N grams for Sentiment Analysis

0
votes

I am doing sentiment analysis on reviews of products from various retailers. I was wondering if there was an API that used n grams for sentiment analysis to classify a review as a positive or negative. I have a CSV file filled with reviews which I would like to run it in python and hence would like an API or a package rather than a tool. Any direction towards this would be great.

Thanks

1
pythonnltksentiment-analysisn-gram
Why don't you try a NN approach for sentiment analysis? Example: github.com/SqrtPapere/SentimentAnalysis_CNN - Luca Angioloni
@Luca Angioloni My dataset isn't big enough. I only have 30k reviews. As far as I'm aware, Neural network requires a much bigger dataset. - paddy
Not necessarily, in the example I provided the NN is trained with 25k reviews. - Luca Angioloni
But you need a training dataset. If this guy has 30k reviews and he has no labels, there's no training data available. - Zizzipupp

1 Answers

0
votes

Use textblob package. It offers a simple API to access its methods and perform basic NLP tasks. NLP is natural language processing. Which process your text by tokenization, noun extract, lemmatization, words inflection, NGRAMS etc. There also some other packages like spacy, nltk. But textblob will be better for beginners.