Legal NLP releases Legal Summarization, 15 new NDA NER models, 5 Legal Clasification models, and Visual Legal NLP demos
Legal NLP 1.11 comes with a lot of new capabilities added to the 800+ models and 125+ Language Models already available in previous versions of the library. Let’s take a look at each of them!
Native Legal Text Summarization: new annotator and 2 models
Legal texts are usually long and very verbose, and mining the information in them is a task that may be crucial to take decisions.
By using our new Legal Summarizer() module, you can get state-of-the-art, short versions of your legal documents, without losing any information.
We included 2 models for Legal Summarization:
- The base model, with generic capacities for summarizing legal documents;
- A specifically finetuned model trained to summarize Legal Agreements . For this task, we finetuned our base model with more than 8K sections from different legal commercial agreements.
15 new Legal Named Entity Recognition on NDAs (50 entities)
Named Entity Recognition (NER) models were include to extract information from 15 different clauses. Here is a summary of the clauses and the more than 50 NER entities pretrained.
This is an example of some NER on NDA agreements:
4 French and Portuguese Legal Multilabel and Multiclass Text Classifiers
We add to our more than 810 classifiers, these recently trained ones:
- Portuguese Unanimity prediction of judgement summaries;
- Portuguese appeal approved / denied / favourable classifier;
- French and English COVID-19-related multilabel legal classifiers, predicting if a specific measure restricts fundamental rights or daily liberties, talks about mobilization of police or armies, state of emergency, suspension of agreements, etc.
Example:
“Par dérogation à l’alinéa 1er, sont autorisés :- les cérémonies funéraires, mais uniquement en présence de 15 personnes maximum, et sans possibilité d’exposition du corps ;”
- Restrictions_of_fundamental_rights_and_civil_liberties
- Restrictions_of_daily_liberties
New Visual NLP + Legal NLP demos
Legal may contain elements as signatures or handwritten sections which can complicate their automatic processing with NLP.
Fortunately, by Using Visual NLP and Legal NLP, we can detect those as shown in the following new demos :
In addition, new Visual NLP Question Answering models on legal texts have been published:
➤Q1: At what time 25 June 2012 must be submitted to be sworn in? 2:30 o’clock
➤Q2: To whom was this subpoena sent? Mr. Vicente Yagpes
➤Q3: What is the NPS DOCKET NO of this document? vii-11–12e
Fancy trying?
We’ve got 30-days free licenses for you with technical support from our legal team of technical and SME. This trial includes complete access to more than 800 models, including Classification, NER, Relation Extraction, Similarity Search, Summarization, Sentiment Analysis, Question Answering, etc. and 120+ legal language models.
Just go to https://www.johnsnowlabs.com/install/ and follow the instructions!
How to run
Legal NLP is very easy to run on both clusters and driver-only environments using johnsnowlabs
library:
!pip install johnsnowlabs
nlp.install(force_browser=True)
nlp.start()