Inspiration
Our project was inspired by Google’s Natural Language API, particularly its sentimental analysis. Sentimental analysis allows a software to tell how positive or negative a string of text is. We wanted to see if we could write software to work in the opposite direction. Could we provide some text and get the computer to output a modified version that sounded happier? We accomplished this by searching through a list of synonyms for random words and choosing a replacement that matched the desired sentiment.
What it does
Our program takes a string of text and modifies words in it to make the sentiment of the text more positive or negative.
How we built it
We built our project in Python, using PyDictionary and Google's Natural Language API.
Challenges we ran into
Initially, we were unable to change the sentiment of the text, while keeping the meaning the same.
Accomplishments that we're proud of
We are proud of making it work in the desired way, because, for a long time, the text sounded like gibberish.
What we learned
We learned that it is hard to get a software to understand the context of a word. We learned to work in a team and how to use the Google API.
What's next for Text Sentiment Modifier
Right now, we have a restricted set of words that we can actually modify the sentiment of. We could expand this set of words in the future.
Built With
- natural-language-api
- natural-language-processing
- pydictionary
- python
- python-package-index
Log in or sign up for Devpost to join the conversation.