GitHub - HelenaJulie/n-gram-from-scratch--Classification-of-positive-and-negative-reviews-nlp: Using python, implement n-gram approach starting from n = 1 to n = 5 for feature representation. Let f1 ...fk be the vectors representation of the n-gram features, whereby f1 is bag of words, f2 is the frequency of pairs of words, as deﬁned for the n-gram model, and so forth. Using the n-gram representations, build and test a binary classiﬁer on the three datasets attached (the datasets are sentiments; binary labels). Study the impact of n on the performance of the classiﬁer (more details are below). Details. Evaluate your classiﬁer on the three datasets, using the features extract with the n-gram model and 10-fold cross-validation. In reporting the results, calculate 7 metrics: false positive rate, false negative rate, true positive rate, true negative rate, accuracy, precision and recall, and report them in that order for each experiment. Report the results for each dataset, and for n = 1...5, combined and separated. That is, report the results for features representation of f1, f2, ..., f5 (5 experiments for each dataset; in total you’ll have 15 experiments), and [f1f2], [f1f2f3], [f1f2f3f4], [f1f2f3f4f5] (4 experiments for each dataset; in total you’ll have 12 experiments). Provide those 27 tables in your answer sheet.

HelenaJulie / n-gram-from-scratch--Classification-of-positive-and-negative-reviews-nlp Public

Notifications You must be signed in to change notification settings
Fork 1
Star 2

Using python, implement n-gram approach starting from n = 1 to n = 5 for feature representation. Let f1 ...fk be the vectors representation of the n-gram features, whereby f1 is bag of words, f2 is the frequency of pairs of words, as deﬁned for the n-gram model, and so forth. Using the n-gram representations, build and test a binary classiﬁer on…

2 stars 1 fork Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
.gitignore		.gitignore
4735186_Arpudaraj_hw2sol.pdf		4735186_Arpudaraj_hw2sol.pdf
4735186_Arpudaraj_hw2sol.py		4735186_Arpudaraj_hw2sol.py
amazon_cells_labelled.txt		amazon_cells_labelled.txt
imdb_labelled.txt		imdb_labelled.txt
yelp_labelled.txt		yelp_labelled.txt