Here is  my solution. I had scale the similarity 0.9 between -10 ~ 20 which is 17

hsankav wrote:

Pallavi0205 wrote:

Here is  my solution. I had scale the similarity 0.9 between -10 ~ 20 which is 17

Hi Pallavi

Acccording to paper the optimal alignment score is A(m,n) which is score of  A(P4,P7) = 25 but not A(P4,P6) please check.

 

No we have to take the maximum value  in the matrix. actually i had made mistake in the solution .This is what I corrected and I got 44 as best alignment score.

HelloWorld wrote:

I'm curious.. what are you talking about here:

 

The range of score is between -10 to 20

The range of page similarity is between 0 to 1

so we need to scale the Similarity of page P4 and P6 in the range of -10 ~20

so after scaling we get 17

i.e 30*0.9 -10 = 17

so now the S(Xi,Yj) i.e S(P4,P6) = S(P6,P4) =17

if pages are similar then S(Xi,Yj) = 20

else S(Xi,Yj) = -10

 

nevermind, i got it..

HelloWorld wrote:

for this one though.. isn't it suppose to be..

Optimal Alignment Score = 44

and 

Session Similarity = Optimal Alignment Score / Length of Longer Session = 44 / 5 = 8.8

No it is given in the paper that when we carry out the computaion from upper left corner to lower right corner, A(m,n) is optimal alignment score between X1.......Xm and Y1..........Yn.