Replies to This Forum

Permalink Reply by Langmuanjohn Tonsing on February 12, 2012 at 23:18

hello Sangeeta,

Let me give a try

Formula is Recall = Number of relevant result retrieve / Number of non-retrieve document + Number of relevant result retrieve

so, Number of relevant result retrieve=10,

     Number of non-retrieve documents=20 (50-30=20)

Recall=10/20+10

         =10/30

         = 0.33

         =0.33*100=33% (i.e. 33/100=0.33)

This is wat i think according to the formula, if it is wrong let me know.

• ▶ Reply

Permalink Reply by Langmuanjohn Tonsing on February 12, 2012 at 23:26

Sori, the last line " =0.33*100=33% (i.e. 33/100=0.33)" should be =0.33 =33% (i.e. 33/100=0.33).

• ▶ Reply

Permalink Reply by Vishal Sharma on February 12, 2012 at 23:47

10*100/30=33.33%

• ▶ Reply

Permalink Reply by Langmuanjohn Tonsing on February 13, 2012 at 0:28

Dear Vishal Sharma 

Can you please explain in details?

• ▶ Reply

Permalink Reply by Dr.Harmanpreet Singh Sandhu on February 13, 2012 at 1:50

Very Good Question asked, with regards to upcoming pattern of the UGC NET June 2012.   Recall percentage is the percentage of documents retrieved which were of actually requested or relevant documents with reference to the query.

To find out the Recall % = Total no. of relevant documents delivered / Total no. of documents deliverd * 100

  i:e   10/30*100 = 33.33 %

When searching in google we don't know the total relevant documents uploaded on the net ..... so the recall % is found out by this formula only.

• ▶ Reply

Permalink Reply by Langmuanjohn Tonsing on February 13, 2012 at 1:58

Dear Harmanpreet,

Can you also please give the formula for precision?

• ▶ Reply

Permalink Reply by Sangeeta Dubey on February 13, 2012 at 18:41

Lot of thanks for your opinion.

• ▶ Reply

Permalink Reply by Vishal Sharma on February 13, 2012 at 18:53

welcome ma'am

• ▶ Reply

Permalink Reply by K. Nagaraja Rao on February 14, 2012 at 2:43

According to me: Recall is intersection of relevant documents and retrieved documents divided by relevant documents in the system:

thus the intersection gives only 10 out of 50 relevant documents. 1/5*100

Which is 20%. To get 33%, the question should say 30 relevant documents from a total collection.

formula taken from wikipedia.

• ▶ Reply

Permalink Reply by Sangeeta Dubey on February 14, 2012 at 2:52

Thank you for your efforts sir.

I ll check it. n inform you.

• ▶ Reply

Permalink Reply by rawish on March 4, 2012 at 21:28

mam if u got the recall value for your questions plz tell me the correct answer and also the method.thanks

• ▶ Reply