“A tag is a (relevant) keyword or term associated with or assigned to a piece of information (a picture, a geographic map, a blog entry, a video clip etc.), thus describing the item and enabling keyword-based classification and search of information.” [Wikipedia]
Tagging is the process of assigning keywords or phrases to items. To be more concrete, many of us may have collections of bookmarks in our web browser. Tagging each bookmark with a relevant term allows them to be classified and categorized semantically, or by meaning.
But one major downside of tagging is that the user has to actually do the tagging; quite a bit of work if you have many thousands of bookmarks (or emails or photographs). Rashmi Sinha was one of the first people to start thinking about the cognitive requirements (i.e., what happens in the head) when users tag. Here is a figure from her analysis:
Her bottom line is that tagging is efficient (compared to other methods of organization) because when we are trying to think of keywords, we have lots of choices that come to mind (stage 1). However, I see this as a potential downside. It’s a heavy decision step that must be repeated for every item one needs to tag.
Several new products are on the horizon that aim to automate this very step:
The first (which is in private beta and thus unavailable) is Twine. The New York Times wrote an interesting story about Twine and how it automatically scans your documents to obtain relevant keywords.
Sarah Miller, a librarian at Illinois Wesleyan University in Bloomington, became a member of Twine’s test group in November, partly because she and her husband, Ethan, a doctoral candidate, needed a place to organize all the documents they wanted to share with each other about teaching and learning.
Ms. Miller likes Twine’s mechanized tagging abilities.
“If I save the URL of a Web page into my Twine account,” she said, “Twine will skim the page and turn it into tags automatically. It’s a way to tie together things that my husband and I find over days, and months and years.”
Twine has an option that allows people to do their own descriptive tagging, just as they might, for instance, use the Web service del.icio.us to assign labels to Web sites to help keep track of them.
“But my tagging is inefficient,” Ms. Miller said. “Personal vocabulary changes. It’s difficult to be consistent.”
A less automated solution is a new product (also in private beta) called zigtag. Zigtag relies on you entering a keyword, but afterwards will suggest additional keywords.
After entering an appropriate tag for a page, the user is presented with a list of matching keywords, each of which has been defined in Zigtag’s database. For example, after entering “Apple” into the search field, I was able to choose from “the computer company”, “the pomaceous fruit”, and “the record company”, among others. The process is painless and the integrated dictionary is fairly comprehensive. If you happen to stumble across a term that isn’t defined, you can easily request to have it added to the dictionary (and can place your own temporary tag). [Tech Crunch]
While these are nice solutions, i’ve always imagined that one side benefit of tagging was that the very effortful process of tagging could contribute to a more durable memory trace (the classic “generation effect“). Incidentally, some limited research of mine (PDF) has not borne this out. But in reality, how well do we want to really remember our bookmarks? Most of us are satisfied that it is stored somewhere and are less interested in retrieving it later unaided.