Basic ChatBot using NLTK in Python

from google.colab import drive
drive.mount('/content/drive')

f=open('/content/drive/My Drive/chatbot.txt','r',errors = 'ignore')
raw=f.read()
raw = raw.lower()

In the above code snippets, we placed our corpus text into a text file and mounted it. Here, we're calling it chatbot.txt The f.read() would help us read the file. The raw. lower() would help us convert the text into lowercase. This is ideal as for any NLP project, we need to pre-process it to make it effective for working. Text-processing basically involves,

• Converting the whole text to either upper-case or lower-case
• Tokenization

The NLTK package contains a pre-trained Punkt tokenizer for English.

#Tokenization
sent_tokens = nltk.sent_tokenize(raw)
word_tokens = nltk.word_tokenize(raw)

In the above code snippets,

• sent_tokens converts text into a list of sentences
• word_tokens converts text into a list of words

#Pre-processing
lem = nltk.stem.WordNetLemmatizer()
def LemTokens(tokens):
    return [lem.lemmatize(token) for token in tokens]
remove_punct_dict = dict((ord(punct), None) for punct in string.punctuation)

def LemNormalize(text):
    return LemTokens(nltk.word_tokenize(text.lower().translate(remove_punct_dict)))

In the above code snippet, we define a function called LemTokens where we input tokens and return normalized tokens.

#Keyword Matching
GREETING_INPUTS = ("hello", "hi", "greetings", "sup", "what's up","hey",)
GREETING_RESPONSES = ["hi", "hey", "*nods*", "hi there", "hello", "I am glad! You are talking to me"]
def greeting(sentence):
 
    for word in sentence.split():
        if word.lower() in GREETING_INPUTS:
            return random.choice(GREETING_RESPONSES)

We can see above that for every input greeting, we have a standardized response. This forms the base for initiating conversation with our chatbot.

# Cosine Similarity
def response(user_response):
    robo_response=''
    sent_tokens.append(user_response)
    TfidfVec = TfidfVectorizer(tokenizer=LemNormalize, stop_words='english')
    tfidf = TfidfVec.fit_transform(sent_tokens)
    vals = cosine_similarity(tfidf[-1], tfidf)
    idx=vals.argsort()[0][-2]
    flat = vals.flatten()
    flat.sort()
    req_tfidf = flat[-2]
    if(req_tfidf==0):
        robo_response=robo_response+"I'm sorry! I don't understand you"
        return robo_response
    else:
        robo_response = robo_response+sent_tokens[idx]
        return robo_response

Now, let's write code for our conversation initiation and ending. This is the fun part.

lag=True
print("BOBtheRobo: My name is Bob. I'm here if you wanna chat. If you don't, type Bye!")
while(flag==True):
    user_response = input()
    user_response=user_response.lower()
    if(user_response!='bye'):
        if(user_response=='thanks' or user_response=='thank you' ):
            flag=False
            print("BOBtheRobo: You are welcome!")
        else:
            if(greeting(user_response)!=None):
                print("BOBtheRobo: "+greeting(user_response))
            else:
                print("BOBtheRobo: ",end="")
                print(response(user_response))
                sent_tokens.remove(user_response)
    else:
        flag=False
        print("ROBO: Bye! See ya later")

This way, we can build a basic chatbot and understand how NLP works

Thank you.