logo

NJP

Google DialogFlow | Intents | Chatbots | part 2

Import · Jan 10, 2022 · video

hi everyone this is the second video in the series of dialog flow uh in a previous session we discussed about the first and the very basic step of about creating agents for a particular chatbot we created an agent named agent snow which you can also see on the left top corner and when we clicked on that create button the default screen which comes in front of us is the intent screen which you can also see in front of you right now so what are basically intense intents are actions that can be performed by the backend logic or maybe the backend code and these are basically configured by developers like us and used to determine what action is to be taken by the code in simple language i can say it as intent maps what a user sees and what your software should execute so it works on the basis of what expressions or what content the user has to present and on the basis of that it corresponds with a suitable answer which is coded in the backend so there are two kinds of intents which are available to us by default which you can also see in the screen one is the welcome intent which is partially configured out of the box and the second one is fallback intent which is completely and fully configured out of the box so let's talk about the fallback intent first when we talk about fallback intent we don't need to specify a user expression for this because this intent comes into picture when none of the other indent matches the user expression for example uh let's say there is a conversation going on with the bot uh and you're talking about one of your orders which is not delivered yet and in the middle of that you ask a question about some xyz store which the bot is not aware about in that case it triggers the default fallback intent which in other terms also means that hey sorry i don't know what you're talking about and then the fallback intent is triggered so let's talk more about the fallback intent you can see that here the context is available event text is available and certain training faces and action which gets triggered when you're talking about fallback intent is input dot unknown and these are the default responses like sorry i didn't get that sorry what was that and we can also test this by typing certain text over here so i'm going to write how are you and the default response which my fallback intent has returned is say that one more time and you can also check for the action which was triggered and that is input unknown and the query score is 0.0 so this is how a fallback intent works and we can try with couple of other texts as well see again my fallback intent is bonded with one of the default text response which i configured over here and now let's get go back and we'll talk about the welcome intent so this is partially configured as i already said and we can enter certain user expressions over here if we want to for example if hello already added yeah it is there you can add so we can keep adding the training faces so that our bot gets trained and whenever we type any of those one of these responses will be triggered so let's try with one of them i will type hello and it says good day what can i do for you today because it was able to recognize whatever i'm talking about and let's see the action it is input dot welcome and the same is being mentioned over here and let's type what i just added to it and it oh for some reason it did not recognize it because i did not save my default welcome intent let's save it and then try and i'm going to give it another try no it doesn't work because i haven't trained my bot yet to respond to whatever i'm adding a custom training face over here so we'll see that moving forward how we can add a custom one and it should still work and so these are basically the user expressions which if you remember were not part of the fallback intent because fallback intent is supposed to be the one which is without the user expressions so basically there are four sections when we talk about intents first is the expression user expression second is the action so user expression is something whatever user says for example i'm facing i'm facing a login issue i'm not able to log into my computer or how are you doing today so all these are the expressions now second section is the action now whatever action is taken on the basis of the user expression and third section is the response which is supposed to be sent back to the user when user says something and an action is taken in the back end and on the basis of that action a response is returned to the user and the fourth section is the context and context specifies the environment under which my intent operates okay and if required we can also create our own intent these out of the box are simply for giving us an idea of how it works but whenever we are talking about a chatbot to be used in our organization definitely there are thousands of use cases which come into picture and we'll be also taking one of those into account once we plan to include a demo for the integration between let's say dialog flow to service now so we'll talk about it more in the upcoming sessions thank you so much for watching

View original source

https://www.youtube.com/watch?v=o5V8Bu4kRvw