def test_run(data_type, lower_and_remove_punctuation, remove_stop_words, distance_method):
"""
Performs a test run, according to the given parameters
:param data_type: Defines how to store the sentences, expects: 'boolean' / 'tf' / 'tfidf'
:param lower_and_remove_punctuation: bool, if true turns all words to lower case and removes punctuation
:param remove_stop_words: bool, if true removes all stop words
:param distance_method: defines how to calculate distance, expects: 'euclidean' / 'cosine'
:return: accuracy, the accuracy of the test run
"""
file_name = "./dataset/amazon_cells_labelled_full.txt"
train_file_name = "./dataset/amazon_cells_labelled_train.txt"
test_file_name = "./dataset/amazon_cells_labelled_test.txt"
data = FileReader(file_name, lower_and_remove_punctuation, remove_stop_words)
train_set, _ = data.build_set(data_type, train_file_name)
test_set, _ = data.build_set(data_type, test_file_name)
classifier = RocchioClassifier(train_set)
accuracy = calc_accuracy(test_set, classifier, distance_method)
return accuracy
def retrieve_best_k_related_reviews(K, query):
"""
Utilizing the Rocchio Classifier,
finds and displays the k most relevant reviews to a given query, and their score.
:param K: int, number of related documents desired
:param query: string, the query to check the reviews against.
:return:
"""
file_name = "./dataset/amazon_cells_labelled_full.txt"
data = FileReader(file_name, True, True)
tfidf_set, text_set = data.build_set('tfidf', file_name)
parsed_query = data.parse_query(query)
retrieve(K, parsed_query, tfidf_set, text_set)