Machine Learning So Easy, Even Your Cat Could Do It (Part 1): Sentiment Analysis
September 30, 2015 / Business
I’m Julie, the Ops Manager over here at indico, and I’m pretty solidly on the non-technical side of things. I’m going to show how you, yes you, can use our APIs to bring machine learning into your life. I brought them into mine, and it was super easy. So I welcome you to a new series entitled “Wizard Hats + Cats” brought to you by little ole me.
Back to that non-technical part. Let’s just say that until recently, my expertise had nothing to do with technology. Think: real estate (EOP), accounting (Ernst & Young), and construction (Shawmut Design and Construction). The one thread that runs through all the phases of my career is my hunger for knowledge. I have always been my own IT help desk. I blogged in my dorm room in ’98 at Salve Regina and used scribble.com. I taught myself the Adobe Creative Suite. I was one of the first people to get the original iPod. I then posted on Craigslist hoping to meet my soulmate, the only requirement being that he must also have said iPod.
Sadly, there was no riding off into the sunset based solely on a consumer product. (Oh how I miss pre-Tinder society.)
When I joined indico as Operations Manager, I found a whole new area to dive into: computer science. APIs, Python, machine learning…I needed to dig in, but I didn’t know how to code. I could pretend to on HackerTyper.net — look Ma, no hands! — but let’s get serious.
Every article I read on machine learning made my head hurt (although this one made it hurt a little less). I didn’t want to know the algorithms behind the code, or parse the seemingly nonsensical stream of jargon — neural networks, Big Data (which, by the way, doesn’t exist), blah blah blah — I just wanted to know why machine learning would make my life easier. I watched the first ten minutes of Introduction to Deep Learning with Python (by our very own Head of Research, Alec Radford) before my head started spinning. Take a look at this screen grab of his code:
In times like these, the only solution is to give up and head to Lil Bub.
Whoa, what? Is that even English? I had only used basic HTML while stumbling through DreamWeaver. This was way over my head, until I discovered indico’s APIs.
That’s why I’m on a mission to tell you that indico does all the dirty work for you.
Using machine learning doesn’t have to involve hundreds of lines of crazy code, a math PhD, or a cat in a robe and a wizard hat.
All of indico’s machine learning goodness is neatly packaged in whichever API you want to run. Literally, all you need is the following for sentiment analysis:
import indicoio indicoio.config.api_key = "YOUR_API_KEY" print(indicoio.sentiment_hq("indico is so easy to use!"))
That’s THREE lines of code. Just three.
Those three lines are mostly for your software developers to get excited about. You and I can poke around at the stuff that’s been beautified for our non-coding eyes: the indico demos.
But before we head over there, let’s run through some of the ways indico’s machine learning tools are helpful for you in this series of absolutely non-technical blog posts. Spoiler: There will be cats.
Today, let’s take a look at Sentiment Analysis.
Basically, this tool tells you how positive or negative a message is. So as we all know, it’s cool with the world that I have
17 cats two cats rampaging around my apartment…but we can all agree that it’s probably not so great if one of them is plotting my demise.
But how can you use it to help your business, your image, or the product you’re trying to improve?
As our CEO, Slater Victoroff, states in his latest post, “People earning six figures a year scroll through a spreadsheet reading tweets and deciding whether they are positive or negative. For weeks at a time. This is not a joke.” Ew. Gross. You don’t need to hire someone to do that when we’ve already done it for you.
I recently met a scientist at Johns Hopkins (name drop like a boss!) who explained the grant-writing process to me. Turns out, the second a grant starts to sound a little negative, the reviewers trash it and move on. They only like positive grants. All that time spent? Gone. Using our Sentiment Analysis API can save you from rejection letters before they happen. This cat right here? Totally into it.
If you’ve got a social media profile of any sort (if you don’t, you’re obviously not reading this right now), you probably wonder how you come across to others. Are you upbeat about things or did you recently go through a rough patch? If you’re running a business, do people have good things to say about your brand or product, or is there something your customers want you to improve?
We’ve used our Sentiment Analysis API to create the Twitter Thumbprint tool, which gives you a full view of how you behave online. Take a look at your Thumbprint to gauge whether or not you need to share more funny cats videos. (It’s the key to ruling the world.)
Here’s a peek at mine:
You can also use the Sentiment Analysis API for media analysis (social, political, entertainment, etc.), financial analysis, marketing campaigns, etc. The list goes on! The main take away? Use indico to do something awesome with your data.
Don’t just write something and send it into the universe. Make sure your sentiment is on point. Go ahead — play with the demo and see for yourself.
For a more in-depth look at sentiment analysis, head on back to our CEO’s article “What Even is Sentiment Analysis?”. It’s a truly great read.
ps: I just ran this blog post through our Sentiment_HQ model. This post was brought to you by the number 0.9348799586296082. (Pretty damn positive.)