• contact:

  • project group:

    WG1: Process Engineering

  • funding:

    Industry project

  • startdate:


  • enddate:



Project name: Analysis and semantic structuring of product information with the example – distribution & sales (Consis)


Keywords:  heterogeneous data sets, knowledge management, ontology-based information retrieval, classification schemes



Issues & Challenges:

Data and information on developed products, as well as any associated documents often exist in several of the companies’ departments. The different information types need to be differentiated:

  • Textual information – information presented through alphanumeric characters
  • Graphic information – information in the form of diagrams, graphs, drawings, sketches
  • Image information – information presented through general objects and their correlations
  • Film information – information presented through general actions, processes, dynamic procedures
  • Audio information – recordings of mostly natural targets (tones, sounds, language) or synthetically manufactured acoustic information

The magnitude of digital data grows exponentially over time during the product development phase. The complexity of the problem is increasing due to the unavailability of adequate information logistics concerning the variety of sources and their presentation formats. This is furthermore complicated by the fact that technical data generated in the product development processes is supposed to be interdepartmentally accessible. The improvement of the access to the relevant product information – and at the same time its analysis and structuring – is perceived as a great challenge.



The objective of this industrial research was the development of a concept for the analysis and structuring of product information, which supports the user in the administration and procurement of relevant information from a heterogeneous data set (e.g. documents, files), by means of a unified terminology. The following sub objectives can be defined:

  1. Survey of the enterprise systems and their generated output information. The priority within this activity lies on the examination of the relevance of the output, as well as on enhancing existing terminology, should this be necessary.
  2. Examination of the heterogeneous system landscape with regard to generation of data. The manner in which data/information is entered into the systems is most important here. The data is often still inserted manually by the user. Therefore, ways of how to automate the information procurement have to be investigated. The information’s origin should also be retraced, in order to comprehensibly document the development background.
  3. Identifying the interactions between the enterprise systems. The systems should be analyzed, in order to discover the dependencies between data/information used by the systems. These dependencies can then be used to reduce the information redundancy often encountered in enterprise systems.
  4. Analysis of possible information structures with regard to usability. The structure of the information is essential for its efficient application. However, not every structure is suited for all information content. Therefore, possible visualization methods (e.g. sequential structure, hierarchical, structures, column-based structures etc.) have to be analyzed and evaluated in the course of this research.


Research Strategy & Solution Statement:

A feasible approach consists of generating an interdepartmental, application-oriented nomenclature or classification. This can result in improved possibilities being created for linking requirements to new products with information from different areas of the company, such as development and construction, distribution etc. Furthermore, the relevant information can be obtained from the data and be stored in a database. Access to the information is ensured through a web-based platform. Furthermore, in parallel to the sub objectives defined, the following research tasks are outlined:

  • Development of a cross departemental concept for a controlled generation, processing and disitribution of product information throught an neterprise.
  • Analysis of the implications of newly generated customer feedback on the enterprise processes and IT-systems
  • Development of functional interdisciplinary classification scheme
  • Conception and implementation of an ontology-based intelligent knowledge management system and integration of this system into the existing system landscape
  • Development of a concept for an ontology-based information search.
  • Evaluation of the performance of the ontology-based information search in predefined use cases.

Figure 1 Knowledge Management System



      Daimler AG