It's UWAweek 48

help2003/help4407

This forum is provided to promote discussion amongst students enrolled in Open Source Tools and Scripting.

Please consider offering answers and suggestions to help other students! And if you fix a problem by following a suggestion here, it would be great if other interested students could see a short "Great, fixed it!"  followup message.

How do I ask a good question?
Displaying the 11 articles in this topic
Showing 11 of 564 articles.
Currently 3 other people reading this forum.


 UWA week 19 (1st semester, week 10) ↓
SVG not supported

Login to reply

👍?
helpful
11:27pm Sun 15th May, Amandeep AS.

hi Micheal my code does produce the desired output as required for the profiles exactly same as shown in sample output for both profile creation and normalized profile creation. but when i try to compare two texts the answers are completely different. one reason for this that i have seen is that i believe that your comparison is a bit wrong. for instance in the lawrence and lawson samples there are 0 instances of "also" and 0 instances of "although". but in your provided comparison you have shown also 2.338 although 0. how can also difference of squares be equal to 2.338 when there is no "also" word in any of the texts. please fix this asap. THANKS


 UWA week 20 (1st semester, week 11) ↓
SVG not supported

Login to reply

👍?
helpful
6:38pm Mon 16th May, Amandeep AS.

??


SVG not supported

Login to reply

👍?
helpful
7:21pm Mon 16th May, ANONYMOUS

Michael's probably busy watching the Eurovision Grand Final.


SVG not supported

Login to reply

👍?
helpful
12:23pm Tue 17th May, Justin H.

Agreed about the "also" field, I noticed that too, I think it's a mistake.


SVG not supported

Login to reply

👍?
helpful
12:54pm Tue 17th May, Justin H.

Just to add: Michael's definitely compared different texts to what he's listed. We'll just have to wait for an answer and see. If you look at the values of certain words present in one sample text but not the other, for example "which" and "yet", since in your distance metric it'll be <somenumber>-0 squared, and then rooted, you should arrive back at the same <somenumber> again. Michael's does not. I'm not willing to share actual values due to rules against collusion, but since the mathematical formula is shared, I think I'm allowed to make that point. tldr: we'll wait for an update, but I'm sure the current numbers are incorrect :)


SVG not supported

Login to reply

👍?
helpful
1:21pm Tue 17th May, Justin H.

Never mind, I misread "What I have also chosen to report is the SQUARES of the style-metric differences", i.e. his values aren't rooted yet.


SVG not supported

Login to reply

👍?
helpful
1:52pm Tue 17th May, Michael W.

Hi Everyone, While I occasionally get things wrong - and am very appreciative of people who let me know so I can correct things - but this time I really did say, "... What I have also chosen to report is the squares of the style-metric differences,..". That is the finally metric requires square-root, but the individual items, which I report (but you don't have to), are the squares of the differences, which are summed, and the square-root is taken. Does that make sense, or have I missed something? Cheers MichaelW


SVG not supported

Login to reply

👍?
helpful
1:59pm Tue 17th May, Justin H.

Hi Michael, when you compared the "Lawrence and Lawson samples" , I believe your example comparison is incorrect. For example, neither text contains the word "also", but after your comparison, "also" has a value other than zero, which can't possibly be true. Screenshots attached. This leads me to believe you didn't in fact compare the lawrence sample with the lawson sample. Apologies if I've misunderstood something.



This article has 3 attachments:

 

 

 

SVG not supported

Login to reply

👍?
helpful
2:00pm Tue 17th May, Justin H.

Just to add: those are screenshots of your examples, not my own work. But as you can see, the square of the differences for "also" must also be zero, since the two inital profiles were also zero. So something is wrong here.


SVG not supported

Login to reply

👍x1
helpful
2:45pm Tue 17th May, Chuanhe L.

hi, i believe that Michael are using Kangaroo(Lawrence_sample.txt) in the comparing part, which i got really similar results with Michael's, you may need to download kangaroo one.


SVG not supported

Login to reply

👍?
helpful
9:26pm Tue 17th May, Michael W.

"Justin Hart" <22*4*5*[email protected]*u*e*t*u*a*e*u*a*> wrote:
> Hi Michael, when you compared the "Lawrence and Lawson samples" , I believe your example comparison is incorrect. For example, neither text contains the word "also", but after your comparison, "also" has a value other than zero, which can't possibly be true. Screenshots attached. This leads me to believe you didn't in fact compare the lawrence sample with the lawson sample. > Apologies if I've misunderstood something.
Hi Justin, There are 2 instances of "also" in Lawrence, 0 in Lawson. However, when that counts for each profile are normalised (as per the spec), Lawrence goes to 1.52905 . Lawson, of course remains 0. The difference therefore is 1.52905, which when squared is 2.338 (to 3 decimal places), which is what is seen as the value for square difference of _normalised_ values for "also". No? Cheers MichaelW

The University of Western Australia

Computer Science and Software Engineering

CRICOS Code: 00126G
Written by [email protected]
Powered by history
Feedback always welcome - it makes our software better!
Last modified  1:17AM Sep 14 2022
Privacy policy