Summary of Gerps-compare: Comparing Ner Methods For Legal Norm Analysis, by Sarah T. Bachinger et al.
GerPS-Compare: Comparing NER methods for legal norm analysis
by Sarah T. Bachinger, Christoph Unger, Robin Erd, Leila Feddoul, Clara Lachenmaier, Sina Zarrieß, Birgitta König-Ries
First submitted to arxiv on: 3 Dec 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper applies Named Entity Recognition (NER) to a specific sub-genre of legal texts in German, focusing on administrative processes. The goal is to identify stretches of text that correspond to one of ten classes defined by public service administration professionals. Three methods are compared: Rule-based systems, deep discriminative models, and deep generative models. Surprisingly, deep discriminative models outperform the other two, with rule-based systems and deep generative models performing similarly but less well than the winner. The reason for this difference is likely due to the heterogeneous nature of the classes used in this task, which deep discriminative models are better equipped to handle. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper uses computer technology to help identify important information in a type of German legal text. These texts deal with rules and procedures for administrative tasks. The researchers tested three different ways to do this: using simple rules, using powerful artificial intelligence models, and generating new text based on existing patterns. They found that one type of AI model worked best, which is good because it can handle situations where the information being looked at has different meanings or structures. |
Keywords
» Artificial intelligence » Named entity recognition » Ner
