Update list of stopwords

[nathanbegbie]

Stopwords were made to include countries, which was actually introduced when we attempted to do the clustering. We should not remove them in the stopwords approach, to give more context.

[tomwilsonsco]

Testing revealed country names were also not in the list of 'English words', so needed to add a list of the countries here to ensure they make it through pre-processing.
Next consideration Nathan raised was the countries with spaces in their names "Democratic Republic of the Congo" these will still not survive pre-processing intact and our TDM stores unigrams, not bigrams etc at the moment.

Short-term solution we can split the multi-word countries into separate words in the input so their key words are retained e.g. search would pick up word "Congo".