2011-12-08

HW4 - MapReduce problem 14.7.

Originally Posted By: sheetalg
The question asks to design a MapReduce that computes the conditional probability Pr[t1 | t2] of seeing the term t1 in a document that contains the term t2.

So my questions are:
1. What will be the input to map function ? map(k, v) - What should be k and v ? (My guess is k = url, and v= docid)
2. How do you compute this probability ? My guess is P(a | b) = P (b | a) * P(a) / P (b). In this case as well, how do I compute P (b | a) ?
3. The question talks something about map function emitting secondary keys to enforce a certain ordering. Can you highlight what ordering is it talking about ?

Thanks,
Sheetal
'''Originally Posted By: sheetalg''' The question asks to design a MapReduce that computes the conditional probability Pr[t1 | t2] of seeing the term t1 in a document that contains the term t2.<br><br>So my questions are:<br>1. What will be the input to map function ? map(k, v) - What should be k and v ? (My guess is k = url, and v= docid)<br>2. How do you compute this probability ? My guess is P(a | b) = P (b | a) * P(a) / P (b). In this case as well, how do I compute P (b | a) ?<br>3. The question talks something about map function emitting secondary keys to enforce a certain ordering. Can you highlight what ordering is it talking about ?<br><br>Thanks,<br>Sheetal

-- HW4 - MapReduce problem 14.7
Originally Posted By: sheetalg
Anyone, any clue ?

Thanks,
Sheetal
'''Originally Posted By: sheetalg''' Anyone, any clue ?<br><br>Thanks,<br>Sheetal
X