Using word vectors to enable better search in our legal products
Download as PPTX, PDF1 like209 views
The document discusses the relevance of artificial intelligence (AI) for lawyers, emphasizing its potential to enhance quality and efficiency in case processing and to introduce new consulting products. AI technologies like word2vec are highlighted for their ability to improve search capabilities and user experience, demonstrating significant user interest in continuing to use these tools. The conclusion calls for ongoing exploration and collaboration in AI to ensure value creation in legal services.
Using word vectors to enable better search in our legal products
- 1. 1.0 Relevant legal information in context [email protected] AI at Haufe Lawyers
- 2. • range of new services • value creation process The use of the opportunities and the adaptation of the technology could be critical for survival in the market. How relevant is Artificial Intelligence for us?
- 3. opportunities • quality and efficiency of case processing can be significantly increased • new consulting products can be offered Artificial Intelligence For Lawyers
- 4. OBJECTIVES technologie word2vec use case selection search Our Current project together with TUm
- 5. Searcher intent Query context Conceptual match SEMANTIC Search
- 6. • vector representations of words • unsupervised learning • sematic relationships (k- m+w=q) • cosine similarity for comaprison word2vec
- 7. RESULTS
- 8. 15 people so far Attendees Experience probands
- 9. 15 people so far Attendees Behavior
- 10. SUS
- 11. WORD2VEC SENT provides the best results participants confirm the quality of the search results quality measured with the ground truth
- 12. • selection search is good as starting point • but users also want the ability to control the search results • 90% have said that they would continue to use the selection search in the future. Ovservations
- 13. best search results for his question what is not found is not there Value FOR OUR Customers
- 14. In exploration stage TUM collaboration accelerated the entry We should stay on it. conclusion
- 15. Questions? Thank you very much for your attention. Hans.Lecker @haufe-lexware.com
Editor's Notes
- #3: range of new service = improved search for expert knowledge, case analyse tools, decision support tools value createn process = index of keywords, content creation (Kommentare, Schmerzengeldtabelle), content analyse
- #4: quality and efficiency … - https://www.lawgeex.com/resources/aivslawyer/, contract review new consulting products can be offered – highly automated small cases (Legal Tech)
- #5: Live show prototype WordAddIn Haufe-Recht
- #11: SUS Score
- #14: Welches Fachgebiet bis Du unterwegs?
- #15: Exploration stage – understanding the technology and it‘s framework conditions is necessary TUM collaboration - The cooperation with the university has considerably accelerated the entry into ML.