How to perform regression algorithm on a textual data IMDB reviews

Question

                 reviews              label
0   i admit the great majority of...    1
1   take a low budget inexperienced ... 0
2   everybody has seen back to th...    1
3   doris day was an icon of b...       0
4   after a series of silly fun ...     0

I've a dataframe of movie reviews and the label column(1-postive , 0-negative review).
I've another similar test dataset with only review columnI need to build a sentiment analysis model using linear regression to predict the label column of test dataframe.
Desired output: Test dataframe with label column. Regression is performed on numerical data , how do convert text review to numeric form to be able to fit it?

Answer 1

You can use either word2vec or tf-idf to convert words to vectors.

Word2vec's key benefit is that words with similar meanings will have similar encodings. This is particularly intriguing when attempting to predict the positive of a review. This is not possible with td-idf because it is based solely on the occurrence of words.

Word to vector To map words into vector space, the procedure employs language models. Each word in a vector space is represented by a vector of real integers. It also allows for similar representations of words with similar meanings.

"Term Frequency — Inverse Document Frequency" is abbreviated as TF-IDF. This is a method for calculating the number of words in a collection of documents. We usually assign each word a score to indicate its prominence in the document and corpus. This method is commonly used in text mining and information retrieval.