Prettygraph is a Python-based Web application developed by @yoheinakajima that demonstrates the generation of text to knowledge graphs
New UI model. The project is a quick crack project, not meant to be a powerful framework, but a simple UI idea for dynamically highlighting text input when generating a knowledge graph.
The application uses Flask as the backend, rendering templates as the front-end, and LiteLLM to generate predictions that convert text input to graphical data in JSON format. The data was then visualized using Cytoscope.js. It is worth noting that the chart is regenerated each time a period is inserted and updated in real time in the UI, providing an interactive experience.
features
Text-to-graph generation: Convert the text entered by the user into a knowledge graph.
Dynamic UI updates: The graphics update with each text input that ends in a period.
Color-coded visualization: Nodes and edges in the graph are color-coded for better visual differentiation.
future improvements
Improved tracking: Enhance the logic of tracking nodes and edges through their flow in text, moving from simple keyword matching to more complex relational mappings.
Manual graphic updates: Introduce a manual update button to generate graphics as needed, reducing computing load and API costs.
Load animations: Add load animations during graphics generation to improve the user experience by indicating that processing is in progress.
Incremental graph construction: Similar to Mindgraph, you modify existing graphs by adding new elements and removing duplicate nodes, rather than recreating the entire graph using each input.
If you want to learn more, you can click on the link below the video.
Thank you for watching this video. If you like it, please subscribe and like it. thank
Github：http://github.com/yoheinakajima/prettygraph
Oil tubing: