As the amount of information we share online grows, so does our ability understand the true meaning behind what people write on the Internet. Isis Forensics is a new company with potential to capitalize on this in a big way.
This UK-based startup has developed the ability to create accurate profiles of people online based entirely on the way they use language in their writing. The first commercial use the technology has found is with a law enforcement body in the UK, where it’s being used to help identify cases where paedophiles are posing as children on social networks in order to groom potential victims.
Isis Forensics analyses text by comparing it against reference data, to identify factors such as the likely age and gender of the author. I witnessed a demo of the process recently and, using data from my Twitter account, I was correctly identified as a male in his early 30s, who is naturally cautious and measured in his approach to life. The system decided this based on the kinds of topics I discuss on Twitter, the words I use (cautious people like me are more likely to use “maybe,” “perhaps” etc.), and more subtle factors related to how language is used differently by different types of people.
Isis Forensics is a spin-out from Lancaster University in the north of England, building on twenty years of research and development in the field of natural language processing. The company has just announced “significant” joint funding from The North West Fund for Venture Capital and Rosebud, via Enterprise Ventures (EV). Although the amount hasn’t been disclosed, EV typically invests between £25,000 ($40,100) and £2m ($3.2m) in its portfolio companies.
Serious potential
Although Isis’ technology is initially being used as guidance in gathering evidence of online child abuse, the potential applications here are really broad – from helping financial institutions tackle fraud, to allowing businesses to better understand their customers. For example, a 40-year old person may traditionally be targeted with marketing aimed at general 40- to 50-year-olds, but if that particular person tends to use language more like a 30-year-old, they may well respond better to advertising that uses that kind of language.
Isis is currently working on an API that will allow third-party developers to create ‘toolkits’ to adapt its core technology for all sorts of purposes. We’ll be keeping an eye on this company’s progress.
Disclosure: The author is a co-founder of TechHub Manchester with Isis Forensics’ VP of Business Development, Ben Hookway.
Image credit: iStockPhoto / Thinkstock
Get the TNW newsletter
Get the most important tech news in your inbox each week.