def similarityWords(dictOxfordNouns):
print "\n --------------------------------------------------"
print "new round"
total_precision = 0;
total_recall = 0;
total_accuracy = 0;
missingWORD = 0;
total_word = 0
count_step = 0
for word in dictOxfordNouns:
# print dictOxfordNouns[word]
# if word == 'baby':
matrix_result = similarity_by_synsets_synsets_nbest_withword_average(word,dictOxfordNouns[word]);
if matrix_result == None:
missingWORD += 1;
continue
# print "\n"
print word
(precision, recall, accuracy) = CompareWithGold.compareGoldWithResult(matrix_result,word)
# print (precision, recall, accuracy)
if precision != -1:
total_precision += precision
total_recall += recall
total_accuracy += accuracy
total_word += 1
count_step += 1
# print "result -------"
# print precision
# print recall
# print accuracy
# print "\n"
else:
missingWORD += 1
# total_word = len(dictOxfordNouns) - missingWORD
precision = total_precision/total_word
recall = total_recall/total_word
accuracy = total_accuracy/total_word
print "total:"
print total_word
print precision
print recall
print accuracy
WriteParametersAndResult.append_result_to_file(precision,recall,accuracy)
return (precision, recall, accuracy)
def train_with_random(dict_nouns):
_best_result = 0;
_best_params = Parameters.get_current_params()
for i in range(0,1000):
# random_all_feature
Parameters.random_params_values()
# calculate_result -> current best
(precision, recall, accuracy) = SimilarityWordnetOxford.similarityWords(dict_nouns)
if accuracy > _best_result:
_best_result = accuracy
_best_params = Parameters.get_current_params()
(best_result_loop, best_params_loop) = train_from_base(dict_nouns)
if best_result_loop>= _best_result:
_best_result = best_result_loop
_best_params = best_params_loop
Parameters.set_params_from_arr(_best_params)
WriteParametersAndResult.append_params_and_result_to_file(_best_params)
