[ Prev ]
2022-05-15

-- May 11 In-Class Exercise
Resource Description for Screen Shot 2022-05-15 at 11.09.08 PM.png
(Edited: 2022-05-15)
((resource:Screen Shot 2022-05-15 at 11.09.08 PM.png|Resource Description for Screen Shot 2022-05-15 at 11.09.08 PM.png))

-- May 11 In-Class Exercise
Given: 
N = 5000, X = 100
n(Batches) = N/X = 50
t = N-X = 4900 
 
Pseudocode:
1. Iterate over each batch and shortlist docs with relevance > t
2. Check whether each doc is relevant
3. Compute prec
4. Compute avg prec
5. Compute agg p@100
<pre> Given: N = 5000, X = 100 n(Batches) = N/X = 50 t = N-X = 4900 Pseudocode: 1. Iterate over each batch and shortlist docs with relevance > t 2. Check whether each doc is relevant 3. Compute prec 4. Compute avg prec 5. Compute agg p@100 </pre>

-- May 11 In-Class Exercise
N=5000
X = 100 documents
No of batches n = N / X = 5000/ 100 = 50 batches
t = N - X = 5000 - 100 = 4900
For every batch get the docs with a score > 4900.
Check if the docs are relevant
Calculate precision
Calculate avg precision
Calculate P@100
N=5000 X = 100 documents No of batches n = N / X = 5000/ 100 = 50 batches t = N - X = 5000 - 100 = 4900 For every batch get the docs with a score > 4900. Check if the docs are relevant Calculate precision Calculate avg precision Calculate P@100

-- May 11 In-Class Exercise
N=5000 X = 100 documents No of batches n = N / X = 5000/ 100 = 50 batches t = N - X = 5000 - 100 = 4900 For every batch get the docs with a score > 4900. Check if the docs are relevant Calculate precision Calculate avg precision Calculate P@100
N=5000 X = 100 documents No of batches n = N / X = 5000/ 100 = 50 batches t = N - X = 5000 - 100 = 4900 For every batch get the docs with a score > 4900. Check if the docs are relevant Calculate precision Calculate avg precision Calculate P@100
2022-05-16

-- May 11 In-Class Exercise
Given: N = 5000, 
       X = 100, 
       Number of batches = N/X = 50 
       Threshold score = N - X = 4900
For every batch: 
    1. Find documents above the score of 
       Threshold score (4900) and determine whether
       it is relevant
    2. Calculate precision 
    3. Calculate the average precision of all batches
    4. Calculate aggregate P@100
((resource:may11_pk.txt|Resource Description for may11_pk.txt))

-- May 11 In-Class Exercise
Resource Description for Screenshot 2022-05-16 at 11.25.51 PM.png
((resource:Screenshot 2022-05-16 at 11.25.51 PM.png|Resource Description for Screenshot 2022-05-16 at 11.25.51 PM.png))

-- May 11 In-Class Exercise
Given: 
N = 5000, X = 100
n(Batches) = N/X = 50
t = N-X = 4900 
 
Pseudocode:
1. Iterate over each batch and shortlist docs with relevance > t
2. Check whether each doc is relevant
3. Compute prec
4. Compute avg prec
5. Compute agg p@100
(Edited: 2022-05-17)
<pre> Given: N = 5000, X = 100 n(Batches) = N/X = 50 t = N-X = 4900 Pseudocode: 1. Iterate over each batch and shortlist docs with relevance > t 2. Check whether each doc is relevant 3. Compute prec 4. Compute avg prec 5. Compute agg p@100 </pre>
X