Monetize data, the most valuable asset of Machine Learning | Technology

Monetize data, the most valuable asset of Machine Learning | Technology
Monetize data, the most valuable asset of Machine Learning 

Data Monetization is the most important asset of Machine Learning 

The data associated with machine learning can be extremely valuable, but, Kimberley Bayliss of Haseltine Lake Kempner writes in this co-edited article, before it can be monetized, there are some major issues to be resolved.

One of the things I hear over and over again from inventors is that data is the most valuable asset in machine learning (ML). After all, an ML model is only as good as the quality and quantity of data on which it is trained.

In many fields, good quality data that can be used to train ML models is proprietary and can be plagued (depending on your point of view) by privacy issues. Also, as soon as you give a third party access to your data, it can be copied, making it easy to lose control completely quickly.

If data is really that valuable, the burning question is whether it can be successfully protected and monetized.

READ MORE: How AI and Machine Learning are changing the corporate landscape | Technology and Business

Be clear about what you have and the responsibilities of each individual

Conduct a data audit to determine what you have and how it is used. Find out what’s in each database, who in your organization has permission to access it, and for what purposes.

Just as employees must be aware when they access a trade secret, and the responsibilities that come with it, employees must also be aware of their responsibilities when accessing and using company data. This minimizes the chances of your employees accidentally sharing valuable data.

READ MORE: How to develop machine learning skills in all of your company’s employees | Technology

Keep records on how the data was compiled and processed

Both copyright and database rights can be used to protect databases in the UK and the EU.

Copyright protects original (eg, creative) selections or arrangements of material in a database. However, the content of a database is protected by database rights, if there has been a substantial investment in obtaining, verifying or presenting the data.

Therefore, in both cases, it is good practice to document how your data was collected and processed to demonstrate that these rights exist.

The following are some points to consider when looking to turn data into a potentially revenue-generating asset.

READ MORE: An Artificial Intelligence technology shows promise in predicting the danger of pulmonary nodule cancer | Technology

Consider Patent Protection 

Patent Protection is not necessarily the first thing that comes to mind when we think of protecting a database, but Article 64 of the European Patent Convention explicitly provides protection for products obtained directly through patentable processes.

Furthermore, the EPO Examination Guidelines in Section 3.3.1 state that: “When a classification method has a technical purpose, the steps of generating the training set and training the classifier may also contribute to the technical nature of the invention. if they help achieve that goal. technical purpose.

Therefore, it seems theoretically possible to obtain a patent with claims on a data processing method (for example, to optimize the data for use in training a machine learning model), which also extends to a produced database. for that process. If your data is changed to improve a technical process, patent protection is worth considering.

READ MORE: Artificial Intelligence will make Indian roadways safer to travel on | Technology

One size does not necessarily fit all

Once you’ve audited your data collections, the next question is how valuable each data set is to your business. In other words, what data gives you a competitive advantage?

While you may want to keep the data that contributes the most to your business a trade secret, this may be overkill for other data assets that provide less benefit to your organization. It is this data that is ripe for monetization.

READ MORE: The use of Augmented Reality in Real Estate is no longer a gimmick | Technology

Derivative or Real Products?

Once the data has been selected for monetization, a range of options opens up to you. These include the licensing and sale of data; for example, through an IP or a data broker.

Both the data itself and products derived from it, such as trained models or other predictive tools, may be sold to third parties. Derivative products may be hidden behind application programming interfaces (APIs) and made available to third parties; for example by using a subscriber model.

While it may feel safer to sell or give access to derivative products, this does not necessarily protect the underlying data, as datasets can be reconstituted from ML models. Pull attacks, in which an attacker makes a large number of requests to a model, can be used to create a database of input/output pairs, or to probe the boundaries of the model to determine the underlying logic. Therefore, both the training data and the model structure can potentially be reconstructed, simply by querying a model.

READ MORE: Using AI in the Financial Services Industry | Technology

No Magic Bullet

As is often the case, there doesn’t seem to be a magic bullet that allows for data protection and easy monetization. However, a deliberate approach with a clear paper trail is likely to offer the best chance of realizing the value of your data sets while maintaining your rights.

Source: IAM Media, Direct News 99 

Leave a Comment