1. what did you find (articulate the answers in your own words)

1.1. interesting about the article?

I found the section explaining the three different critiquing mechanisms the most interesting. It outlines design patterns for critic rules. When a specific condition is met, the action associated with that critic rule is performed. This is the real heart of the critic technique, and the design must be broken down into these mechanisms in order to provide the criticism based on the domain sensitive constraints of a problem.

Also, the supporting three-tiered problem approach gives the user all the information about a problem they will need. They are first notified, and a hypermedia link is available to further explain reasons for the problem, and specific explanations of when and how a problem surfaces. By notifying the user in the design phase the system is not only teaching the user domain specific knowledge, but is also paving the way for a smooth deployment of the finished design.

1.2. not interesting about the article?

A lot of the information past section 5.2 was too abstract for my liking.

2. what do you consider the main message of the article?

The main message of the article was that embedded critiquing systems can be applied to complex, iterative design processes, and that using such a system can provide a way make sure a design meets specific requirements. The system critiques the user by making suggestions about their design and the user can make decisions about which critiques to follow. One other message was that a better design can be realized by embedding the critiquing operations into the design environment itself. In this fashion, the user can see suggestions or problems with their design as it is manifested.

3. are themes discussed in the article which you would like to know more about?

How is a problem broken down into critic rules? Certainly domain-specific knowledge is required. How can critic rules be managed so as to not overwhelm the system? I would imagine that given a complex enough problem domain, this method of embedding the critiques in the design breaks down as the system is not able to respond to input fast enough.

4. do you know of other papers, ideas, and systems which are closely related to

4.1. DODEs

I cannot think of any. There are plenty of computer systems which support designers (CAD programs, 3d modelling programs, etc), but I don't know of any which use a knowledge-based approach to provide assistance to their users.

4.2. Critiquing?

Spelling and grammar checkers are the obvious ones. In some sense, educational games, which correct the user after making a mistake and show why the answer is wrong, could be considered critiquing systems. Compiler warnings for code which is only incorrect under certain circumstances is another example, however it is limited to being a generic critic.

4.3. analyze “spelling correctors” as a critiquing system

Spelling correctors can only provide generic criticisms. A grammar checker might go into the specific or interpretive realms, but a spelling checker is just checking a single word against a dictionary of words. The base rule it works from is whether or not a word exists in its dictionary. If the word is found, no criticism is necessary. Otherwise, the dictionary is searched for other words that may have a similar spelling. This list of words is presented to the user as concrete examples.

This system works fairly well in general, but is far from suited to be automatic. The system does not know the correct spelling when it encounters an error which has more than one similar spelling. Or the user may have used a proper spelling, but for a word that had a different meaning or context. It is thus very likely for a spelling checker to present information which is mistakenly believed to exist in the design environment.

5. what does the article say about

5.1. design

The important aspects of design mentioned in the article are accounted for by providing the user with an interactive design area. Relevant design rules are presented to the user during the design process.

5.2. learning

The user is learning while designing because they are using a interactively critical system. If the system raises a cristicism the user doesn't understand, they can bring up the hypermedia page representing the reason behind the criticism. A set of arguments for/against a certain design decision are presented to the user, who can then make a more informed decision, taking into account facets of their design they may not have been aware of. The next time they make a design, they will have learned some background on key design decisions. In some sense the user is learning about design from the critic rule writer.

5.3. collaboration

The collaboration taking place is between the design environment and the user, between the design environment and the person writing the critic rules, and between multiple users of the system. The user and the critic rule writer are indirectly collaborating through the system. Multiple users are collaborating with each other because previous designs are accessible through the system, allowing newcomers to examine the design decisions previous users made.

5.4. innovative media support for these activities?

The hypermedia is innovative in that it provides the context of a criticism when the user does not understand on their own. It is the major force behind learning in the system, since inherent in the critic rules are some base guidlines found in any design of that type.

6. do you have any ideas how this research could / should be extended based on your own knowledge and experience?

How are the critic rules implemented in the Hydra-Kitchen system? Can there be a generic way to describe critic rules as they can be applied to any design environment? Are there any facilities for sharing a set of specific critic rules?