Logo des Repositoriums

Auflistung nach Autor:in "Thimm, Heiko"

1 - 5 von 5
  • Konferenzbeitrag
    Assessment Power of ChatGPT in the Context of Environmental Compliance Management – Experiments with a Real-World Regulation Cadastre
    (EnviroInfo 2023, 2023) Thimm, Heiko
    In multiple research disciplines use cases built on Large Language Models in particular ChatGPT are at the centre of today’s discussions. For example, in various ongoing projects of the LegalTech area ChatGPT is evaluated in terms of its potential to replace routine work of lawyers. In a recently started project we are investigating the use of ChatGPT for a specific corporate compliance management task. In particular, based on a real-world test data set ChatGPT is prompted to assess the relevance of environmental regulations. The ChatGPT output is compared to the respective judgements of the human experts in order to obtain a first indication of the assessment power of ChatGPT in the compliance management domain. This research in progress article gives an overview of the evaluation approach and presents first results of a set of 142 test cases covering regulations from four different areas of environmental legislation.
  • Textdokument
    Global Environmental Website Disclosure of Large Production Companies and their CDP Scores
    (INFORMATIK 2021, 2021) Thimm, Heiko; Boye Rasmussen, Karsten; Tanik,Milan
    This work contributes to the exposure of the current state of environmental disclosure through an empirical transnational study of 344 websites of large production companies from China, Europe, Japan, and the USA. The focus of the investigation is on four formative indices based on variables covering observation of disclosure extent of environmental management and environmental compliance management. All countries show low levels of disclosure. However, some variation on the extent of disclosure was found. The lowest level of disclosure was found for the Chinese websites. Higher - though not remarkable higher - levels of disclosure extents were observed for the companies from Europe, Japan, and the USA. Also, the correlation between disclosure scores based on these observational data and the CDP carbon disclosure scores of the year 2020 did only reveal insignificant correlations. However, websites having webpages containing a higher number of terms of the corporate environmental management terminology also achieve higher CDP scores.
  • Konferenzbeitrag
    ICT Support of Environmental Compliance - Approaches and Future Perspectives
    (Advances and New Trends in Environmental Informatics - Stability, Continuity, Innovation, 2016) Thimm, Heiko
  • Konferenzbeitrag
    A monitoring framework for guidance and risk control assistance of environmental compliance officers
    (INFORMATIK 2015, 2015) Thimm, Heiko
    It is the ultimate goal of environmental compliance management to assure corporate compliance with given regulations. The work processes that are required to fulfill and maintain compliance are usually complex and long running processes that are composed of many interdependent human lead activities. Errors of human work activities and also material and equipment defects can lead to non-compliance. A monitoring framework is proposed that based on data entries of compliance officers can track compliance work processes and provide compliance officers with activity guidance. Main activity failures can be detected through failure detection functions that analyze the activity log, failure indicators, and also external data. Based on the proposed framework environmental information systems can be extended towards actively responding systems that guide compliance officers and provide them with assistance for risk control tasks.
  • Konferenzbeitrag
    Relation Extraction from Environmental Law Text Using Natural Language Understanding
    (EnviroInfo 2022, 2022) Thimm, Heiko; Schneider, Phil
    In the last decades the highly active area of environmental legislation has produced a vast amount of text documents that contain laws and regulations enacted by various types of rule setters. This large body of legal text documents is still growing with an increasing speed. In order to assure compliance with the regulations, today, corporate specialist spend a lot of time with the reviewing and assessment of these documents. It seems that through the use of text processing assistance tools these important corporate environmental compliance management tasks can be completed in less time. To develop corresponding assistance tools has been the broader goal of this work in which initial text processing experiments with a common Natural Language Understanding pipeline are described. The obtained results confirm that in order to extract meaningful relations from text documents of the environmental legislation area, domain-specific processing techniques that are tailored to the specific language and format of legal text are required.