Fibonacci Heaps in AI NLP QA Model - Suggestions?

[heff001]

TL;DR

Fibonacci (min) heaps

I am using them in an AI NLP Question Answering Model - Root to Frontier Hierarchical Trees. Is this too academic? I have tried Red Black Trees with little success. What do you suggest?

 

[pbuk]

I am using them in an AI NLP Question Answering Model - Root to Frontier Hierarchical Trees.

Is this a practical implementation? You might be better off asking in the Programming and Computer Science topic.

Fibonacci (min) heaps... Is this too academic?

Not for me :) but if I were you I would start prototyping your model using a high level language (e.g. Python) and a pre-built tree package (e.g. anytree). Once you have a working Proof of Concept you can start looking for bottlenecks and only then if you need to consider rolling your own low-level tree handler.

 

[heff001]

I thank you...

This for an NLP/NLU Startup

The Fibonacci heap of (min) trees so far has been #1 on my list. Adjacency trees are too simple for NLP/NLU. I need to build an Answer Model from a Question model.
Here, I can build hierarchical trees within a heap, build strings from trees, for the NLU response.
The major up-front work is the NL classifications. I just want a flexible data structure that I can adjust and eventually populate in Hadoop DB.

I do Python.

I am not at a point of using recursive features yet.