Mental Mechanisms for Topics Identification

Algorithm 1

ReAD algorithm.

(1)   Inputs: document, dictionary, list of stop and undesirable words,

decay rate η,  and STM capacity

(2)   For each sequential word 𝚠_𝚒  from a document (only if 𝚠_𝚒  is not in

list of stop words):

     (2.1)   Place 𝚠_𝚒  in STM and remove word that has been in STM the longest

     (2.2)   Query dictionary with 𝚠_𝚒

          (2.2.1)  𝚃_𝚒  =  items from dictionary corresponding to word 𝚒

     (2.3)   Remove stop and undesirable words from 𝚃_𝚒

     (2.4)   For each item k  in 𝚃_𝚒  (initial activation for new item = 0)

          (2.4.1)   Increase item activation: 𝚊_𝚔  = 𝚊_𝚔  +  1

     (2.5)   For each item k  for whichword 𝚠_𝚒  is not in STM anymore

          (2.5.1)   Decay item’s existing activation: 𝚊_𝚔  = 𝚊_𝚔  −  1/η

(3)   Output: Select the most activated items as topics