the Design Experience Weblog Archive

My purpose for having students take ownership of the class blog, work through the Segue CMS site to solidify them in the f2f class world, and then re-create themselves in their blogs is simple: I want the students to remove me from the class/course. I want to be invisible. I want them to realize that, through ownership, students can participate in the world using the best available tools. I want them to command the technology--not be neutral to it. Hector J. Vila

I love the idea of the instructor becoming invisible. That is the key here. Unlocking the students potential to learn. Too often online learning really means "online teaching". Until someone can own the learning process and realize they need to take information and knowledge and make it their own, learing doesn't really take place.

The idea that students can take technology and control it to direct their own learing is quite related to unschooling, which means many things, but most of all means giving someone the freedom and power to learn.

I have a lot more to think about, but I believe that any online learning attempt that does not give the potential learner the power to control and own the learning process will not be effective.

tsearch2 works reasonably well and is configurable with different dictionaries that you can build yourself based on ispell word lists and snowball stemmers. It has a parser built into the database and so far has made it easy to integrate full text search into various SQL queries.

One additional way folks seem to like to organize content is categorization. Generally this takes the form of a dropdown or multiple select widget when you can pick from a predefined flat list, or hierachy of categories. An additional kind of categorization based on free text "tag's has been recently popularized by sites suck as de.icio.us and flicr. I think taking both of these ideas, and full text indexing the category names and synonyms, and full text indexing the free form tags would make a neat combination in addition to full text indexing of the "content" of a particular page or item.

The art would be in weighting the category names, tags, and content to return the most relevant results. Of course, it depends on who is doing the categorization. I think it would be a fun experiement to try out categorization vs. free form tags vs. full text indexing on the OpenACS discussion forums database.

One advanced idea might be to extract the most common tags from the forum posts. That is, once a sutiable library of the most popular tags is established, it might be a good idea to apply that list to the text of a forum posting based on the words in the posting. A domian specific idea would be to prepare a list of OpenACS package names and tcl procedures and to tag those as well. This way, a visitor to openacs.org could go to the online API documentation and find a tcl procedure or package name and then pull up the list of forums postings that refer to those packages or procedures.

All this is possible if I ever have a free minute to play with these ideas.