Inspiration
We are currently juniors who take AP US History, and we thought it would be interesting to develop an easier way to analyze historical texts such as speeches to answer questions that cannot be easily googled.
What it does
We chose to implement a tool that specifically analyzes Presidential Inaugural speeches. Our program prompts the user to input a President, year, and question. It then uses the question as a model to analyze the text and return an appropriate answer.
How we built it
We used the nltk package in python to process and analyze the language of the questions and speeches.
Challenges we ran into
We were originally planning to link our python file with a DialogFlow agent, which would have created a Facebook Messenger interface. However, we were unable to connect our python file to the javascript files required for DialogFlow.
Accomplishments that we're proud of
In some cases, our program yielded better results than Google search.
What we learned
We were excited to learn about Conversational UI and expand our knowledge about NLP.
What's next for AskSCS: Text Analysis
In the future, we would hope to be able to connect our program to a DialogFlow agent for a more accessible interface. Also, within DialogFlow, we would be able to account for synonyms and related words to make searches more comprehensive.
Log in or sign up for Devpost to join the conversation.