Is there a ready-made AI system (program) for a search robot for analyzing a database on jurisprudence and ranking the result?

Question

Artificial intelligence

Task: Display the narrowest information on the user's question (excerpts from regulatory legal acts) on the legal issue of interest to the user.
Given:
1. The system of existing normative legal acts (in fact, a database of certain terms and legal cases from life).
2. A user who does not know legal terminology (terms from the database) and who cannot correctly qualify even which branch of law (criminal, civil, administrative, etc.) his question belongs to.
Need:
a self-learning program (algorithm), which, based on the analysis of the entire set of normative legal acts uploaded to the database and the user's question (specifying words entered by him), will "direct" (narrow) the circle of normative acts and their norms, eventually leading the user to the final result on his own - a set of excerpts from normative legal acts (articles, paragraphs);
or you need a highly specialized search robot that will search for keywords in the legal database and rank the search results in a digestible way.
P.S. If not, then I assemble a team to possibly develop the theme. I have legal knowledge (the conditions for which words and questions will be cut off), you have the software part (I programmed a little in Delphi in my youth, but it was a long time ago, but this will allow us to speak almost the same language).

Answer 1

Yes, brains! )))

Answer 2

Sorry. What do you propose to leave lawyers without work? How did cars leave coachmen and grooms without work? And who, then, will protect "sort of copyrights" and patents for developments in the field of AI (expert systems, big data, etc.) and prohibit others from using developments in this area?

You need to look around kodeks.ru , garant.ru , consultant.ru , although this does not seem to be the case.
Maybe here the Analysis of Relationships is a norm ... (or nearby) that the sensible is written.

Answer 3

As far as I understand, the following approaches may be useful for you:
1) Information retrieval.
Indexing data from your database and customizing the search engine according to the subject area (lists of synonyms, thesaurus, expanding search queries, etc.)
2) Approach based on machine learning.
Creation of specialized classifiers that could, with some degree of confidence, relate the user request to a particular topic (you can use any classifiers - naive Bayes, decision trees, SVM, etc.). In this case, you should carefully consider the set of features (these can be words) that are used as a parameter vector for the classifier.