Communications of the ACM
Semantic Solution, A Patent Commercial Success
Credit: iStockPhoto.com
Information services company Matrixware has teamed up with a consortium of European researchers to enhance access to its database of patents. The commercial deal points to a bright future for software designed to help organizations make the leap to semantic systems.
The agreement includes almost a million euros in funding for researchers at the University of Sheffield in the United Kingdom and Bulgarian semantic technology company Ontotext to annotate patent information with semantic data and implement a semantic search engine to make existing patents faster to find and retrieve.
"The goal is to enable patent professionals to do their work more efficiently. Keeping track of patent information is a huge challenge because you're dealing with terabytes of data and the number of patents is constantly growing," explains Kalina Bontcheva, a senior researcher at the University of Sheffield.
The deal with Vienna-based Matrixware is one of several likely commercial implementations for the semantic knowledge and content management solutions developed by researchers in the EU-funded TAO project, which was coordinated by Bontcheva and involved seven partners in five countries.
"We've had a lot of interest and have been approached by several companies with a view to implementing our technology," Bontcheva says.
The software developed in the TAO project is designed to address the demands of companies and organizations that have to deal with large amounts of data, be it patent information, Web site content or databases of manufacturing components.
Semantic data, which allows computers as well as humans to understand the meaning of text, images and other so-called data artifacts, makes information easier and faster to retrieve. But transitioning from legacy databases and applications, which often use obscure programming languages and complex structures, to semantic systems that require ontologies — vocabularies of related concepts used to assign machine-readable meaning to information — is far from a simple task.
Automating Transitions
"Transitions from legacy systems to semantic ones can be performed manually, but it takes a lot of time and effort, and it is often impractical due to the sheer volume of data involved," Bontcheva explains. "In TAO, we created tools to automate much of the process, allowing organizations to reap the benefits of semantic systems in a much more cost-effective way."
The approach requires implementing software that builds ontologies to describe products, services and processes within an organization by learning from the descriptions, file locations and other information already contained in legacy systems and databases. Different data artifacts are then automatically categorized and annotated semantically, ready for rapid retrieval via a semantic search engine.
"For example, an aircraft door that in a legacy database probably just had a part number will now return information about its location, components, maintenance requirements, date of manufacture and so on," Bontcheva notes.
That is the sort of structured and explicit information Dassault Aviation was seeking when the French aerospace group and TAO partner implemented the tools as part of the project. The company has since gone on to lead standardization efforts within the aviation industry for plane and component maintenance data using the ontology developed in TAO as a basis.
"Manufacturers, who often have large databases of information on designs, components and products accessed by subsidiaries in different countries, are a key industry for semantic solutions. However, with more and more data being generated and stored by companies of all kinds, almost any business in any sector could benefit," Bontcheva says.
Intense Demand
For that reason, the TAO project coordinator does not foresee companies cutting back on plans to adopt semantic information systems in the near future despite the global economic crisis. In fact, the economic problems could intensify the demand for semantic solutions from some industries, such as banking and financial services, that are facing increasing regulatory requirements.
"For others, despite the initial investment, it is a way to increase competitiveness and save money in the long term," Bontcheva says.
The TAO tools, most of which are open source, could be used by almost any company with any kind of database and software requirements to transition to a semantic system. For example, the software has also been implemented by the company behind videolectures.net, a Web site dedicated to distributing recorded presentations.
"Because the software is open source, companies can implement it as they want. However, it is likely that they will turn to us to customize the tools to meet their specific needs," Bontcheva says.
Though the TAO partners are not currently planning to launch a spinoff company as a result of their work, they have not ruled out setting one up in the future. In the meantime, they are actively seeking investment and business partners.
The TAO project received funding from the ICT strand of the EU's Sixth Framework Program for research.
From ICT Results
No entries found