I use Spacy to extract named entities from the texts. But it extracts a lot of entities represented by different words, but actually they relate to the same real-world entity. For example, President of the USA and POTUS, Great Britain and the UK, International Monetary Fund and IMF, etc. In future, I need to process entities and it is good if I will be able to stick them together. So, in the result, I need to have just a single word (or phrase) that represent a single real-world entity. How this can be done?