Thomson Reuters to open-source company data analysis tools

Beverley Head
Financial information services company Thomson Reuters has revealed plans to open-source the identifier system it developed to index every listed company in the world.

Having grown through acquisition over the years, Thomson Reuters developed the identifier system to bridge and make sense of the multiple silos of data that it developed or has access to. The company takes in data from 1400-plus sources worldwide.

In order to turn those silos into an accessible content marketplace that could then be analysed for insights, Thomson Reuters developed a standard ontology that establishes a permanent identifier for every piece of data, using a 256-bit code to identify every company in the world.

That capability has already been implemented in the Accelus Org ID system currently used by a handful of organisations worldwide, including one Australian bank. A know-your-customer service, it cross-checks data from 240 countries, 400 sanction, watch and regulatory lists, and 100,000 media sources to identify high-risk individuals.

Thomson Reuters now plans to open-source the identifiers used in that solution.

It's part of a bid by the company to "externalise the information model" and "create an ecosystem for interoperability" according to Debra Walton, chief content officer for financial and risk at Thomson Reuters.

Walton said Thomson Reuters takes information feeds from 1400 separate sources, including the Sydney based and newly fledged RoZetta Technology, an offshoot of university-owned big data specialist Sirca. RoZetta sells access to its big data services, including a service that processes, analyses and graphs every trade on every financial market in the world in two and a half hours.

Walton said that it was not sustainable to bring all these different data feeds into a central location for analysis. By instead externalising the information model it was however possible to create an "ecosystem for interoperability" that could be harnessed by any organisation choosing to adopt the open source identifiers.

Walton, an Australian-born, now New York resident, said during a visit to Sydney last week that; "In financial markets information analysis has always been the lifeblood." The advent of so-called big data, which can include conventional information from computers, insights from social media and feeds from sensors, means the challenge to collect, collate and analyse that data are magnified.

"We are now in a paradigm shift...we see data come to the fore - holding the potential of advantage through insight," said Walton. She said that information from wearable technologies, smartphones, social media as well as more traditional data sources was all now available for analysis by the financial sector.

Thomson Reuters itself has a patent for a system able to analyse data for patterns that predict credit rating downgrades. She said the StarMine system had already predicted problems in two US companies that subsequently went bankrupt.

A similar approach to data analysis could, said Walton, be used to predict potential future mergers and acquisitions.

Following a 2011 acquisition of Chicago based Lanworth, Thomson Reuters is also able to take in geospatial data mapping the world's crops, which when combined with data such as weather information, can be used to predict crop yields, giving traders an early warning about when to buy futures.