Possible missing subscript on Equation for CRFs

[lzfelix]

Dear professor Eisenstein,

Thank you very much for making the book publicly available.
I would like to point a possible typo on the equation on section 7.5.3.1:

Should not the s function be written s_m, indicating the transition between labels y_{m-1} and y_m on the i-th position of the underlying w sequence?

Thank you very much,
Luiz C F Ribeiro

[yuvalpinter]

Hi Luiz,
IIRC s is a global transition matrix, which is not dependent on position.
Yuval

[lzfelix]

Hi Yuval,

You are right regarding the transition matrix being global, irrespective to elements position.

I actually forgot the complete definition of s_m(y_m, y_{m-1}) from an example on the book, where the function is defined as a sum of the score of the m-th element of the sequence having the label y and the transition between tags. This is defined formally on Equation 7.16.

What caught my attention is that s has always a subscript on this chapter, except on this equation.

[jacobeisenstein]

Luis, you are correct. s_m is indexed by m because it describes the score for the tag transition y_{m-1} -> y_m specifically on position m. I will make an update. Thanks! -Jacob

…

On Tue, Jun 19, 2018 at 6:17 AM, Luiz Felix ***@***.***> wrote: Hi Yuval, You are right regarding the transition matrix being global, irrespective to elements position. I actually forgot the complete definition of s_m(y_m, y_{m-1}) from an example on the book, where the function is defined as a sum of the score of the m-th element of the sequence having the label y and the transition between tags. This is defined formally on Equation 7.16. What caught my attention is that s has always a subscript on this chapter, except on this equation. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#29 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA3QHG6DXPubzxYylcp_nlk5AEnU059cks5t-Pn5gaJpZM4UtdDW> .