Streer et al
by Carson Reynolds
I had thought this would be an article discussing adpated social interaction versus direct manipulation techniques. It turns out that this is actually just the abstract for a panel which took place at CHI ’95 which Streeter, Nass, Shneiderman, and Keisler organized.
As penance for a bad article choice, I did a brief literature search to find some other documents that address this debate. I read:
The introduction to Concept and realization of intelligent user support in interactive graphics applications, the Dissertation of L. Miguel Encarnação. This seeks to find a definition for intelligent user interfaces, but also includes a reference to the next related reading:
Another panel was held a t CHI ’97 to debate interaction with Software Agents vs. Direct Manipulation. Thankfully, this panel included a document detailing the positions. The crux of the debate seems to be between the proven attributes of direct manipulation: “predictable and controllable” vs. the potential benefits fo “autonomous, adaptive, intelligent” systems. Schneiderman seems skeptical that the anecdotal evidence from a few adaptive help systems is enough to warrant a more full scale exploration of less predictable interfaces. As an aside I think that it is interesting that Pattie Maes seems to conflate intelligent with adaptive. It seems right now that “intelligent” systems are grouped according to a vague attribute, while adaptation can be more rigorously defined.