Showing posts with label Boolean searches. Show all posts
Showing posts with label Boolean searches. Show all posts

Monday, 3 March 2014

Boolean only makes sense: Boolean Logic

Boolean logic can best be described as a system of filtering results to both expand or narrow concept sand terminology. Based on three simple "filters"--AND, OR, or NOT--people can open up or narrow their search results. 

For our topic of research, and considering how recently the Rob Ford scandal took place, it was not difficult to turn up numerous results with simple searches for popular sources. Boolean was most helpful on the library databases when looking for research on satirical news reporting. To search this topic I used such terms or combinations as: 
Satire OR Humour AND News or Reporting 
Jon Stewart And Satirical AND News 
Boolean was successful in finding not only a vast amount of sources, but sources that were more related and relevant to my article because of the "filters" I applied. 

I recently used Boolean searching for an art history assignment. I was looking into indigenous representation in the early artworks of Emily Carr. My knowledge of indigenous culture, history and art gives me the background to know that the identifier for indigenous people has changed almost as much as Rachel's hair cut on Friends. In the past 30 years the terms Aboriginal, First Nations, Native, Native American, and Indian have been applied as their cultural name. To this day they label differs based even on.region. Currently in Canada we use the term indigenous, while in the States the term American Indian is in practice. 

This is all to say that when I search for an article, for efficiency I search something like: 
Indigenous OR native OR aboriginal AND Emily Carr AND representation OR identity

Boolean searches are essential when subject take on multiple names, reference systems or terminology. 

Sunday, 23 February 2014

The Search is On

Well, I know it's a little early for March 2, but Becky's here, all the same, reporting from the trenches of search terms and such.

As I think I mentioned in my last post, I already had a lot of my sources found and available to me. From what I didn't have, I kept a narrowed-down research question of "Where do the distinctions between pronoun cases come from and why do they exist?" constantly on hand for my remaining research.

I think the biggest problem, for me, with my topic is that it doesn't have any specific terms unique to it that don't apply to other, fairly unrelated, topics. This means that no matter how narrowly I search, I'm going to be wading through a lot that's not useful-- and, taking advice from the research advice in lecture, I wasn't prepared to wade through nineteen pages of search results in the hopes of finding something relevant on the twentieth.

Google Scholar was particularly not my friend in this respect. Google by its nature just casts so wide a net, that with the added difficulty of separating my topic from all the others that share key terms, it becomes impossible to find something useful. The one time I did find something that seemed relevant to my search, based on its description from Google Scholar, I opened up the article only to find after a page that it was completely tangential to my research. At least when searching through the library and its associated databases, I was able to narrow down my search to a couple of relevant fields, to help sift the dross from the gold.

What I found most useful from the research advice in lecture was the practice of picking up synonyms and connected terms as I went. Every time I found a source that, while relevant to my topic, didn't give me any new information on my research question, I'd plumb that source for new terms on the topic to reuse in my search: I went from "subject," "object," and "pronoun" to "pronoun cases" and then "nominative" and "objective" and finally "accusative," as I travelled ever deeper down the rabbit hole of pronoun-related terms.

These varying terms are, of course, great for combining in a Boolean search, but because of the limitations to Boolean searches on most databases, I found it easier to just search various terms in random combinations that seemed to work for me. What limitations? Well, a Boolean search is essentially a mathematical equation: You instruct the search engine to search for a specific combination of terms, with the ANDs and ORs in place where you want them. My problem is I want a full order of operations. I want to be able to say
(Pronoun AND Case) OR ((Subject OR Object) AND Pronoun) OR (((Objective OR Accusative) OR Nominative)  AND (Pronoun OR Case))
 Unfortunately for my obsessiveness (but probably fortunately for anyone trying and failing to parse that), very few Boolean searches allow for that level of nested brackets. Therefore, like I said, I mostly abandoned the Boolean approach for just trying the various combinations that my Boolean statement specified, one at a time. It worked. At least I'm good as a nested Boolean searcher, if I can't qualify as a human.