A Functional Model of Sensemaking in a Neurocognitive Architecture

Table 2

The list of sub-symbolic mechanisms in the ACT-R architecture.

Mechanism	Equation	Description
Activation	𝐴𝑖=𝐵𝑖+𝑆𝑖+𝑃𝑖+𝜀𝑖	𝐵𝑖: base-level activation reflects the recency and frequency of use of chunk i; 𝑆𝑖: spreading activation reflects the effect that buffer contents have on the retrieval process; 𝑃𝑖: partial matching reflects the degree to which the chunk matches the request; 𝜀𝑖: noise value includes both a transient and (optional) permanent components (permanent component not used by the integrated model)
Base level	𝐵𝑖=ln∑^𝑛𝑗=1𝑡𝑗^−𝑑+𝛽𝑖	𝑛: the number of presentations for chunk i; 𝑡𝑗: the time since the jth presentation; 𝑑: a decay rate (not used by the integrated model); 𝛽𝑖: a constant offset (not used by the integrated model)
Spreading activation	𝑆𝑖=∑𝑘∑𝑗𝑊𝑘𝑗𝑆𝑗𝑖,	𝑘: weight of buffers summed over are all of the buffers in the model; 𝑗: weight of chunks which are in the slots of the chunk in buffer k; 𝑊𝑘𝑗: amount of activation from sources j in buffer k; 𝑆𝑗𝑖: strength of association from sources j to chunk i
	𝑆𝑗𝑖=𝑆−lnfan𝑗𝑖	𝑆: the maximum associative strength (set at 4 in the model); fan𝑗𝑖: a measure of how many chunks are associated with chunk j
Partial matching	𝑃𝑖=∑𝑘𝑃𝑀𝑘𝑖	𝑃: match scale parameter (set at 2) which reflects the weight given to the similarity; 𝑀𝑘𝑖: similarity between the value k in the retrieval specification and the value in the corresponding slot of chunk i; The default range is from 0 to −1 with 0 being the most similar and −1 being the largest difference
Declarative retrievals	𝑃𝑖=𝑒^𝐴𝑖/𝑠∑𝑗𝑒^𝐴𝑗/𝑠	𝑃𝑖: the probability that chunk 𝑖 will be recalled; 𝐴𝑖: activation strength of chunk 𝑖; ∑𝐴𝑗: activation strength of all of eligible chunks 𝑗; 𝑠: chunk activation noise
Blended retrievals	𝑉=min∑𝑖𝑃𝑖1−Sim𝑖𝑗^2	𝑃𝑖: probability from declarative retrieval; Sim𝑖𝑗: similarity between compromise value 𝑗 and actual value 𝑖
Utility learning	𝑈𝑖(𝑛)=𝑈𝑖(𝑛−1)+𝛼𝑅𝑖(𝑛)−𝑈𝑖(𝑛−1),	𝑈𝑖(𝑛−1): utility of production i after its 𝑛−1st application; 𝑅𝑖(𝑛): reward production received for its nth application; 𝑈𝑖(𝑛): utility of production i after its nth application
	𝑃𝑖=𝑒^𝑈𝑖/𝑠∑𝑗𝑒^𝑈𝑗/𝑠	𝑃𝑖: probability that production i will be selected; 𝑈𝑖: expected utility of the production determined by the utility equation above; 𝑈𝑗: is the expected utility of the competing productions j