Big Data Aspect-Based Opinion Mining Using the SLDA and HME-LDA Models

Algorithm 2. The initialization of Gibbs sampling.

1: for to Ddo

2:  //Randomly assign topic t to sentences in Corpus

3:  t = randomint from (1, T);

4:  documentTopics [d] = t;

5:  documentTopicsCount [d][t]++;

6:  for to Ndo

7:   //Randomly assign value to y

8:   y = randomint from (0,1);

9:   ;

10:   //Randomly assign value to u

11:   u = randomint from (0,1);

12:   ;

13:   ify ==0 then

14:    //statistic of aspect targets with topic t plus 1

15:    aspectWordCount [w][t]++;

16:   end if

17:   ify ==1 and u ==0 then

18:    //statistic of positive opinion with topic t plus 1

19:    positiveWordCount [w][t]++;

20:   end if

21:   ify ==1 and u ==1 then

22:    //statistic of negative opinion with topic t plus 1

23:    negativeWordCount [w][t]++;

24:   end if

25:  end for

26: end for

Algorithm 2. The initialization of Gibbs sampling.