chatbot使用_如何使用Python构建Chatbot项目

pipinstall tensorflow keras pickle nltk 

import numpy as np
from keras.models import Sequential
from keras.layers import Dense, Activation, Dropout
from keras.optimizers import SGD
import random
import nltk
from nltk.stem import WordNetLemmatizer
lemmatizer &＃61; WordNetLemmatizer()
import json
import pickle
intents_file &＃61; open( &＃39;intents.json&＃39; ).read()
intents &＃61; json.loads(intents_file)

words&＃61;[]
classes &＃61; []
documents &＃61; []
ignore_letters &＃61; [&＃39;!&＃39; , &＃39;?&＃39; , &＃39;,&＃39; , &＃39;.&＃39; ]
for intent in intents[ &＃39;intents&＃39; ]:for pattern in intent[ &＃39;patterns&＃39; ]:#tokenize each wordword &＃61; nltk.word_tokenize(pattern)words.extend(word)#add documents in the corpusdocuments.append((word, intent[ &＃39;tag&＃39; ]))# add to our classes listif intent[ &＃39;tag&＃39; ] not in classes:classes.append(intent[ &＃39;tag&＃39; ])
print(documents)

# lemmaztize and lower each word and remove duplicates
words &＃61; [lemmatizer.lemmatize(w.lower()) for w in words if w not in ignore_letters]
words &＃61; sorted(list(set(words)))
# sort classes
classes &＃61; sorted(list(set(classes)))
# documents &＃61; combination between patterns and intents
print (len(documents), "documents" )
# classes &＃61; intents
print (len(classes), "classes" , classes)
# words &＃61; all words, vocabulary
print (len(words), "unique lemmatized words" , words)
pickle.dump(words,open( &＃39;words.pkl&＃39; , &＃39;wb&＃39; ))
pickle.dump(classes,open( &＃39;classes.pkl&＃39; , &＃39;wb&＃39; ))

# create the training data
training &＃61; []
# create empty array for the output
output_empty &＃61; [ 0 ] * len(classes)
# training set, bag of words for every sentence
for doc in documents:# initializing bag of wordsbag &＃61; []# list of tokenized words for the patternword_patterns &＃61; doc[ 0 ]# lemmatize each word - create base word, in attempt to represent related wordsword_patterns &＃61; [lemmatizer.lemmatize(word.lower()) for word in word_patterns]# create the bag of words array with 1, if word is found in current patternfor word in words:bag.append( 1 ) if word in word_patterns else bag.append( 0 )# output is a &＃39;0&＃39; for each tag and &＃39;1&＃39; for current tag (for each pattern)output_row &＃61; list(output_empty)output_row[classes.index(doc[ 1 ])] &＃61; 1training.append([bag, output_row])
# shuffle the features and make numpy array
random.shuffle(training)
training &＃61; np.array(training)
# create training and testing lists. X - patterns, Y - intents
train_x &＃61; list(training[:, 0 ])
train_y &＃61; list(training[:, 1 ])
print( "Training data is created" )

# deep neural networds model
model &＃61; Sequential()
model.add(Dense( 128 , input_shape&＃61;(len(train_x[ 0 ]),), activation&＃61; &＃39;relu&＃39; ))
model.add(Dropout( 0.5 ))
model.add(Dense( 64 , activation&＃61; &＃39;relu&＃39; ))
model.add(Dropout( 0.5 ))
model.add(Dense(len(train_y[ 0 ]), activation&＃61; &＃39;softmax&＃39; ))
# Compiling model. SGD with Nesterov accelerated gradient gives good results for this model
sgd &＃61; SGD(lr&＃61; 0.01 , decay&＃61; 1e-6 , momentum&＃61; 0.9 , nesterov&＃61; True )
model.compile(loss&＃61; &＃39;categorical_crossentropy&＃39; , optimizer&＃61;sgd, metrics&＃61;[ &＃39;accuracy&＃39; ])
#Training and saving the model 
hist &＃61; model.fit(np.array(train_x), np.array(train_y), epochs&＃61; 200 , batch_size&＃61; 5 , verbose&＃61; 1 )
model.save( &＃39;chatbot_model.h5&＃39; , hist)
print( "model is created" )

