Perform Natural Language parsing using statistical approaches

Currently the Natural Language parsing feature is done through regular expression matching based on a finite set of common language patters. This approach has several limitations:

- it's limited to the set of language patterns we defined
- it's limited to the languages for which we have a parser
- it's not very flexible has the user has to perfectly follow our patters structure if he wants to be understood

I believe that by using statistical approaches we could heavily improve this feature, overcoming the two mentioned limitations. Moreover, if we use the proper tools we could be able to extend the same feature to more language with little or no effort.

The downside of this approach would be that we would be forced to use a language different from Vala as there is no proper library or tool which would let us implement it using the elementary OS standard programming language.

There exist several NLP libraries, probably the best one for our case is SpaCy: [https://spacy.io](https://spacy.io)

---
Want to back this issue? **[Post a bounty on it!](https://www.bountysource.com/issues/62038897-perform-natural-language-parsing-using-statistical-approaches?utm_campaign=plugin&utm_content=tracker%2F60388692&utm_medium=issues&utm_source=github)** We accept bounties via [Bountysource](https://www.bountysource.com/?utm_campaign=plugin&utm_content=tracker%2F60388692&utm_medium=issues&utm_source=github).