EloquenC

Inspiration

Every day, 23 billion text messages are sent around the world. Unlike normal interactions, body language, eye contact, emphasis, and other indicators can not be used to understand a speaker's purpose in texts or emails. As a result, many text messages and emails often evoke a different tone than a sender intends. Thus, our group wanted to devise a way to quickly analyze the tone of writing in order to help the writer better understand how their message will be perceived by readers. Also, although there are many guides to help people with the delivery of speeches and presentation, there remains a high demand for a system that will analyze the content of a speech. Therefore, our group decided to create EloquenC to address this common problem.

What it does

EloquenC is a skill for the Amazon Alexa. Users simply open Eloquence helper and are prompted to read their writing to the Alexa. The Alexa then processes the messages and analyzes its tone by focusing on diction and syntax. Finally, the Alexa returns the tone conveyed by the message aloud.

How we built it

We built this skill by configuring a set-up where Alexa Skills Kit triggers an AWS Lambda function which proceeds to ping the IBM Watson Bluemix Tone Analyzer V3 API and parse the response to return a speech output to the Alexa API. This was built fully in Python and had .json and .txt files as refrence to create intents and sample utterances. We connected AWS and Bluemix by building the entire Tone Analyzer in the .zip file we uploaded as source code.

Challenges we ran into

The biggest challenge we ran into was creating a working AWS Lambda function. We initially attempted to code our solution in Java but the inability to debug without a complex set-up on eclipse led us to to switch to Python.

Accomplishments that we're proud of

We’re proud that we were able to configure something like this in our first hackathon. With almost no experience in making any sort of apps or using any product from Amazon Web Services, we were extremely happy to find that our product actually worked.

What we learned

We learned how to create lambda functions, program in python, use IBM Watson’s API, use git, and use the lambda functions to create an Alexa skill.

What's next for EloquenC

In these 24 hours, we were only able to figure out how to implement the tone analyzer function onto Alexa. Although this is one of the most useful features that Watson provides, there are many more opportunities for additional features and improvement with the other IBM Bluemix services.