Writing a letter and send it to someone far away is one of the most ancient ways of communication. Okay, maybe it’s not that ancient, but you get the idea. A man at war sent a love letter to his wife telling that he’s okay, and the wife replied that she missed him so much; a customer wrote a complaint in a newspaper that he/she had a bad experience when using a company’s services; a teacher sent a letter to the parent telling their son might be expelled if he cannot pass the exam.
Text contains emotions. Love, hate, anger, happiness, or even sadness.
In recent days, it’s become rare for us to send a letter, there are many better ways for us to send messages. Technology find its way to change our way to communicate with each other. In the other hand, millions of words are posted in the social media and chat apps daily, and making them a huge source of text.
It was somewhen in 2012, when I began to understood that there was something called 'bot’. Someone on Twitter replied to my tweet whenever I had an exact phrase in my tweet. I visited the profile and began to realize that it was an automatic reply. Almost all the tweets from that account were in the same format.
An example of auto-reply tweet; image from Labnol
The Journey Begin
Since that day, I began to love that kind of bots. Simply because it can ‘understand’ what you write and giving it a proper response. The simplest way to do it is find a tweet that contains an exact keyword and reply it with a predefined response. From this experience, I made a currency converter Twitter bot, @TukarUang. This bot will give you the latest exchange rate, based on Google Currency API. As a software developer, it didn’t take me much time to create this bot. And here is the example of my Twitter bot’s response.
The tweet was written in Bahasa Indonesia, and it said, “@ankszm, the exchange rate for 1000 Lebanese pounds is 6,456.3 Indonesian Rupiahs”. Pretty cool, huh? The response above was triggered when ankszm mentioned the bot using
1000LBP to IDR format. However, the bot already stopped working since 2013. I decided to shut it down since the Google API had changed, and I just didn’t want to work on it.
Still in 2013, I decided to take an MEng degree in in Universitas Gadjah Mada to have a deeper and better understanding about Information Technology. And my researches were focused on.. Text! C’mon, it’s not that hard to guess, right? During my years of study, I was more amazed knowing that there’s a field focusing on how to make the machines able to understand our language, it’s Natural Language Processing.
It’s not only
when someone says '100LBP to IDR', then calculate it, but the machine can understand when someone says
Hello bot, I want to know the current weather in New York City. Do I need to bring an umbrella? Indeed, this requires far more advanced and complex computing process to extract the information and the intent, but seeing the machine replies with
Hi, Tom! It's sunny now in New York City, and there is no need for you to bring an umbrella for the next 3 weeks! Have a good day! always give me a pleasant feeling.
That is what we call a chatbot.
Even though I didn’t conduct a chatbot research for my master thesis (instead, I conducted a research in sentiment analysis), but I’m getting more and more interested in chatbot. And since I graduated in 2016, I still following the research in this area, mainly for business purpose.
Chatbot for business? Is that a thing?
Answering this question gonna lead us to another article, but the short answer is YES. People nowadays tend to use chat apps to interact with each other rather than meet them face to face. Imagine that you are a business owner and have a chatbot implemented in your social media page, which allows people to chat and have a conversation with your chatbot, and the chatbot will lead them buying your products. Isn’t it nice?
There are so many companies that already have a chatbot in their social media pages, and the number still growing. Can you see that this is an opportunity? That is why I love chatbot more and more.
As my closing statement, both for developers and business owners, chatbot is the next big thing in industry, and it’s still not too late to dip your toes into it.
By the way, it’s just a random quote in the subtitle, and does not related to the article.