This issue is a part of Google Code-in 2019 task. As a part of the task I'm reporting my findings on running the pre-trained SentimentAnalyzer with the Amazon review dataset as requested.
I found the accuracy of the model to be 51.71916007215394 percent.
Also, I found this issue https://github.com/JuliaText/TextAnalysis.jl/issues/160 and for that I simply ignored all such instances where the error was being reported. I found out the difference is very low and hence can be neglected.
I've also uploaded the notebook file to the GCI website.
Ayushk4
commented
4 years ago
This issue is a part of Google Code-in 2019 task. As a part of the task I'm reporting my findings on running the pre-trained SentimentAnalyzer with the Amazon review dataset as requested.
I found the accuracy of the model to be 51.71916007215394 percent.
Also, I found this issue https://github.com/JuliaText/TextAnalysis.jl/issues/160 and for that I simply ignored all such instances where the error was being reported. I found out the difference is very low and hence can be neglected.
I've also uploaded the notebook file to the GCI website.
cc @aviks @Ayushk4 @dhairyagandhi96