Tuesday, 24. August 2010 24.08.10 12:00 Age: 8 Jahre

Predicting the Future of Markets
s-lab develops automatic market, trend, and mood internet analysis for firms

By: Katharina Bätz, English translation: Katrijn van Oudheusden

Accessing data faster: Henning Wachsmuth (s-lab) develops the software for automatic market analysis (Photo: Nadija Pejic)

Paderborn, Germany. For private internet users, a search engine can be a big help when trying to quickly gather relevant information. But when the goal is to gather multiple views on a topic, the search can take a long time. For example, what if a company wants to obtain a prognosis concerning a specific industry? Then efficient and trustworthy data gathering via the internet becomes much more important.

In order to provide not just private users with easy-to-gather data on the internet, but to also enable small- and medium-sized companies to find what they need, researchers at the Software Quality Lab (s-lab) at the University of Paderborn are currently working on the development of so-called “Information Extraction Technologies for Business Applications” (InfexBA). The work is part of a project funded by the Federal Ministry for Research and Education.

The project goal is to develop, together with three further partners, techniques for data mining to provide market, trend and mood analyses. “We want to develop technology that can provide an automatic prognosis on a specific market based on internet texts, as well as automatically identify trends and moods concerning products,” explains s-lab employee Henning Wachsmuth.

With the help of information-processing technology, the documents that contain information on company or market profits and losses are automatically selected from the masses of articles available online. Whenever the documents contain both time and monetary information, an artificial intelligence program automatically determines whether the information concerns profit statements.

Using information extraction, the search process for firms aiming to create a market, trend, or mood analysis, is greatly simplified. “It is certainly clear to us that the quality of our prognoses cannot compete with those of an expert,” says a realistic Henning Wachsmuth. “However, our technology helps to access the data faster and more comprehensively.”

A next step for the researchers will be to integrate a learning process into the technology that identifies company and market names. “We are very satisfied with the project to date; we have achieved important sub-goals,” explains Henning Wachsmuth. But that is not enough for the research team under the lead of project coordinator Prof. Dr. Gregor Engels: “In the next six months we need to work really hard to successfully complete the project,” says Wachsmuth.

In addition to the Paderborn s-lab, two software firms (Resolto Informatik and Digital Collections), as well as the Bauhaus University in Weimar, are collaborating on the project. The project is further subsidized with € 750,000 by the BMBF Program “KMU-innovativ: IKT”.

Henning Wachsmuth
s-lab – Software Quality Lab
+49 5251 60-3359