Projects with this topic
Sort by:
-
Evaluation of various Word2Vec models for sentiment analysis You are given the reviews dataset. These are 194439 amazon reviews for cell phones and accessories taken from https://jmcauley.ucsd.edu/data/amazon/ Use the “reviewText” and “overall” fields from this file. The goal is to predict the rating given the review by modeling it as a multi-class classification problem.
• Average of word embeddings • Learn word2vec models using gensim on this dataset with the following settings: (a) Size=100, 200, 300, (b) Window=3,7, (c) Min_count=2, 5. Use skipgram. • This will give 12 word2vec models.
Updated