Add self-reference field to ISI table.

Working on this now. It actually seems pretty nicely done, unlike some formats I know. There are a couple weird things though.

Some papers claim one name for the journal they are published in, but are ALWAYS cited using another name, for instance I have a paper that claims it was published in "EARTH PLANET SCI LETT" but absolutely everything cites it as "HILGEN FJ, 1991, EARTH PLANET SC LETT, V107, P349". It seems that each token in the title of the citation is always shorter than or equal to the length of the title in the original paper, so this should be semi-easy to work with.

Another interesting one is a paper always cited as "DAWES C, 2004, ARCH ORAL BIOL, V49, P665, DOI

10.1016/j.archoralbio.2004.02.007", which does not contain any DOI information in the paper itself. How they know the DOI is 10.1016/j.archoralbio.2004.02.007 is beyond me. This makes it impossible to create this a self-reference using the given data. We can however "solve" this one by simply chopping the DOI off of any citations we see, which forces them to conform to the self-reference we give the original paper.

Current Status:
Completed implementation.
Currently we normalize the journal names with an heuristic which attempts to match the journal name a paper claims it belongs to with a journal name that other papers cite. In almost all cases it makes a correct match, but since the journal names appear to be human-entered there is no guarantee of a correct match.

Future Work
* Add "all citations" column, which combines the self-citation with the normal citations (used for applying co-authorship style algorithm to citations) (?)

Task mentioned in "Future Work" above is no longer needed.