Concise-concepts

- Resolve several breaking bugs.
- Perform a large optimization regarding string normalization and KeyedVector membership.
- Perform simple code optimization each localized to just a few lines.

17 resolved an issue that caused duplicate logging for the same missing keys. Also introduced `verbose` option.
18 introduced `json_path` to export matching patterns to a custom path.
19 defaults to key if example words are not present in the embedding model.

Added correct character escaping Regex lowercase match
Added correct n-gram join.

11 added support for more custom matching patterns via 4 config variables.
- ´exclude_pos´: A list of POS tags to be excluded from the rule based match.
- ´exclude_dep´: A list of dependencies to be excluded from the rule based match.
- ´include_compound_words´: If True, it will include compound words in the entity. For example, if the entity is "New York", it will also include "New York City" as an entity.
- ´case_sensitive´: Whether to match the case of the words in the text.

10 resolved an issue where gensim Word2Vec and FastText models were not processed as KeyedVectors. Hence, the model did not load due to mis'interpretting it as an iterable object.

**Also unified code regarding checking whether string are present in a model.**

**Also made sure that n-grams models and word matches are supported.**