I got some job for you . I got 3 tables as attached. All metrics are from 0 to 100 only NIST is from 0 to 15 and TER - the lower the better.
IN this tables row are test samples. I asked 15 people to do some task. They had to do transcipting from text that was read for them. The same was done using automatic computer system. In excel I have 3 sheet, sheet 1 is comaprison betwwen original text and work done by humans, sheet 2 has comaprison between original text and automatic system work. sheet 3 has comparison between human work and automatic tool - the better scores the more simillar.
What I need is some analysis of how different automatic work is that the one done by humans, information if automatic work correlates with work done by humans, some information which metric is most reliable - sometimes for example BLEU decreases when EBLEU rises.
For all of those I would need significance tests.
FInally I would need some results desription - as detailed as possible with description how analysis was done. In addtiion I would need some visualization for example using some diagrams - the resutls can be norlalized for easier undestanding. Could you help me with that ??
IN some time I will add a column with human jundgements and I will need to compare it to automatic methods. So the methouds you choose should be easy to upgrade results when I got new data
and t is quite urgent :) I Got 48h for this
21 freelancer menawar dengan rata-rata $169 untuk pekerjaan ini
Hello, Expert here, I am a Statistician involved in United Nations Projects in Geneva. I would be very happy to assist you in this project. I am available to discuss now, Hear from you soon, Best Regards Fouad
Greetings. I am well experienced in data analytics and I am also a proficient Excel, SPSS and R user. I am sure that I can do the work. Feel free to message me when you wish to discuss further. Thanks
I have a significant background in this kind of tasks for academic and publication level. I would need to know in which format do you need the analysis and if you do need the code to reproduce the results.