Predict sentiment of new data

[ravilkashyap]

@linanqiu Thanks for the wonderful presentation of your program!
However, I want to know how this model can be used to predict the sentiment of a given sentence.

[KseniaCadaques]

@ravilkashyap, hi, I ran into similar issue and the only way I could resolve was that I trained the model and sentence vectors for all the data (testing and training together), how @linanqiu seems to do in his notebook, and then I just used test docvecs to predict the labels for a classifier trained on docvecs for train data. It seems to me that in this way I am overfitting the data because I train word embeddings for all the data at the same time using the test data.

Otherwise I am sure that there is a way to update a model receiving sentence embeddings for the test data but I don’t understand how to do it without training all of the model. Hope somebody else will enlighten us!

[liberocks]

I also tried to predict new given sentence(s). So, this is my version of solution. Replace s with your sentence. Thus you don't need to train again whenever you want to predict sentence.

Code :

s = "its ten minutes of script strecthed to two and half hours of boredom i was really surprised how weak this is spoiler alert it s really boring"

s = s.rsplit()
s = model.infer_vector(s)

s_vector = numpy.zeros((1, 100))
s_vector[0] = s

classifier.predict(s_vector)

[KseniaCadaques]

Thank you, in the end I did it with the cycle for the test set and predict as well! Very beautiful solution. <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free. www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

…

On Tue, Jan 30, 2018 at 11:49 AM, linerocks ***@***.***> wrote: I also tried to predict new given sentence(s). So, this is my version of solution. Replace s with your sentence. Thus you don't need to train again whenever you want to predict sentence. Code : s = "its ten minutes of script strecthed to two and half hours of boredom i was really surprised how weak this is spoiler alert it s really boring" s = s.rsplit() s = model.infer_vector(s) s_vector = numpy.zeros((1, 100)) s_vector[0] = s classifier.predict(s_vector) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#20 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AX3_jq0Tc_vJTycG28tuY7IDSeDQscQpks5tPvPXgaJpZM4PwErd> .

-- Regards, Ksenia Kharitonova