import nltk
from nltk.stem import WordNetLemmatizer
lemmatizer &＃61; WordNetLemmatizer()
import pickle
import numpy as np
from keras.models import load_model
model &＃61; load_model( &＃39;chatbot_model.h5&＃39; )
import json
import random
intents &＃61; json.loads(open( &＃39;intents.json&＃39; ).read())
words &＃61; pickle.load(open( &＃39;words.pkl&＃39; , &＃39;rb&＃39; ))
classes &＃61; pickle.load(open( &＃39;classes.pkl&＃39; , &＃39;rb&＃39; ))
def clean_up_sentence (sentence) :# tokenize the pattern - splitting words into arraysentence_words &＃61; nltk.word_tokenize(sentence)# stemming every word - reducing to base formsentence_words &＃61; [lemmatizer.lemmatize(word.lower()) for word in sentence_words]return sentence_words
# return bag of words array: 0 or 1 for words that exist in sentence
def bag_of_words (sentence, words, show_details&＃61;True) :# tokenizing patternssentence_words &＃61; clean_up_sentence(sentence)# bag of words - vocabulary matrixbag &＃61; [ 0 ]*len(words)  for s in sentence_words:for i,word in enumerate(words):if word &＃61;&＃61; s: # assign 1 if current word is in the vocabulary positionbag[i] &＃61; 1if show_details:print ( "found in bag: %s" % word)return (np.array(bag))
def predict_class (sentence) :# filter below  threshold predictionsp &＃61; bag_of_words(sentence, words,show_details&＃61; False )res &＃61; model.predict(np.array([p]))[ 0 ]ERROR_THRESHOLD &＃61; 0.25results &＃61; [[i,r] for i,r in enumerate(res) if r>ERROR_THRESHOLD]# sorting strength probabilityresults.sort(key&＃61; lambda x: x[ 1 ], reverse&＃61; True )return_list &＃61; []for r in results:return_list.append({ "intent" : classes[r[ 0 ]], "probability" : str(r[ 1 ])})return return_list
def getResponse (ints, intents_json) :tag &＃61; ints[ 0 ][ &＃39;intent&＃39; ]list_of_intents &＃61; intents_json[ &＃39;intents&＃39; ]for i in list_of_intents:if (i[ &＃39;tag&＃39; ]&＃61;&＃61; tag):result &＃61; random.choice(i[ &＃39;responses&＃39; ])breakreturn result
#Creating tkinter GUI
import tkinter
from tkinter import *
def send () :msg &＃61; EntryBox.get( "1.0" , &＃39;end-1c&＃39; ).strip()EntryBox.delete( "0.0" ,END)if msg !&＃61; &＃39;&＃39; :ChatBox.config(state&＃61;NORMAL)ChatBox.insert(END, "You: " &＃43; msg &＃43; &＃39;\n\n&＃39; )ChatBox.config(foreground&＃61; "#446665" , font&＃61;( "Verdana" , 12 ))ints &＃61; predict_class(msg)res &＃61; getResponse(ints, intents)ChatBox.insert(END, "Bot: " &＃43; res &＃43; &＃39;\n\n&＃39; )ChatBox.config(state&＃61;DISABLED)ChatBox.yview(END)
root &＃61; Tk()
root.title( "Chatbot" )
root.geometry( "400x500" )
root.resizable(width&＃61;FALSE, height&＃61;FALSE)
#Create Chat window
ChatBox &＃61; Text(root, bd&＃61; 0 , bg&＃61; "white" , height&＃61; "8" , width&＃61; "50" , font&＃61; "Arial" ,)
ChatBox.config(state&＃61;DISABLED)
#Bind scrollbar to Chat window
scrollbar &＃61; Scrollbar(root, command&＃61;ChatBox.yview, cursor&＃61; "heart" )
ChatBox[ &＃39;yscrollcommand&＃39; ] &＃61; scrollbar.set
#Create Button to send message
SendButton &＃61; Button(root, font&＃61;( "Verdana" , 12 , &＃39;bold&＃39; ), text&＃61; "Send" , width&＃61; "12" , height&＃61; 5 ,bd&＃61; 0 , bg&＃61; "#f9a602" , activebackground&＃61; "#3c9d9b" ,fg&＃61; &＃39;#000000&＃39; ,command&＃61; send )
#Create the box to enter message
EntryBox &＃61; Text(root, bd&＃61; 0 , bg&＃61; "white" ,width&＃61; "29" , height&＃61; "5" , font&＃61; "Arial" )
#EntryBox.bind("", send)
#Place all components on the screen
scrollbar.place(x&＃61; 376 ,y&＃61; 6 , height&＃61; 386 )
ChatBox.place(x&＃61; 6 ,y&＃61; 6 , height&＃61; 386 , width&＃61; 370 )
EntryBox.place(x&＃61; 128 , y&＃61; 401 , height&＃61; 90 , width&＃61; 265 )
SendButton.place(x&＃61; 6 , y&＃61; 401 , height&＃61; 90 )
root.mainloop()

python train_chatbot. py