I was trying to find, for purely selfish marketing reasons, what terms are most popular in the world of chatbots. I was trying to “measure the zeitgeist”, so to speak. 🙂
There was only one small problem. I had no idea how to do it.
Turns out, I didn’t really have to look very far.
Chatbots Magazine is one of the more popular blogs about chatbots. They recently published an article about the top 100 articles on chatbot magazine.
So I thought this would be a good way to measure the chatbot zeitgeist. Scientific or unscientific, it is certainly an interesting experiment. 🙂
It is quite manual. I simply copied and pasted the titles of the top 100 articles into a text file.
Then I headed over to WordItOut, where you can create nice looking tag clouds.
When I pasted the entire list of article titles and generated a tag cloud, this is what I got:
You can see that the following words dominate the tag cloud without giving as much insight: chatbots, chatbot, bots, bot. After all, what else would you expect?
Excluding the obvious words
But luckily, the WordItOut website also provides a way to exclude some words (and the UI for doing it is pretty nice too), so I then excluded those top 4 words.
This is what the tag cloud looks like after the change.
The second tag cloud is much better because it gives us a lot more insight. For one thing, it makes the image a lot easier to read.
This is not very surprising. Facebook did popularize the recent craze for chatbots.
People are both fascinated and afraid of AI. Don’t believe me? Notice that “kill” is on the top right quadrant. 🙂
My personal view? You don’t have much to fear. For a long while. Meet the job taker:
Build and Building
Well, the website is called chatbots magazine. So yes, build/building is certainly going to be in that list. Here are some articles from that list with build/building in the title:
Build Chatbots Using Node.js in Motion AI
So You Want To Build A Chat Bot?
How To Build And Launch A Chatbot Over The Weekend
Building a Facebook Messenger Bot Within a Week
Delbot: Building An NLP-based, Voice-driven Bot From Scratch in Python
19 Best UX Practices for Building Chatbots
Now this one is important.
People are clearly thinking about not just functional but also conversational chatbots. Turns out, the more conversational you want to make your bot, the more things you need to think about.
You also see design/designing make an appearance. It is similar to build/building, so makes complete sense.
The words “machine” “learning” and “intelligent” make an appearance, as expected.
Notice that because the sample size is pretty small in a sense (there are only 100 sentences, and we are trying to get the top 50 frequent terms), many words make an appearance even though they only appear twice.
Well, I told you it wasn’t super scientific!
The one big disappointment for me is that I didn’t find any surprises. Did you?