NER task using dandelion with python

[Massimo Guarino]

Good evening,

I'm working on a project and I've to extract named entities and their uris and labels (person, work, org, places, etc...). I'm also interested in tuning all the parameters

I found in the demo version (top_entitites, more tags/more precision). The problem is that I should do all the previous tasks in Python (I've already installed dandelion_eu). 

While I've already red your api documentation, I would be very glad if you can help me to handle all this stuffs

in Python! (Is there an online tutorial I can follow in order to achieve a better understandigs of Dandelion in python, e.g. to extract categories as person, work, etc...?)

Many thanks in advance for all your assistance,

Best Regards,

Massimo

[Giacomo Berardi]

Hi Massimo,

the Python client allows to use any parameter documented in https://dandelion.eu/docs/api/datatxt/nex/v1/

The method for Entity Extraction in Python is documented in https://python-dandelion-eu.readthedocs.io/en/latest/datatxt.html#nex-named-entity-extraction

Following the first example, you could perform a call like:

datatxt.nex(text='Oh my, arduino is super cool, so #opensource', top_entities=10)

regards

Giacomo Berardi

Dandelion team

[Massimo Guarino]

Dear Giacomo,

I appreciated your tips,

really helpful,

many thanks,

Massimo