Martinez-Gil, Jorge A Novel Neurofuzzy Model for the Comparison of Legal Texts., 2022 [Preprint]
Preview |
Text
legal_neurofuzzy.pdf Download (179kB) | Preview |
English abstract
The daily work of legal professionals is often hampered by characteristics such as the high speed with which new legislation is generated. In addition, the generation of such legislation is almost always done using unstructured formats that are not prepared for automatic processing by computers. As a result, a large amount of heterogeneous information is generated in a highly chaotic manner, leading to an information overload. We have designed a new model for comparing legal texts that combine the latest advances in language processing through neural architectures with classical fuzzy logic techniques to overcome this problem partially. In this regard, we have evaluated such a model with the lawSentence200 benchmark dataset, and the first results we have obtained seem promising.
Item type: | Preprint |
---|---|
Keywords: | Neurofuzzy systems, Legal Intelligence, Semantic Textual Similarity |
Subjects: | I. Information treatment for information services > ID. Knowledge representation. I. Information treatment for information services > IF. Information transfer: protocols, formats, techniques. |
Depositing user: | Dr Jorge Martinez-Gil |
Date deposited: | 25 Aug 2022 08:37 |
Last modified: | 25 Aug 2022 08:37 |
URI: | http://hdl.handle.net/10760/43487 |
References
Downloads
Downloads per month over past year
Actions (login required)
View Item |