Page 121 - Ai_C10_Flipbook
P. 121

Modelling

                 AI Modelling refers to developing algorithms, also called models which can be trained to get intelligent outputs.
                 That is, writing codes to make a machine artificially intelligent. The model is trained using data. There are mainly
                 two types of AI models:
                                                                           AI Model




                                             Learning-based                                         Rule-based



                                      Machine Learning                                Deep Learning




                         Supervised    Unsupervised   Reinforcement          Artificial Neural      Convolutional
                          Learning       Learning       Learning                Network             Neural Network
                 Rule-based Approach


                 The  Rule-based  Approach  is  one  of  the  earliest  and  simplest  methods  of  implementing  artificial  intelligence.
                 It relies on predefined rules and facts created by developers to enable machines to perform specific tasks and
                 generate desired outputs. Developers manually define a set of rules that determine how the machine processes
                 data and responds to various scenarios. The main drawback of this approach is that the machine's learning is static.
                 Once trained, the machine does not adapt to changes made in the original training dataset. If the machine is tested
                 on a dataset that differs from the rules and data provided during the training stage, it will fail to produce accurate
                 results and will not learn or adjust to the new conditions it encounters.

                 CASE STUDY: Banks' Chatbot

                 A bank's website features a chatbot to assist users with basic queries related to account management, such as
                 checking account balance, resetting passwords, or locating nearby ATMs.
                 1.  Data: The data required to train the chatbot is a simple scripted document based on questions and their
                    corresponding answers.
                 2.  Rules: The chatbot uses a simple decision tree approach with defined rules under the category of “Yes” or “No”
                    to complete the conversation. Some of the rules which bank's chatbots follow are:
                    a.   Rule 1: Predefined Questions and Answers. It prompts the user for “Account Balance”, if the answer is Yes, it
                       suggests a mobile app of the bank or login to an online banking account. Based on the user choice, it does
                       the needful.
                    b.   Rule 2: Keyword Matching. The chatbot recognises user input through keywords and matches them to
                       pre-set rules in its database. For example, when a user mentions "ATM," the bot asks for location details and
                       retrieves nearby ATMs from a stored database.
                    c.   Rule 3: Guided Flow Under the “Change Password” option. It will prompt the user to answer the details like
                       registered email or phone number. It will direct the user to the login page.
                    d.   Rule 4: Options and Redirection. Provides links to relevant pages (e.g., login, directions) and redirects to
                       human support for more complex issues.
                 3.   Interaction: When a user communicates with the chatbot, it processes the message by matching it with the
                    predefined rules. Depending on the situation, the chatbot replies with a ready-made response or asks for more
                    details to address the query.


                                                                          Advanced Concepts of Modeling in AI   119
   116   117   118   119   120   121   122   123   124   125   126