Scrape public data from a public search in Orbis such as for [all titles with "New Haven" in the LC subject heading] (http://orbis.library.yale.edu/vwebv/search?searchArg1=new+haven&argType1=all&searchCode1=SKEY&combine2=and&searchArg2=&argType2=all&searchCode2=GKEY&combine3=and&searchArg3=&argType3=all&searchCode3=GKEY&year=2012-2013&fromYear=&toYear=&location=all&place=all&type=all&medium=all&language=all&recCount=50&searchType=2&page.search.search.button=Search) (where parameters can be manipulated in the URI to return different/more results) This would have the added benefit and detriment of returning dirty data. E.g. Several titles have a subject heading of Manuscripts, Medieval--Connecticut--New Haven. but obvs this is not about NHV so much as it is an indicator that the mss reside here.
Get data from PubMed API about research done in NHV, published here, etc. They have an API, which would make data acquisition smoother.