Inspiration
Finding a new or better word to fit a sentence or phrase is sometimes difficult, but what if you also wanted to modify an entire text? What if you're looking for some new inspiration and need a new way to brainstorm ideas and topics? Well, we found ourselves at that exact crossroad as we were discussing aquatic mammals and how to describe them. NarWord was founded on the idea that sometimes you're looking for an entirely new wording or perspective on a topic and developing a quick alternative to looking multiple terms and ideas up.
What it does
The purpose of NarWord is to take a text input of any length and return it with the adjectives, nouns, and verbs contextually replaced with the user's choice of synonyms, antonyms, or rhymes. This likely wont make the original text better, but it's a great opportunity to expand your vocabulary, discover new meanings and perspectives, or burn a couple of minutes copying and pasting wikipedia articles.
How we built it
We use Google's Natural Language API to first identify and grab the key parts of speech we want, (adjectives, nouns, and verbs) from the inputted text. Then using Datamuse's word-finding query engine we search for synonyms, antonyms or rhymes that fit the surrounding context of those word (or don't), and substitute them back into the original text. Then the user's left with a nearly useless, but "grammatically equivalent" mess of their original text!
Challenges we ran into
- Deployment on heroku was very problematic due to dependency issues.
- Google cloud API authentication was led to many problems since we were developing on different platforms.
Accomplishments that we're proud of
- We learned how to use Google's Natural Language API and figured out how to do real time word-by-word analysis.
- We worked together to create a legitimate product that anyone could publicly use for any reason and on any platform.
What we learned
- Learned how to best divide tasks amongst individuals of varying skill levels.
- How to iterate on design and methodology in order to achieve a more efficient or effective result.
- How to prioritize new ideas and their implementation in order to accomplish the most that we could in a given timeframe.
What's next for NarWord
- Make changes to the contextual aware algorithm to search for more relevant, fitting, and diverse terms for word replacement.
- Allowing users to upload documents and other files to be scanned rather than having to type or copy everything.
- Allowing users to download a document copy of their converted text in a variety of formats.
- Allowing users to save previous submissions and returns for revision and reuse.
Log in or sign up for Devpost to join the conversation.