5 ms·

> Note the position of the point on the x/y axis is meaningless - only the proximity matters. Only the relation of the point to each other can be interpreted.

> Note the position of the point on the x/y axis is meaningless - only the proximity matters. Only the relation of the point to each other can be interpreted. I don't understand the algorithm used to interpret the relations, but judging by Marxism between positioned between Math Logic, Philosophy of Biology, Theory of Science and Recursion Theory, I doubt its usefulness.

Emma_Goldman 8y ago
Same. I don't really get it. The explanation is far too elliptical.
olooney 8y ago
UMAP[1] and t-SNE[2] are two very similar and popular ways to map a set of points in a high dimensional space to a low-dimensional (usually 2D or 3D) space by assuming all the points are close to some smooth curved surface (a manifold[3]). These are unsupervised non-linear dimensional reduction techniques; if you're familar with linear dimensional reduction techniques like PCA[4] you can think of UMAP and t-SNE as curved generalizations of that. t-SNE is widely known to be hard to interpret and unstable with respect to arbitrary hyper-parameters[5] and I would assume UMAP suffers similar problems. It's also extremely difficult to define "similarity" between documents and authors. Techniques like LSA[6] and word embeddings[7] exist but are still an open area of research. Based on similar concerns as yours, I doubt any of these issues were seriously addressed when this visualization was developed. [1]: https://arxiv.org/abs/1802.03426 https://arxiv.org/abs/1802.03426 [2]: https://en.wikipedia.org/wiki/T-distributed_stochastic_neighbor_embedding https://en.wikipedia.org/wiki/T-distributed_stochastic_neigh... [3]: https://en.wikipedia.org/wiki/Manifold https://en.wikipedia.org/wiki/Manifold [4]: https://en.wikipedia.org/wiki/Principal_component_analysis https://en.wikipedia.org/wiki/Principal_component_analysis [5]: https://stats.stackexchange.com/questions/263539/clustering-on-the-output-of-t-sne https://stats.stackexchange.com/questions/263539/clustering-... [6]: https://en.wikipedia.org/wiki/Latent_semantic_analysis https://en.wikipedia.org/wiki/Latent_semantic_analysis [7]: https://en.wikipedia.org/wiki/Word_embedding https://en.wikipedia.org/wiki/Word_embedding
claudiawerner 8y ago
Although Marxism is related to the theory of science (though not so much G.A. Cohen, who I'm not very sure why he's so "contemporary", he's not) it is positioned strangely, especially on the exact opposite side of the diagram where it seems it belongs, with continental philosophy. Also odd to see Habermas in theory of democracy so far away... and complete lack of mention of heavily cited figures in critical philosophy (Adorno, Marcuse...)

Emma_Goldman 8y ago

Same. I don't really get it. The explanation is far too elliptical.

olooney 8y ago

UMAP[1] and t-SNE[2] are two very similar and popular ways to map a set of points in a high dimensional space to a low-dimensional (usually 2D or 3D) space by assuming all the points are close to some smooth curved surface (a manifold[3]). These are unsupervised non-linear dimensional reduction techniques; if you're familar with linear dimensional reduction techniques like PCA[4] you can think of UMAP and t-SNE as curved generalizations of that. t-SNE is widely known to be hard to interpret and unstable with respect to arbitrary hyper-parameters[5] and I would assume UMAP suffers similar problems. It's also extremely difficult to define "similarity" between documents and authors. Techniques like LSA[6] and word embeddings[7] exist but are still an open area of research. Based on similar concerns as yours, I doubt any of these issues were seriously addressed when this visualization was developed. [1]: https://arxiv.org/abs/1802.03426 https://arxiv.org/abs/1802.03426 [2]: https://en.wikipedia.org/wiki/T-distributed_stochastic_neighbor_embedding https://en.wikipedia.org/wiki/T-distributed_stochastic_neigh... [3]: https://en.wikipedia.org/wiki/Manifold https://en.wikipedia.org/wiki/Manifold [4]: https://en.wikipedia.org/wiki/Principal_component_analysis https://en.wikipedia.org/wiki/Principal_component_analysis [5]: https://stats.stackexchange.com/questions/263539/clustering-on-the-output-of-t-sne https://stats.stackexchange.com/questions/263539/clustering-... [6]: https://en.wikipedia.org/wiki/Latent_semantic_analysis https://en.wikipedia.org/wiki/Latent_semantic_analysis [7]: https://en.wikipedia.org/wiki/Word_embedding https://en.wikipedia.org/wiki/Word_embedding

claudiawerner 8y ago

Although Marxism is related to the theory of science (though not so much G.A. Cohen, who I'm not very sure why he's so "contemporary", he's not) it is positioned strangely, especially on the exact opposite side of the diagram where it seems it belongs, with continental philosophy. Also odd to see Habermas in theory of democracy so far away... and complete lack of mention of heavily cited figures in critical philosophy (Adorno, Marcuse...)