Decide on the article type inclusion/exclusion and grouping

[krassowski]

As noted by @biswapriyamisra, we should treat review differently than original research articles.

My early idea was to group 'Review', 'Editorial', 'Comment', 'Letter' together as 'Descriptive articles' (although they are not of equal importance). We may want to disregard the article types below a certain threshold like 5 or 10.

Currently we have:

Article type	Count
Journal Article	3126
Research Support, Non-U.S. Gov't	1275
Review	696
Research Support, N.I.H., Extramural	435
Research Support, U.S. Gov't, Non-P.H.S.	152
Editorial	43
Comparative Study	36
Comment	35
Clinical Trial	26
Multicenter Study	21
Published Erratum	21
Research Support, N.I.H., Intramural	13
Letter	12
Evaluation Study	12
Dataset	9
Case Reports	7
Twin Study	7
Observational Study	7
Video-Audio Media	6
Introductory Journal Article	6
Systematic Review	6
Validation Study	5
Research Support, U.S. Gov't, P.H.S.	5
Randomized Controlled Trial	5
English Abstract	4
Meta-Analysis	4
Congress	3
Interview	3
News	2
Practice Guideline	1
Controlled Clinical Trial	1
Consensus Development Conference, NIH	1
Historical Article	1
Clinical Trial, Phase II	1