No reviews yetBe the first to leave a review for Emoj
Cant seem to be able to make an API call. Whenever I make an API call (Ex: https://emoj.ai/api/classify?tex... , but the token is fake obv), it tells me that the site cant be reached. Seems like a correct API call to me
This could be useful for sites that want to layer in a visual representation of one's emotion. E.g. if we had this for Product Hunt comments, we could use emojis to identify positive 😊 vs. negative ☹️ reactions.
@rrhoover Agreed, that could be pretty cool! I initially started thinking about this purely in the context of bots, but I think giving any product emotional intelligence can be pretty powerful, with lots of use cases.
Let me know if you end up building that ;)
@keyul Your welcome! Let me know how it works for your use case.
Report
Can you see the big picture?
.
Humans will get more and more rude with support rep. Why? To trick the system. Cie will manage based on the fear the client quits instead of common sense.
Hi Product Hunters!
I’m excited to be back today launching Emoj, an API that lets anyone add emotional intelilgence into their product with just a few lines of code.
Understanding how someone is feeling in the moment is relatively easy when we talk to other people face to face. When we communicate via text messaging, understanding emotion can sometimes be difficult, but with emojis and other indicators, it’s getting easier. When it comes to communicating with robots, or even just software in general, it’s extremely difficult for the machine to understand us. But it doesn’t have to be this! Context is key in order to build a great bot or conversational assistant. Having knowledge about how a person is feeling is crucial to delivering personalized and humanized conversations. But it doesn’t have to be this way! Context is key in order to build a great bot or conversational assistant. Having knowledge about how a person is feeling is crucial to delivering personalized and humanized conversations.
About 8 months ago, I started working on a bot called Joy (www.hellojoy.ai). Joy tracks and measures your mental health/wellness by analyzing your conversations. I started building Emoj out of a personal need to improve Joy's capabilities. I was initially using a combination of things like IBM Watson, Microsoft Luis, and general keyword tagging, but Joy was only accurately detecting emotion around 67% of the time. Through the natural flow of Joy’s conversations, I started to collect a gold mine of data — (anonymous) text conversations labeled with an emotion as confirmed by the end user. With this dataset, I was able to train a model that is accurate over 85% of the time and is constantly improving.
I’ve been using Emoj to power Joy for the last several months and am excited to see what others will use it for! I’ll be around to answer questions and I’d love to hear any thoughts or feedback!
P.S. I wrote more on the topic here: https://medium.com/@dannyfreed/t...
P.P.S. There’s a special deal for Product Hunters 😺
@joshbarkin Sort of, but not really. Sentiment analysis is great for a lot of use cases, but it usually only classifies things into positive, negative, or neutral.
Emoj classifies things into 8 different types of buckets and provides a confidence rating for each one:
- Happiness
- Sadness
- Stressfulness
- Anxiousness
- Disgust
- Anger
- Fear
- Tiredness
So, while a sentence like "I had a long day at work" might read as neutral or negative when run through sentiment analysis, Emoj will classify a sentence like that as "Stressfulness" being present. It can do this because of the data it's trained on (which comes from http://www.hellojoy.ai)
Emoj
Product Hunt
Quick Code
Roundups
reason8
