hatbot . A word that is coming back more and more in the modern era. And no, unfortunately we are not talking about our feline friends – but rather a very intelligent robot with which we communicate to collect information, get answers to our questions and even make purchases. In 2018, more than half of companies planned to use a chatbot, and of the rest, a quarter of the others are seriously thinking about it! Having become an increasingly widespread phenomenon, it would not be surprising if you have become acquainted with a chatbot in the last week.
Today I will try to analyze the chatbot through its functions and its future in addition to giving you some good examples of its implementation.
First, what is a chatbot?
There are several definitions of a chatbot, but the simplest would be to describe a chatbot as a conversational agent that dialogues with a user . The idea: to simulate a conversation as natural as possible so that the person chatting with the Chatbot does not feel like they are talking to a robot that provides them with automatic and predefined responses. The Chatbot must therefore analyze the messages albania telephone number data sent by the user to react appropriately according to the user’s request. Interestingly, chatbots have been among us since the 1960s, with the influence of the Turing test process, which consists of giving the illusion that a program is thought out by a sensible dialogue. But it was in 2016 that Marc Zuckerberg, CEO of Facebook, announced the launch of the Chabots platform on Messenger, the “Messenger Bot Store”.
Concretely, how does it work?
There are two types of chatbots: those that integrate artificial intelligence , and those that do not.
With AI: The smartest chatbots are the fall programming is launched also the most difficult to produce. In fact, they are created from complex algorithms and use technologies such as natural language understanding . They are able to detect the user’s intentions, going so far as to recommend specific products according to their tastes or reassure them if they detect panic. They can even learn by themselves. This phenomenon, called machine learning , refers to the way in which a computer acquires new knowledge and skills and improves its efficiency based on previously obtained results.
Without AI : Without artificial intelligence, humans must enter as much information as possible and answer several questions before receiving answers. The chatbot is similar to a basic program that searches for the answer associated with the question present in its database and returns an error if the question is not there. They are easier to implement and less expensive to produce.
Nowadays, most chatbots are used to address specific issues, or to reduce internal costs. They have very varied uses, as the following examples clearly demonstrate:
Make a reservation at a restaurant or make an appointment
Order a product and have it delivered
Recommend products, services or gifts
Answer practical questions (timetables, access, services)
User experience is crucial to the success of a chatbot. In fact, the user must feel at the center of the experience and must receive answers to their questions. It is not just about coding a technically efficient chatbot so that it pleases users.
Prominent examples
Ralph the Lego bot helps people easily find the perfect Christmas gift
The one from the American basketball business to consumer reviews league NBA , which allows you to have statistics on players, game schedules and much more.
Mastercard, which uses artificial intelligence to enable consumers to transact, manage finances and shop through messaging platforms.
Skyscanner travel comparison chatbot , which allows you to book a trip on Facebook
This is just the beginning!
We can already see this with Desjardins, which uses voice commands to enable certain actions without the need for an agent on the other end of the line, such as saying “yes”, “no” or reciting a phone number to find a customer in the database.
The chatbot of tomorrow will also have to understand the complexity of conversations. For example, we can very well imagine a scenario where we would have an interaction with a Bot through a connected device like Google Home that would continue on another text platform like Messenger and end on a computer by voice command. To achieve this, it will be essential for the chatbot to be able to interact on several different platforms, without ever losing information from previous conversations.