Page 122 - Ai_C10_Flipbook
P. 122

The following is an example of the user interaction:
                     ✶ User: "How can I check my account balance?"

                    Chatbot: "To check your account balance, you can use our mobile banking app or log in to your online
                    banking account. Would you like me to guide you to the login page?"

                    o  Option 1: Yes       o  Option 2: No
                     ✶ User: "How do I reset my password?"
                      Chatbot: "You can reset your password by clicking the 'Forgot Password' link on the login page. Let me
                    provide you the steps:
                    o  Click on "Forgot Password."

                    o  Enter your registered email or phone number.
                    o  Follow the instructions sent to your email/phone.
                    Would you like to visit the login page now?"


                    o Option 1: Yes        o    Option 2: No
                     ✶ User: "Where is the nearest ATM?"
                    Chatbot: "Please share your current location or city name, and I'll find the nearest ATM for you."
                    User: "I'm in New Delhi."

                    Chatbot: "The nearest ATM to your location is at [ATM Address]. Would you like directions?"
                    o Option 1: Yes        o    Option 2: No

                     ✶ User: "What are your customer care hours?"

                      Chatbot: "Our customer care team is available from 9:00 AM to 6:00 PM, Monday to Saturday. For urgent
                    issues, you can call our helpline at [contact number]."
                     While rule-based approach as seen in the above chatbot is effective for simple, repetitive tasks, they are
                    limited in scope and flexibility. Once trained, rule-based models cannot learn or improve over time. They
                    require manual updates to handle new types of queries or information. It cannot be used to personalise

                    responses based on user history or preferences. Hence, learning-based models are introduced.
              Learning-based Approach


              This approach refers to the model where the relationship or patterns in the data are not defined by the developer.
              Random data is fed into the machine and the machine develops its own patterns or trends based on data outputs.
              It is an alternative method to address some of the challenges of rule-based systems.

              In a learning-based approach, the machine analyses data to identify patterns and make predictions. With unlabelled
              data, it clusters similar points based on shared features. For labelled data, it learns from past examples to map
              inputs to outputs. This approach enables data-driven decision-making by extracting meaningful insights.
              For example, suppose you have a dataset of some images of balls. Now you do not have any clue as to what trend
              is being followed in this dataset as you don’t know their names, colour or any other feature. Thus, you would put

              this into a learning approach-based AI machine and the machine would come up with various patterns it has
              observed in the features of these images. It might cluster the data on the basis of colour, size, shape, etc. It might
              also come up with some very unusual clustering algorithm which you might not have even thought of.




                    120     Artificial Intelligence Play (Ver 1.0)-X
   117   118   119   120   121   122   123   124   125   126   127