An Efficient Parallelized Ontology Network-Based Semantic Similarity Measure for Big Biomedical Document Clustering

Algorithm 3

Algorithm of MapReduce-based document similarity calculation.

Document similarity calculation

Input: <m, list(d)>

Output: <pair of d, similarity>

Notation: Write (k, v) outputs <k, v>

Class mapper

 Method map (heading, list(d))

  m ← heading

  For each d₁ ∈ D

   r ← Sim (m, d₁)

   For each d₂ in list(d)

    s ← string (d₁ +” &” + d₂)

    Write (s, r)

   End for

  End for

Class reducer

 Method reduce (s, list(r))

Sum ← 0, count ← 0

 For each r in list(r)

   Sum ← sum + r

   Count ← count +1

  End for

  Write (s, sum/count)