Mining Symptom-Herb Patterns from Patient Records Using Tripartite Graph

Algorithm 1

Tri-TSPA.

Input: Weight Matrix

Output: Top- List

(1) Define Tri-TSPA()

(2) Begin

(3) queue ← Discover nearest entities using the similarity measures

(4) Case  1. for do

(5) +

(6)    

(7)    End for

(8) Case  2. for do

(9) = +

(10)    

(11)    End for

(12) Top- list ← Get the predicted patterns list in the term of

(13) or

(14) Top- list ← Filter the Top- list using relevance function

(15) Return Top- list

(16) End