Wolfram Alpha Teams with Xignite to Deliver Financial Data

By SemanticWeb.com on August 21, 2009 12:58 PM

Ron Miller
SemanticWeb.com Contributor

Wolfram Alpha, the data-driven search engine that was released to so much fanfare last spring announced an agreement recently to use Xignite financial-information web services to deliver real-time financial data to Wolfram Alpha. This agreement marks the beginning of what could be a series of vertical data sources that could be added to Wolfram Alpha’s growing database over time.

Xignite supplies real-time financial information such as market data, interest rates and exchange rates via web services explains Joel York CMO at Xignite. “We are a pure software service provider of market data,” York says. Peter Overmann, who is executive director of research and development for Wolfram Research, Inc., says Xignite has contracts with many different sources, which saved them from having to try to integrate information on a more ad hoc basis. “We just get it all directly from Xignite,” he says.

How it Works

Overmann explains that integrating Xignite was a snap because Wolfram Alpha has developed expertise linking to all kinds of data sources including web services.

“From our point of view, there was no technical difficulty,” he says. He says a user starts by entering a natural language English question such as “biggest automotive company, or What is General Electric’s dividends per share in Euros? (See graphics.)

In this query, Wolfram Alpha uses Xignite data to determine the largest automotive company.

In this query Wolfram Alpha uses Xignite data to compute General Electric’s dividends per share cost in Euros.

Then Wolfram Alpha determines it has financial properties such a a company name, business type or other financial property that enables the search engine to determine that it is a question about finance.

From there, the engine attempts to compare the question to its financial data source, in this case Xignite’s subservice and do the necessary computations to deliver a correct answer to the user. “This happens in a really short time,” Overmann says, “so users wouldn’t notice any delay.”

Looking Ahead

For now, Wolfram Alpha will offer the financial data for free via the Wolfram Alpha web site, but they have plans to add more data sources and eventually plan to release a premium or pro version of the service. “We are working with many data providers,” Overmann explains, adding that Wolfram Alpha is still in the early stages of development with much more to come. “We are growing every week,” he says “and every version has one new data category and we see no end to this.”

Russell Foltz-Smith, Head of Business Development for Wolfram Alpha says they will also be releasing an API very soon, which Foltz-Smith says should enable organizations to integrate Wolfram Alpha into their own platform. As they add additional vertical data sources similar to the financial one they provide with Xignite, Foltz-Smith believes companies will tap into these vertical sources using the API.

In addition, the company, much like Salesforce.com and Amazon.com, plans to make available the power of its server farm, which Overmann points out has more than 10,000 computers. He sees Wolfram Alpha as a platform on which companies can build search applications, store data and tap into the power of the Wolfram Alpha engine. “It’s definitely a platform for [customers] to use their own computations and their own methods to extend the system by asking questions in natural language,” Overmann says.

Semantic Qualities

Semantics come into play with Wolfram Alpha, and Foltz-Smith says, this will be even more apparent once companies tap into the upcoming API. “Developers will take the API and use the linguistic abilities to connect to different data sources.” He admits this might not be the “holy grail of semantics,” but it provides a way to do some interesting things involving machine to machine communications using Wolfram Alpha’s structured data.

Xignite’s York points out that people have compared Wolfram Alpha to Google, but he says, “Google doesn’t have a clue what’s in links. It has no context. It doesn’t know if you are searching for tennis rackets or cheese puffs, but Wolfram Alpha knows what the user means.” He says this enables Wolfram Alpha to take a question and using Xignite or another data source, provide an answer.

When Wolfram Alpha was released in the Spring it was easy to criticize its data out of the box as being too limiting, but by teaming up with companies like Xignite to give users access to growing amounts of real-time data, it increases its usefulness and raises many new possibilities for usage moving forward.

Comments