Fraunhofer Institute for Applied Information Technology FIT
Explorative Text and Data Mining: Use your intuition and experience to exploit the full potential of your knowledge
A large portion of an enterprise's strategically relevant information is contained in text documents. No matter if you look for patents, product descriptions, meeting minutes, letters from your customers, or articles from the trade press: You'll have a hard time getting an overview of the huge amount of unstructured information. Single documents may have many hidden but valuable relationships. But that's not all: Your text data is related to many other pieces of information, like product or project data, customer records, human resources and departments.
Fraunhofer-FIT's text- and data analysis tool SWAPit not only visualises semantic relationships among text documents. It also clearly arranges all associated information on a single screen! This gives you an X-ray view for your document archives. Using your intuition and experience you'll be able to exploit the full potential of your knowledge:
- Your CRM staff detects the characteristics of those customer groups that share common interests in your products and services.
- Analysing inquiries to your customer support centre, your marketing department realises which features could enhance your products - and which segment of your customers is ready to pay for that.
- With just a few clicks your R+D department knows how the patent situation in your technology area has developed over time.
SWAPit has been successfully used in many applications, e. g.
- Business intelligence: Navigate trade news and link news items to internal KPI measurements.
- Insurance companies: Relate customer complaints with master data and contract information.
- Software engineering: Group and structure large collections of use cases.
- Quality assurance: Analyse technical documentation.
- Innovation management: Cluster and analyse ideas and brainstorming notes.
- R+D planning: Analyse patent portfolios.
- Knowledge management: Analyse documents and extract terminologies.
Central component of SWAPit is the DocMINER text mining engine, a technology that automatically detects relationships within text collections and generates graphical overviews (document maps). DocMINER makes it comfortable to find a way through the mass of information. Documents can be linked to relational data from databases and taxonomies (catalogues), thus complementing the overall picture of available knowledge.
SWAPit combines state-of-the-art methodologies from statistics, retrieval, artificial intelligence and visualisation. SWAPit's modular architecture turns the integration into existing infrastructures (e.g. company portals) basically into a matter of customisation. A thin client is served by web services for analysis, visualisation and interaction. Inter-service communication is XML-based (SOAP protocol). The graphical user interface is realised in Java™ technology, making the approach independent of operating system and browser.