Figure 1417

A schematic diagram of the major components and interlinking processes used in Anderson's (1983, 1993) ACT models. Reprinted by permission of the author.

Learning from problem-solving attempts and instruction: Proceduralisation

Anderson (1982, 1983, 1987a, 1990, 1993, 1996; Anderson & Lebiere, 1998) has proposed another mechanism called knowledge compilation in his theory of skill learning within his ACT cognitive architecture (Adaptive Control of Thought). Among other phenomena, his ACT models— successively named ACTE, ACT*, and ACT-R— have modelled the learning of geometry (Anderson, Greeno, Kline, & Neves, 1981), computer programming (Anderson & Reiser, 1985; Pirolli & Anderson, 1985), serial list learning (Anderson & Matessa, 1997; Anderson, Bothell, Lebiere, & Matessa, 1998), and computer text-editing (Singley & Anderson, 1989). The main components of the ACT architecture have remained fairly invariant over the years although representational and processing details have changed (see Figure 14.17):

• A declarative memory, that is a semantic network of interconnected concepts that have different activation strengths (see Chapters 1 and 9).

• A procedural memory, of production rules.

• A working memory that contains currently active information.

Declarative knowledge is represented as chunks, which are just schema-like structures encoding a small bundle of knowledge (see Figure 14.18). Declarative knowledge can be reported and is not tied to the situation in which it can be used (e.g., a memorised textbook procedure to apply a statistical test), whereas procedural knowledge often cannot be expressed, is applied automatically, and is specifically tuned to specific situations (e.g., the knowledge we use when adding numbers). Information can be stored and retrieved from declarative memory by a number of methods. Information in the production memory takes the form of production rules, which are executed when they match the contents of working memory. Production memory can also be applied to itself by application processes; new productions can be learned by examining existing productions. For the most part, Anderson explains skill acquisition as knowledge compilation; as a move from the use of declarative knowledge to procedural knowledge. Knowledge compilation has two sub-processes: proceduralisation and composition.

A network representation of a declarative chunk from Anderson's ACT architecture. This chunk represents a piece of knowledge encoding the fact that 5 and 6 add up to 11.

Stop Anxiety Attacks

Stop Anxiety Attacks

Here's How You Could End Anxiety and Panic Attacks For Good Prevent Anxiety in Your Golden Years Without Harmful Prescription Drugs. If You Give Me 15 minutes, I Will Show You a Breakthrough That Will Change The Way You Think About Anxiety and Panic Attacks Forever! If you are still suffering because your doctor can't help you, here's some great news...!

Get My Free Ebook


Post a comment