Convergence Research Machine Learning

Through constant training, it's software that gets smarter every day.

What is Machine Learning?

In short, machine learning is a process for a computer to make decisions using prior information and experience. People make decisions by being aware of what has transpired in the past and are able to learn from experience. They can apply this knowledge to new information to relate the new information to their past experience. Machine learning leverages the statistical analysis of past events (corpus) in conjunction with knowledge curated by a human (training). This data is built into a model when then can be used to evaluate new information (prediction).

Let's say you had to perform a task. Your task was to read a paragraph and then decide if the paragraph was about sports, news or politics. Although contrived, this is a simple task for a human but less-so for a computer. Traditionally, a computer would scan a block of text looking for words specific to a topic and, if it finds any of those words in the paragraph it would determine the paragraph was about the category containing those words. For example, if it finds the word "baseball" it can say the paragraph is about sports but this is a poor approach.

Why is this poor? Because the computer doesn't understand context the way a person does. Oftentimes when discussing politics, someone might use the term "inside baseball" to describe the inner-workings of Washington, DC. A person recognizes this term refers to politcs but a computer only looking for words in text will think it's about sports. The context is what matters.

This is where machine learning shines because it understands context. It learns the difference between how words are used versus the words themselves. With proper training, this is a powerful combination that acts as a cost-reduction and time-saving measure because you get the intuition of an employee with the speed and consistency of a computer.

How Machine Learning can help CRAs

The screening industry rewards speed and efficiency more than ever. In the past, certain human-centric tasks were more time-consuming, expensive per-hour and impacted your ability to service your clients' needs. Our solution can remove the FTE hours associated with:

  • Risk Factoring If your clients ask you to apply risk-factoring to their results, our Charge Classifier API can save full-time employee hours as well as provide a consistent, customizable solution.
  • Highlighting or diminishing certain types of offenses If your client is a property manager, being able to emphasize property or violent crimes while suppressing things they don't want to see like traffic offenses or ordinance violations.
  • Understanding your criminal results with more clarity Know how much of your criminal data is related violence, theft, traffic and more.
  • Custom Solutions If there's a task your team is doing with full-time employees today, we might have a solution that can save you time and money.

Do you want to learn more about how machine learning can save you money and super-charge your business?