Here’s my time vampire of the day:
I’ve been using SVN until now, but git has intrigued. Flashbake’s ability to track context as well as product seems really interesting. More to follow.
Update: 2009-03-05 Fail. Just couldn’t get it to work. Oh well. I blame my lack of terminal-fu. Interesting project, though.
One of the things that I’ve been working on over the past year is ways to help disciplinary novices create better search terms and search strategies. Research has shown that domain novices are particularly poor at identifying appropriate search terms — largely because they are not yet familiar with the terminology and relationships used within the discipline (Hsieh-Yee, 1993; Vakkari, 2001). As a result I’ve been very interested in different ways of analyzing strengths and weaknesses of search strategies in order to better advise and support students at the beginning of their research. Max L. Wilson and Daniel Tunkelang had an interesting little back-and-forth about ambiguity of terms and clarification vs. refinement. Hopefully they’ll continue this discussion as it is yielding many interesting ideas.
This could prove to be useful in a teaching context. Even though the article that Wilson cites shows that only 16% of search terms were ‘ambiguous,’ one has to wonder if these terms would have been ambiguous had the searcher used a discipline-specific search engine or database rather than a general search engine. Also one has to wonder if there is a difference between the percentage of ambiguous terms found in queries resulting from problems with low complexity (ie. when was “Rubber Soul” released) to queries with high complexity (ie. what was the impact of the trans-atlantic shipping on youth culture in 1950s Liverpool?). Could directing students towards discipline specific sources help with the ambiguity problem? Possibly. Discipline specific sources have the advantage of rewarding correct term selection with high-relevance material (if “relevance” is defined as “from the right disciplinary context”) however they are also very punishing — zero results, for example — for those who do not have the right search terms yet. Developing strategies for the latter group has been the biggest challenge when teaching information seeking skills.
Obviously I need to spend more time thinking about this.
I’m a couple of chapters into Christine Bruce’s Informed Learning and I already feel the need to talk about it. The focus on information seeking and use being a disciplinary act is wonderful. This quote pretty much sums up what makes this book so wonderful:
We need to emphasize both discipline and information-use outcomes in our learning design and implementation; discipline mastery is achieved through the processes of creative and reflective information use. Once we recognize what information is and how we are using it, we can be more in charge of the information environment and how we encounter, source, control, engage with, and use information. We cannot assume that learners are aware of these processes or that they are able to implement them. The learning experience that prepares today’s students for tomorrow’s professional practice brings such practices into the curriculum and encourages reflection upon them … The idea of informed learning comes from recognizing that information use and learning are close companions; in formal learning environments, discipline content and effective information use need to be learned together as interrelated phenomena. (p. 3-4)
I’ve been struggling to articulate what she communicates with such ease. I’m going to reserve further comment until I’m finished the book but I wanted to get something up here as it looks like much of my work over the coming year is going to focus on the relationship between discipline, information seeking, writing and information literacy. It is exciting to know that the little nagging idea that has been rolling around the back of my head has some basis in reality.
No posts lately because I’ve been stuck in a rathole while trying to work through a possibly insane idea that I’ve been toying with. Hopefully it will lead to a fuller post, but in short I’m messing around with ways to use citation information to help people clarify their dreaded “I’m looking for information on the impact of x on y” (where x usually equals globalization and y equals either a non-Enlgish speaking, non G-7 country or women.)
Who knows how it will work out. Results have been limited so far, but I have had some success in taking topical keywords and terms from statistical agencies and producing a list of articles that identify and analyze trends by working with hard data. Unfortunately, at the moment the number of databases that this works with is exactly one (Scopus) and the list of terms from statistical agencies required to get reasonable results is approaching 20. This may be more useful in theory than in practice.
Having said that, it really would be nice if there were more databases and search engines out there where one could specify that the search terms should be found in the references list. Also it would be nice if Scopus wasn’t so stingy in their application of proximity searching. Why, for the love of the Flying Spaghetti Monster (yay! two year old meme!) can’t I search for:
(census OR statistics) w/10 Canada
The answer is because Scopus doesn’t allow the use of OR when proximity searching. Which makes kittens cry.
This one needs a lot more thought, but is there a way to quantify the difference in technique/cognitive approaches between effective searching of structured databases and unstructured databases.
I have a bucketload of strategies that I use for navigating through structured relational databases. I’m also aware of the fact that I approach searches in a search engine like Google very differently than I approach something like the MLA international Bibliography or PSYCinfo. Part of the difference could be that I’m searching with the expectation of finding different kinds of information, another could be that I adjust my strategy in Google knowing that I will be matching against at least 100K of full text rather than abstracts and bibliographic data.
I imagine that people consciously or subconsciously structure their searchers to avoid unwanted material (even if they don’t use ‘not’ or ‘-’). I’d be very curious to know how people take a massive set of potential information like Google structure their searches to compensate for the monumental amount of off-topic material that could possibly interfere with the quality of their search results.
Spent last night reading a few articles on the relationship between searching and serendipity. This morning’s indexed seems to fit the bill perfectly 