Cardiff-based AI research company, AMPLYFI has partnered with Georgetown University’s Center for Security & Emerging Technology (CSET) to develop a new machine learning platform capable of uncovering previously unlisted companies.
The platform, developed by AMPLYFI, uses AI to access previously unexplored data on the Deep Web, finding a large volume of ‘hidden’ Chinese companies that are AI-related, but not listed as such on leading commercial data sets such as Crunchbase and PEData/Zero2IPO.
With policymakers, market analysts, and academic researchers often using these commercial databases to identify artificial intelligence-related companies and investments, this new report suggests they may be making decisions without the full picture.
The Deep Web is the part of the internet that is not indexed by search engines and is usually made up of unstructured data. This data is typically anything that is written by humans for humans, such as patents, papers, reports and other listings and is historically difficult for machines to analyse.
This means most research tends to focus on more structured datasets held in business information platforms built around funding rounds and tax categorisations. The Deep Web is 400-500 times larger than the surface web, while 90% of data available on the web sits in unstructured data sets, meaning adding it into the analysis pool yields a far richer output.
Dewey Murdick, Director of Data Science at Georgetown’s CSET, highlights the value of this kind of study and calls for continued research into the AI sector: “AI has rapidly moved from the research bench to the application production line and it is essential to understand where and how AI is being applied within China, the United States, and countries around the world. CSET has improved its ability to measure AI adoption by working with AMPLYFI Ltd.”
Chris Ganje, CEO at AMPLYFI, said that access to more representative data was key for any future investment strategy: “Organisations that are making decisions without considering the value of unstructured and deep-web data are ‘flying blind’ and may not be aware of the huge steps forward in the abilities of machines to read, analyse and support strategic decision making. We were delighted to have supported CSET on this project and are excited to continue to demonstrate the power of our platform to extract value from this vast untapped online resource”.
