Algorithm 1: Categorization Based on Keywords or Keyword Combinations

1: Load dataset D ← CSV

2: For each record di ∈ D do

3:    If di.k = ∅ then

4:       di.k ← ""

5:    End if

6:    di.k ← lowercase(di.k)

7: End for

8: Define category-keyword map: $\mathcal{C}$ = {c1 ↦ K1,…, cn ↦ Kn}

9: For each di ∈ D do

10:    Si ← ∅

11:    For each (cⱼ, Kⱼ) ∈ $\mathcal{C}$ do

12:      If ∃k ∈ Kⱼ such that k ∈ di.k then

13:        Si ← Si ∪ {cⱼ}

14:      End if

15:    End for

16:    di.c ← Si

17: End for