In the TexPrax project, the PTW and the UKP are transferring solutions for extracting implicit information from unstructured text in order to be able to process it automatically.

What is TexPrax about?

The responsible use of artificial intelligence (AI) is currently strongly discussed in many areas of industry. One such application area that would benefit greatly from AI is production-related problem solving. This is often controlled via a shop floor management system and requires a high level of effort. The automation of parts of the documentation, research or decision-making process would therefore facilitate the process considerably.

How can this goal be achieved?

A large part of the knowledge required for this is implicitly distributed among the employees and not usable for everyone in the organisation. This project demonstrates the use of text analysis algorithms to automatically extract implicit knowledge from chats and add it to the documentation process. In the process, essential relevant correlations, e.g. a cause-effect relationship, are recognised and stored from the company's written communication. The data basis can then be used to build assistance functions like knowledge sharing or the identification of recurring problems.

Who is working on TexPrax?

A close cooperation between production and text analysis is essential to achieve this goal. The cooperation of researchers from the Institute for Production Management, Technology and Machine Tools (PTW) under the direction of Prof. Dr. Metternich and the Ubiquitous Knowledge Processing (UKP) Lab under the direction of Prof. Dr. Gurevych thus creates optimal conditions for applying the knowledge for responsible, automated evaluation of unstructured chat data to improvement of business processes using the example of “Knowledge management in production-related problem solving”. Use cases in other areas of companies are also explored.

Your participation opportunity

TexPrax aims to transfer knowledge from the institutes to industry. In order to adapt the solutions to the needs of industry, interested companies can get involved in a participation forum. Current solutions are demonstrated and discussed in regular workshops. Companies will find support for the introduction of text analysis methods and can help shape the direction of further development. Dates, details and registration for the participation forum can be found at www.mit40.de

Acknowledgements

This project is supported by the Hessian State Chancellery – Minister for Digital Strategy and Development and the European Regional Development Fund (ERDF).

Funding source