Enter a seed word into the search bar to visualise its semantic neighbourhood.
Click on the nodes in the graph to expand the network with related words.
Drag to rearrange nodes.
Hi! I’m Krishiv: I’m a full-time student, part-time coder and a dedicated word-lover. In my senior year of high school, I initially developed this app to help with essay-writing (for which college applications ensured I had plenty of opportunity!) but it’s grown from there into a celebration of the dynamic, intricate living structure of language and thought itself. For insofar as language corresponds to ideas, words map the nuances and contours of human creativity. What could be more exciting?
To me, part of what makes writing so thrilling - and language as a whole so fascinating - is the distinct radiance, connotation, tonality that colours every word. I believe the search for the perfect word is one best conducted through peripheries: the writer’s first mental effort takes them only so far - it’s then their joy and mission to glance carefully along the network of language before them, raise up different words to the light, exchange them and interchange them. Play with, probe, connect semantic characters until le mot juste clicks.
To use this app,
Link distances and node colours correspond to the degree of similarity calculated by the algorithm (and then mathematically normalised.)
The algorithm uses word embeddings, a natural language processing technique, that converts terms to multidimensional mathematical vectors (200 dimensional in this case). These vector representations capture latent semantic relationships between words, learnt via machine-learning algorithms that associate certain words with others, from vast amounts of text data. You’ll probably get some weird results now and then - this is just the nature of the algorithm. Laugh at them: weird results can lead to some pretty amusing rabbit holes!
If you have thoughts, questions or feedback, I'd love to hear it! Please get in touch at [email protected] as I try to further improve and develop this resource. If you find this helpful at all, you might want to show your support: Buy Me A Coffee