Good morning, this is Iino.
The title is as it is. The long-awaited release of the user-local chatbot framework has been released for some users.
It doesn’t seem to have been officially released yet, but I actually made a bot using the framework for Ledge, so I’ll summarize my impressions and features of using it.
A chatbot creation platform that users locals are fully prepared for
The User Local Artificial Intelligence API was released a step ahead, but programming was essential to use it. From a non-engineer’s point of view, the hurdles were high.
This framework allows you to create a chatbot without programming.
I’ve been seeing services that “create a chat without programming” in Japan, but since it was released at this time, there are various functions that other services don’t have. I will briefly introduce each of them.
You can also make a bot of that character.Word ending customization function
The previously announced User Local API made it possible to convert word endings such as “~nyan” and “~wan”. You can choose from cat, dog, and old man, but with this framework, you can customize the endings as you like.
If you use this, you can also use characters with unique endings like FunassyiYou can easily turn it into a botThat’s what it means.
It also supports bots on Twitter.there may be new possibilities
The bot platform supports LINE, Facebook Messenger, and Twitter. What is interesting is that it supports Twitter.
Since it is difficult to respond in real time on Twitter, it seems that the batch will move periodically, but unlike other platforms, it has a great potential because it has the power to spread. It may be interesting to create one account for your company and try it out.
Information about users can be easily understood!
On LINE, I sent a message to the botThe other party’s ID and profile picture are available by defaultIt looks like Of course, you can easily see all the conversation logs with the bot.
From the user’s point of view, it’s a little scary, but from the bot creator’s point of view, it’s quite nice to know what kind of users are using it.
You can also manually send a message to the chat partner from the platform, so you can easily leave the conversation to the bot first and then start the chat manually from a certain point.
Also, how many people are using it? It is also possible to see it graphically.
How you use it is up to you. Knowledge database function that can easily turn web media into bots
One of the major features is that you can set up a knowledge database.
For example, if you store information about ramen shops in a database, you can create a ramen bot that will tell you the recommended ramen shops in Shibuya when you hear “ramen shops in Shibuya.”
in short,what to put in the database?can be used in various ways byabout it. If you put information about web media articles, you can create a bot for that media.
howIt also supports RSSTherefore, if you make it well, it is possible to automatically insert it into the database as soon as new information comes in. This is easy.
I actually made a Ledgebot
I made a LINEbot that distributes Ledge articles using the knowledge database function. For article registration, just enter the RSS URL of each category.
The keywords “article”, “want to know”, and “tell me” are used as hooks to fetch information from the knowledge database. So, they introduce articles such as “Articles on AI” and “I want to know about MA”.
It takes about 5 minutes to create, including linking LINE accounts. Please use it by all means. Apart from whether it can actually be used, it is interesting that web media can be turned into a bot this easily. Of course, the chat function is done with the user local API.
If you have a smartphone, you can also go here.
In order to make it work properly as a media, it seems that search accuracy will not improve unless tuning is added later, so just be careful.
Will the bot rush keep coming?
Every time a service like this comes out, I think, but if you try to do it seriously, the framework will absorb the difficult things.
“I’m an engineer”, “I’m a marketer”… The era of choosing whether or not to do something based on such titles may be over.
As Mr. Ito, the user local representative, said, the fact that anyone can create a chatbot is a big turning point.
What kind of services will be born from this chatbot framework? While looking forward to it, I would like to pay more attention to future trends.