Learn a word2vec model from fake news dataset and a real news dataset. We will use the word2vec model implemented in the Python Gensim library. Now we have two sets of word representations learnt from different datasets. Output the top 5 most similar words to the following ones from each representation. 1. Hillary 2. Trump 3. Obama 4. Immigration In order to do this we first normalize all vector representations (set them to Euclidean length 1). Consider the vector x for a given word w. We compare the cosine similarity between x and the vectors x' for each word w' in the fake news dataset first. We then output the top 5 words with highest similarity. We then do the same for the real news and then see if the top similar words differ considerably. Submit your assignments as two files train.py and test.py. Make train.py take two inputs: the text dataset on which to learn the words and a model file name to save the word2vec model to. python train.py Make test.py take three inputs: text dataset, word2vec model, a query file containing five query words. The output should be the top five most similar words to each word in the query file. python test.py Are the most similar words to the queries considerably different from the fake and real news datasets? Copy both your programs and model file to your AFS course folder /afs/cad/courses/ccs/s20/cs/677/002/. The assignment is due 11:30am on May 4th 2020.