Chatbots caught my attention because over the last few years, there has been an exponential growth of tools to design, mock, build, deploy, manage, and monetize chatbots. I can say with some certainty, that nowadays almost everyone, whether consciously or unconsciously, has been in contact with at least one of them.
A chatbot (“chat robot”) is a computer program that simulates human conversation, either via voice or text communication, using a wide variety of input methods like voice, text or even touch.
Even though the term ‘ChatterBot’ was coined by Michael Mauld in 1994, the first chatbot (ELIZA) was developed in 1966. It was meaning to trick users by making them believe that they were having a conversation with a therapist, who was aimed at asking open-ended questions or responding with follow-ups. In particular, ELIZA recognizes key words or phrases from the input and gives a pre-programmed response using those keywords. For instance, if a person says that ‘My mother cooks good food’. ELIZA would pick up the word ‘mother’ and respond by asking an open-ended question ‘Tell me more about your family’. This created an illusion of understanding and having an interaction with a real human being though the process was a mechanized one.
After ELIZA there has been a wide variety of succesful chatbots, with an increasingly “human component”. In the following I will list the main pillars in the chatbot history:
Parry in 1972 originated as a research tool designed to simulate the thinking of a paranoid individual. During the computer science meeting in 1973, a hilarious conversation between ELIZA and Parry was also set up (PARRY Encounters the DOCTOR). The aim: take the bots to their logical conclusions
Jabberwacky in 1988 was designed to “simulate natural human chat in an interesting, entertaining and humorous manner”. It is an attempt to create a program that learns; it retains all conversations and finds appropriate responses by matching patterns in their context Dr. Sbaitso in 1991 was a computerized psychologist chatbot with a digital voice designed to speak to humans.
ALICE (Artificial Linguistic Internet Computer Entity) in 1995 was inspired by ELIZA, but it used a natural language processing which allowed for more sophisticated conversation. Moreover it was revolutionary, for being open-source and developed with AIML (artificial intelligence markup language) which gave developers the possibility of creating their own chatbots powered by ALICE.
Mitsuku in 2005 was created from AIML technology. It claims to be an 18-year-old female chatbot from England and contains all of Alice’s AIML files, with many additions from the user conversations, and is always a work in progress. Her intelligence includes the ability to reason with specific objects. For example, if someone asks “Can you eat a house?”, Mitsuku looks up the properties for “house”, it finds the value of “made from” is set to “brick” and replies “no”, as a house is not edible. She can play games and do magic tricks at the user’s request.
IBM Watson in 2006 was originally developed to compete on the American TV game show, ‘Jeopardy!’ where it defeated two of the former champions in 2011. Nowadays it has been used in many fields such as decision maker in cancer treatment in Memorial Sloan-Kettering Cancer Center.
Siri in 2010 was launched as a new iPhone app and it brought voice applications into the mainstream consumer market.
Alexa in 2015 caught consumers imagination and launched the market for smart home speakers. It is able to answer in an always less artificial way and to reply to increasingly complex questions.
Facebook Chatbots in 2016 with Facebook’s launch of their messaging platform, they became the leading program for chatbots. In 2018 there were more than 300,000 active chatbots on Facebook’s Messenger platform.
#data-science #natural-language-process #neural-networks #chatbots