reset password
Author Message
jadiagaurang
Posts: 53
Posted 17:08 Jul 22, 2009 |

I have tried to understand from slides and paper. I have taken d = -10 and used score calculation given in paper. Please check my calculation. I have attached file with post. I hope that I did it correct. Innocent

Attachments:
xieguahu
Posts: 50
Posted 19:29 Jul 23, 2009 |

should the similarity(Pa, Pb) same as the similarity (Pb, Pa)?

xieguahu
Posts: 50
Posted 19:45 Jul 23, 2009 |

I did assuming the similarity of p4,p6 is same as similarity of p6, p4.  Other assumptions are the same

Attachments:
Last edited by xieguahu at 18:56 Jul 25, 2009.
jadiagaurang
Posts: 53
Posted 20:36 Jul 23, 2009 |

Hello xieguahu,

If I did not consider similarity of (P6,P4) and (P4,P6) as same. But, If we consider as same then S(P6,P4) = S(P4,P6) = 17. I have corrected my calucation and marked as bold figure. Please take a look!

Attachments:
hsankav
Posts: 18
Posted 10:43 Jul 25, 2009 |

Gaurang,

I solved the Exercise as we discussed yesterday .I got a different value for the pair A(P4,P7) Could you please check my solution

Attachments:
arctica82
Posts: 10
Posted 12:43 Jul 25, 2009 |

I am somehow confused,Please let me know If I am wrong,

First, optimal alignment is the last element in our calculated values,but I have noticed that you have taken the Maximum value.

this is from paper

Therefore, we can carry out the computation from upper left corner to lower right corner,

A(m,n), which is the optimal alignment score between X

 

 

1…Xm and Y1…Yn

Also for calculating the first row and first column you have consider d=-10

so the first row will be 10,20,30,...

Because d is negative itself.

jadiagaurang
Posts: 53
Posted 15:25 Jul 25, 2009 |
hsankav wrote:

Gaurang,

I solved the Exercise as we discussed yesterday .I got a different value for the pair A(P4,P7) Could you please check my solution

Hello Haritha,

You are right for A(A4, P7). I did small mistake there. It should be 7. A(P4,P7) = MAX[(10 + -10);(10 + -10);(17 + -10)] = [0;0;7] = 7. But, I am still not sure about should we consider S(P4,P6) same as S(P6,P4).

jadiagaurang
Posts: 53
Posted 15:44 Jul 25, 2009 |

If I consider S(P4,P6) is not same as S(P6,P4) then I get this table and Best (Optimal) alignment score is 20.

If S(P4,P6) = 0.9 and S(P6,P4) = 0 then A(P4,P6) = 17, A(P6,P4) = -10

 

P5

P4

P1

P6

P7

0

-10

-20

-30

-40

-50

P1

-10

-10

-20

0

-10

-20

P6

-20

-20

-20

-10

20

10

P1

-30

-30

-30

0

10

10

P4

-40

-40

-10

-10

17

7

 

If I consider S(P4,P6) is same as S(P6,P4) then I get this table and Best (Optimal) alignment score is 44. A(P4,P6) = A(P6,P4) = 17.

If S(P4,P6) = 0.9 and S(P6,P4) = 0.9 then A(P4,P6) = A(P6,P4) = 17.

 

 

P5

P4

P1

P6

P7

0

-10

-20

-30

-40

-50

P1

-10

-10

-20

0

-10

-20

P6

-20

-20

7

-3

20

10

P1

-30

-30

-3

27

17

10

P4

-40

-40

-10

17

44

34

HelloWorld
Posts: 88
Posted 17:29 Jul 25, 2009 |

Hi everyone,

here's mine.. feel free to check it

 

Attachments:
hsankav
Posts: 18
Posted 18:27 Jul 26, 2009 |

The table to be considered for computing optimal alignment score is of  2 versions.which one to consider? Do I need to consider d or -d in the table for computing? I am confused? Can anyone help me out!! Please check the attachment for tables.

 

 

Attachments:
Last edited by hsankav at 18:27 Jul 26, 2009.
hsankav
Posts: 18
Posted 19:55 Jul 26, 2009 |

Here is my solution for Exercise 4 considering table from the paper and d = -10. I hope this is correct

Attachments:
jadiagaurang
Posts: 53
Posted 20:21 Jul 26, 2009 |

Hello Guanhua, Harita and Grady,

I am very much confused about this question. I am still thinking that we need to calculate scores. There is one example of Page-1 and Page-2 in paer LuDM05. They found page similarity and score of those pages. I am considering S(P4,P6)=S(P6,P4)=0.9 as I have posted earlier. I guess that if two pages are same then we should consider same page similarity for them. So, Score of both is 17. Here d = -10 and harita, I am using calucation table from lacture slides. So, my gap is like -10, -20, -30, ... and d = -10. I posted 44 as best alignment but i read paper again and lower right corner A(m,n) is best alignment, which is 34 in my case. I am not sure what is corrent approch but guys it is really ambiguous.

 

P5

P4

P1

P6

P7

0

-10

-20

-30

-40

-50

P1

-10

-10

-20

0

-10

-20

P6

-20

-20

7

-3

20

10

P1

-30

-30

-3

27

17

10

P4

-40

-40

-10

17

44

34

Best (Optimal) alignment score is 34 and Session similarity = 34/5 = 6.8

xieguahu
Posts: 50
Posted 20:26 Jul 26, 2009 |
I read the paper again, and I think you are right

jadiagaurang wrote:

Hello Guanhua, Harita and Grady,

I am very much confused about this question. I am still thinking that we need to calculate scores. There is one example of Page-1 and Page-2 in paer LuDM05. They found page similarity and score of those pages. I am considering S(P4,P6)=S(P6,P4)=0.9 as I have posted earlier. I guess that if two pages are same then we should consider same page similarity for them. So, Score of both is 17. Here d = -10 and harita, I am using calucation table from lacture slides. So, my gap is like -10, -20, -30, ... and d = -10. I posted 44 as best alignment but i read paper again and lower right corner A(m,n) is best alignment, which is 34 in my case. I am not sure what is corrent approch but guys it is really ambiguous.

 

P5

P4

P1

P6

P7

0

-10

-20

-30

-40

-50

P1

-10

-10

-20

0

-10

-20

P6

-20

-20

7

-3

20

10

P1

-30

-30

-3

27

17

10

P4

-40

-40

-10

17

44

34

Best (Optimal) alignment score is 34 and Session similarity = 34/5 = 6.8

 

Pallavi0205
Posts: 30
Posted 23:28 Jul 26, 2009 |

Here is my solution for exercise #4. I  scaled the similarity 0.9 in the range of -10 ~ 20.

Attachments:
alomo
Posts: 70
Posted 08:12 Jul 27, 2009 |

 

There are two similarities given in this exercise: similarity(P4,P6) and similarity(P5,P7).

We used the (P4,P6)-similarity. But why do we need another similarity, for (P5,P7)?

Aren’t we missing something in our solutions?