Inspiration
- We got the inspiration from photo album's function of storing people's memory and how museums on a website can combine nostalgia and the educational wisdom from the past.
- We also came up with the idea of using logic and graph theory to construct the connections between historical tales from a csc240 assignment, which uses coloring functions and logical statements to represent different graphs.
What it does
Our website in the form of a photo album, where visitors type in the key words and will receive related photos. Most importantly, we also provide a platform for visitors to type in their own stories, and they will receive a graph representing the logic within the story and some similar stories telling how people in the past deal with the situation or the possible trend of an event.
How we built it
- To build the web site, we use html and css to write the front-end of the project.
- For back-end, we firstly use cohere generate to derive key words and connections from the text, thus constructing a graph for the visitor. We use the same approach to deal with database of historical tales stored in the museum. After that, we convert abstract graphs into matrices. Finally, we compare the visitor's graph with all the graphs( using matrices to compute), and provide the most inspirational stories to the visitor.
Challenges we ran into
- Initially, we had trouble extracting logical connections from text file. We first came up with the idea of using predicate logic P(x) to represent people and events, but found it hard to formalize in the code. But we end up using nodes to solve the problem.
- We also had trouble installing cohere package and finally tried different system environment to figure it out.
Accomplishments that we're proud of
- Our website boasts a unique and elegant design, enveloping both us and our users in an atmosphere of mystery and nostalgia.
- We take pride in bridging ancient wisdom with modern-day life, providing an experience that is both educational and inspiring.
- By using graphs and matrices to analyze language and text information, we optimized the traditional way of focusing solely on the semantics of individual words. This innovative approach allows for a deeper understanding of textual relationships.
- Our project have other applications on history learning and essay reasoning, because it generate historical narratives as examples for reasoning. It can improve essay argument quality and provides relevant historical instances.
What we learned
We learned how to construct a graph from a text by using APIs. Knowing more technologies ( such as natural language processing) and being able to utilize them is beneficial for future study and research.
What's next for Visualizing the Past Wisdom
The graph could be even more specific, by adding value edges to represent how strong the connections are between different nodes(people and events). We also wish to promote the algorithm by coloring the edges, so that we could specify which type of logical connections are for each edges. Specifying the graph enables us to obtain more information from the text, and thus better comparing and learning from the past.
Built With
- cohere
- css3
- html5
- javascript
- python
- pythonlib
Log in or sign up for Devpost to join the conversation.