Goals and Scope
The goal of this package is to provide basic functions and types for building models based on the ACT-R cognitive architecture. There are several approaches for developing ACT-R models. The classic approach involves simulating the internal cognitive processes as a discrete event simulation. A proof-of-concept implementation in Julia can be found in the package ACTRSimulators.jl. Another approach is to derive a likelihood function to enable maximum likelihood or Bayesian parameter estimation. This approach is limited to mathematically tractible ACT-R models, typically of low to moderate complexity. The repo ACTRTutorials.jl provides a collection of tutorials for developing ACT-R models within the likelihood approach.
Installation
In the REPL, type ]
to enter the package model and enter the following:
add ACTRModels
Quick Example
The example below shows how to create a simple ACT-R model and retrieve a memory using (animal,rat)
as a retrieval request.
using ACTRModels
using Random
Random.seed!(28194)
# create chunks of declarative knowledge
chunks = [Chunk(;name=:Sigma, animal=:dog),
Chunk(;name=:Bonkers, animal=:rat)]
# initialize declarative memory
declarative = Declarative(memory=chunks)
# specify model parameters: partial matching, noise, mismatch penalty, activation noise
Θ = (mmp=true, noise=true, δ=1.0, s=0.20, blc=1.5)
# create an ACT-R object with activation noise and partial matching
actr = ACTR(;declarative, Θ...)
# retrieve a memory chunk
retrieve(actr; animal=:rat)
Chunks
┌──────────────────────────────────┬───────┬───────┬──────────────┬────────┬────
│ slots │ N │ L │ time_created │ recent │ a ⋯
├──────────────────────────────────┼───────┼───────┼──────────────┼────────┼────
│ (name = :Bonkers, animal = :rat) │ 1.00 │ 1.00 │ 0.00 │ [0.0] │ ⋯
└──────────────────────────────────┴───────┴───────┴──────────────┴────────┴────
7 columns omitted
References
Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., & Qin, Y. (2004). An integrated theory of the mind. Psychological review, 111(4), 1036.