def simi(path_file_1, path_file_2):
list_A = []
list_folder = []
list_result = []
folder = glob.glob1(path_file_2, '*' + g_.gv_format_csv)
# print(folder)
for file in folder:
if file == g_.gv_profile_config + g_.gv_format_csv:
print(g_.gv_space)
elif g_.gc_folder_path + g_.gv_profile_path + file != path_file_1:
list_result.append(file)
file_content = db_.read_file(
g_.gc_folder_path + g_.gv_profile_path + file, g_.gv_null)
list_folder.append(file_content)
# print(list_folder)
else:
file_content = db_.read_file(
g_.gc_folder_path + g_.gv_profile_path + file, g_.gv_null)
list_A.append(file_content)
vectorizer = TfidfVectorizer(decode_error='ignore',
strip_accents='unicode',
stop_words='english',
ngram_range=(1, 1))
vectorizer.fit(list_folder)
vectors = vectorizer.transform(list_A).toarray()
vectors2 = vectorizer.transform(list_folder).toarray()
s = cosine_similarity(vectors, vectors2)
max_index = np.argmax(s)
# print(list_result[max_index])
return list_result[max_index]
def basic_operations_on_query(query):
query = re.sub(r"[^a-zA-Z0-9]+", ' ', query) # Removing special characters
query = query.strip() # Removing end white spaces
# Checking if users words need to be replaced with official words.
domain_synonym_corpus_list = db_.read_file(
g_.gc_folder_path + g_.gv_f_domain_syn, g_.gc_newline)
# Using domain specific keywords
for line in domain_synonym_corpus_list:
line_str = line.split(g_.gv_comma)
if line_str[0] in query:
query = query.replace(line_str[0], line_str[1] + ' ')
# Removing the stop words.
stop_words = db_.read_file(g_.gc_folder_path + g_.gv_f_stop_words,
g_.gc_newline)
word_list_query = query.split(g_.gv_space)
for word in word_list_query:
if word in stop_words:
word = g_.gv_space + word + g_.gv_space
query = query.replace(word, g_.gv_space)
query = g_.gv_space.join(query.split())
return query
def user_exists(username):
all_content = db_.read_file(
g_.gc_folder_path + g_.gv_profile_path + g_.gv_profile_config +
g_.gv_format_csv, g_.gc_newline)
for line in all_content:
if username == line.split(g_.gv_comma)[0]:
return 0 # Exists
return 1
def get_text_elements(gv_language):
if gv_language != g_.gv_english and gv_language != g_.gv_german:
gv_language = g_.gv_english
gl_texts_dump = db_.read_file(g_.gc_folder_path + g_.gv_text_elements,
g_.gc_newline) # DB
gl_texts = []
if gl_texts_dump != g_.gv_null:
for text_element in gl_texts_dump:
elements = text_element.split(g_.gv_comma)
if elements[1] == gv_language:
if elements[2] == g_.gv_yes_en:
g_.gv_yes = elements[3]
if elements[2] == g_.gc_no_en:
g_.gv_no = elements[3]
gl_texts.append(elements[2] + g_.gv_comma + elements[3])
elements.clear()
#gl_texts.sort() #- is sorting required?
return gl_texts
def get_label(query):
# Training the model.
vectorizer = CountVectorizer(decode_error='ignore',
strip_accents='unicode',
stop_words='english',
ngram_range=(2, 3))
list_of_queries = []
list_of_labels = []
list_new_query = []
one_question = db_.read_file(g_.gc_folder_path + g_.gv_f_queries,
g_.gc_newline)
for question in one_question:
line_format = question.split(g_.gv_comma)
try:
question_text = line_format[1]
list_of_queries.append(question_text)
list_of_labels.append(line_format[2])
except:
ui_.sys_response(lv_null)
list_new_query.append(query)
len_list_of_queries = len(list_of_queries)
vectorizer.fit(list_of_queries)
vectors = vectorizer.transform(list_of_queries)
vectors_test = vectorizer.transform(list_new_query)
vectorised = vectors.toarray()
vectorised_test = vectors_test.toarray()
classifier = MultinomialNB()
classifier.fit(vectorised, list_of_labels)
classifier.predict(vectorised_test)
label = classifier.predict(vectorised_test)[0]
return label
def make_profile(matr):
lv_text_elements_profile = 'profile_elements'
lv_profile_pref = 'profile_'
lv_chat_file_pref = 'chat_rec_'
lv_yes = ',yes'
lv_q_username = '******'
lv_q_chat_rec = 'chat_rec'
lv_q_help_others = 'help_oth'
lv_q_allow_chat = 'allow_chat'
all_rel_questions = []
profile_inf = []
# Getting the Questions from the table.
all_questions = db_.read_file(g_.gc_folder_path + g_.gv_text_elements,
g_.gc_newline)
all_questions.sort()
# Finding the Questions to ask.
for sentence in all_questions:
sentence_str = sentence.split(g_.gv_comma)
if sentence_str[0] == lv_text_elements_profile and sentence_str[
1] == g_.gv_language:
all_rel_questions.append(sentence_str[2] + g_.gv_comma +
sentence_str[3])
#Creates the file for profile.
for q in all_rel_questions:
question_elements = q.split(g_.gv_comma)
field_name = question_elements[0]
field_val = ui_.user_input(field_name)
db_.write_file(
g_.gc_folder_path + g_.gv_profile_path + lv_profile_pref + matr +
g_.gv_format_csv,
field_name + g_.gv_comma + field_val + g_.gc_newline)
profile_inf.append(field_name + g_.gv_comma + field_val)
# Update Master Profile Config
update_info_profile_config = ''
for profile_element in profile_inf:
profile_element_arr = profile_element.split(',')
if profile_element_arr[0] == lv_q_username:
update_info_profile_config = update_info_profile_config + g_.gv_comma + profile_element_arr[
1] # Matr
update_info_profile_config = update_info_profile_config + lv_yes # ProfileActive: Yes
elif profile_element_arr[0] == lv_q_chat_rec:
update_info_profile_config = update_info_profile_config + g_.gv_comma + profile_element_arr[
1]
elif profile_element_arr[0] == lv_q_help_others:
update_info_profile_config = update_info_profile_config + g_.gv_comma + profile_element_arr[1] \
+ lv_yes + lv_yes + lv_yes
elif profile_element_arr[0] == lv_q_allow_chat:
update_info_profile_config = update_info_profile_config + g_.gv_comma + \
profile_element_arr[1] + lv_yes + lv_yes + lv_yes + lv_yes
print(update_info_profile_config)
db_.write_file(
g_.gc_folder_path + g_.gv_profile_path + g_.gv_profile_config +
g_.gv_format_csv, g_.gc_newline + update_info_profile_config)
db_.write_file(
g_.gc_folder_path + g_.gv_chat_path + lv_chat_file_pref + matr +
g_.gv_format_csv, 'chat record')
return matr