail user, I probably know less about myself than Google ... this information is worth money, so much so that all Google services are free.

Endless ideas for Big Data Natural language processing systems

With the ability of modern Big Data systems to understand natural (unstructured) language - some systems even understand video! - you can invent staggering new applications:

• real-time HR ^[3,4,5]
• a very good (!) search engine to search the files on your laptop
• an effective virtual police officer who uses all available court records to catch crooks
• Google ‘shrink’: I may know less about myself than Google does, so Google could start offering a whole psychology service so I can “get to know myself better” 🙂 ...

Who else knows of a great application of Big Data Natural language processing?
Suggestions are most welcome!

Originally posted on July 26, 2012 at webanalists.com

Resources

[1] Big Data University video - BigSheets patent Search NC State University (images screenshots from video)

[2] jstart portfolio - NC State: Matching academic research to business opportunity

[3] How BigData Tools Helps HR Understand You

[4] Why Human Resource Should Care About Big Data

[5] 10 pieces of advice for getting started with HR analytics