def main_process(input_file): sentence_count = 0 gen_questions = [] final_questions = [] with open(input_file, "r") as f: for line in f: sentences = ss.sentence_segment(line, tri_gram=False) sentence_count += 1 for i in range(len(sentences)): questions = qg.generate(sentences[i].pos) for tp in questions: (question, answer, answer_index) = tp ans_eng, ans_index = wnth.get_general_info(answer) answer_item = _wi.word_item(answer, ans_eng, ans_index, None) question_item = _qui.question_item(sentences[i], sentence_count, question, answer_item, answer_index) gen_questions.append(question_item) ranked_questions, qrank_scores = qr.rank_question(gen_questions) for question in ranked_questions: choices = chg.generate_choices(question) if choices != []: question.add_choices(choices) final_questions.append(question) return final_questions
def generate_choices(self, question_item): answer = question_item.answer if self.is_thai_word(answer.word): siblings, eng_siblings, index, hypernyms = self.wn.get_siblings(answer.eng_word, answer.index) choices_set = set() all_choices = [] for i in range(len(siblings)): for j in range(len(siblings[i])): new_item = word_item.word_item(siblings[i][j], eng_siblings[i][j], index[i][j], hypernyms[i]) sibling_thai = siblings[i][j] if sibling_thai not in choices_set and self.is_thai_word(sibling_thai): choices_set.add(sibling_thai) all_choices.append(new_item) if len(all_choices) >= 3: ranked_choices, choices_with_score = self.choice_ranker.rank_choices(question_item, all_choices) choices = [question_item.answer] + ranked_choices[0:3] else: return [] # choices = self.rank_choices(all_choices, answer) else: return [] # random.shuffle(choices) return choices
def __init__(self, *args, **kwargs): if len(args) == 5: (sentence, sentence_no, question, answer, answer_index) = args self.sentence = sentence self.sentence_no = sentence_no self.question = question self.answer = answer self.answer_index = answer_index self.choices = None self.asked_choices = None elif "from_str" in kwargs: attributes = ast.literal_eval(kwargs["from_str"]) for key in attributes: if key == "sentence": self.sentence = _sentence.sentence(from_str=attributes["sentence"]) elif key == "choices": self.choices = [_word_item.word_item(from_str=choice_str) for choice_str in attributes["choices"]] elif key == "answer": self.answer = _word_item.word_item(from_str=attributes["answer"]) else: setattr(self, key, attributes[key]) self.evals = []
def get_question_item(question_file, pos_file):
sentences = []
sentence_with_pos = []
with open(pos_file) as f:
for line in f:
read_list = ast.literal_eval(line.strip())
for a_sentence in read_list:
sentences.append("".join([word for (word, _) in a_sentence]))
sentence_with_pos.append(a_sentence)
sentence_count = 0
all_question_items = []
with open(question_file) as f:
for line in f:
read_list = ast.literal_eval(line.strip())
for a_sentence in read_list:
sentence_item = sentence.sentence(sentences[sentence_count], sentence_with_pos[sentence_count])
for a_question in a_sentence:
(question_sentence, answer, choices) = a_question
answer_item = word_item.word_item(answer)
answer_index = find_blank_index([tp[0] for tp in sentence_with_pos[sentence_count]], question_sentence)
all_generated_choices = _cg.choice_generate(answer_item)
choice_items = []
for a_choice in choices:
if str(a_choice) == answer:
choice_items.append(answer_item)
else:
for gen_choice in all_generated_choices:
if str(gen_choice) == a_choice:
choice_items.append(gen_choice)
break
# print([str(choice) for choice in choice_items])
question = question_item.question_item(sentence_item, sentence_count, question_sentence, answer_item, answer_index)
question.add_choices(choice_items)
all_question_items.append(question)
sentence_count += 1
return all_question_items