New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Improve NLU for more detailed queries #2
Comments
Rasa integration? |
Hello @Immortalin, I do not plan to have a Rasa integration yet. But it does not mean I will not in the future. |
It will make slot filling etc. easier as you can just delegate everything |
In any case if you do decide to use Rasa, please make it modular, it seems that Rasa does not have any serious support for non-trivial morphology. |
Snips-NLU perhaps? |
@Immortalin looks like snips doesn't support languages with substantial morphology and morphophonology either. |
Quick update for this topic: I'm working under the nlu-improvement branch. Leon is able to find several entities from a query thanks to node-nlp. The Is It Down module is now using these entities instead of a custom regex to find the domain names (related to #51). I'll be working on the Trends package now to validate the feature, but I'll also need to dig into other topics regarding that improvement (such as locations and slot filling). I'm not sure that will be in the |
I finished this improvement. It'll be released in the |
The docs according to those changes has been published. Most of the essentials information can be found in this part. |
@louistiti , the "this part" link in the previous post resolves to a 404 here. Could you fix this please? Thank you! |
Hi, sure, here you are: https://docs.getleon.ai/packages-modules#entities. |
Be able to subtract a type of entity from a query. Such as date, number, location, etc.
If it uses a tiers but Node.js, it should operate from the core and send the result object to modules via the child process parameters.
Otherwise, it should fully be implemented in the core itself.
I noted some following leads in this Trello card.
The text was updated successfully, but these errors were encountered: