For Wall Street, a Matter of Semantics
New standards, technologies, products prompting adoption
February 25, 2008
Semantics--the study of meaning in language--may date back to the ancient Greeks, but only now is a technological take on the subject gaining traction on Wall Street. Due in large part to the ever-escalating volumes of data that trading firms must digest on a daily basis--Aite Group reports a near-quadrupling of U.S. equity messaging since December 2006--they are turning to so-called semantic technologies for help.
Semantic technology applies meaning, or markers, to disparate data and text in a standardized manner, making them machine-readable and interoperable, which allows computers to more easily correlate information, communicate about it, and then make the big jump--performing tasks that call for reasoning and inference.
Firms such as Ameriprise Financial, Aon and Citigroup have recently initiated internal projects to explore the potential benefits of semantic technologies, according to "Semantic Wave 2008," a report issued last month by Washington, D.C.-based research firm Project10X. Credit Suisse, HSBC, Morgan Stanley, State Street Corp. and Wells Fargo Bank have been identified by vendors and industry reports as having or considering semantic-based technology initiatives.
There are "pockets of semantic tech development in almost every large organization in the financial services community," said Eric Miller, a computer scientist affiliated with the Massachusetts Institute of Technology and CEO and founder of Zepheira, a Fredericksburg, Va.-based technology consultancy.
"The use of semantic technologies by Wall Street firms is increasing and its implementation is inevitable," remarked Mary Knox, a Timberlake, N.C.-based research director in the banking and investment services division of Gartner.
Christine Connors, global director of semantic technology solutions at Dow Jones, said that, "up until now, the focus at a lot of Wall Street firms has been on the subject of speed, but the next wave of technology focus is going to be, how quickly can we retrieve or produce financial data that has some analysis capability already built into it."
Front-office tasks, such as filtering news feeds and data, and back-office work like data integration have seen early implementation. But the list of relevant activities is much broader, including algorithmic trading, business rules for investment strategies, risk and compliance management, due diligence, sales and customer service, auditing transparency, and fraud detection and prevention.
Services and Standards
Several factors are pushing the adoption of semantic technology, observers say, including new, enterprise products from established database software providers such as Oracle Corp., semantic-based services from financial data suppliers like Reuters and Dow Jones, and the development of easier-to-use semantic language standards--Sparql and Sawsdl.
Sparql is a query language that was officially recognized last month by the World Wide Web Consortium (W3C), while Sawsdl provides a framework for coordinating service-oriented architectures with semantic technologies. The two can be used in concert with previously endorsed W3C standards such as the resource development framework (RDF) and Web ontology language (OWL), giving developers the ability to organize data in such a way that machines can understand contextual meanings and nuances.
Over the past few years, a number of start-ups have emerged that use standards-based semantic capabilities to provide new types of search and data management capabilities, including AdaptiveBlue, Powerset, Radar Networks and Metaweb Technologies. The latter, based in San Francisco and founded by artificial intelligence expert Danny Hillis, last month received $42.5 million in venture capital funding from Goldman Sachs and Benchmark Captial.
