Pleasantly shocked! Shouldn’t surprise me but it does. a) Took 2400+ of my Wikipedia searches from 2013-2021 (not all, just bookmarked) b) Grabbed the short Wikipedia descriptions such as: Adhesion interaction in which dissimilar surfaces or particles stick together c) Threw away the Topic word, leaving behind the meaning. d) Put ALL of the descriptions into one big text file. e) Did an AI topic classification of it, breaking it into 2 topics through 99 topics. (Used Wordstat 8) f) Took all 4700+ new topics and split up the attached keywords. g) Sorted the 2-99 Topics with split keywords so that similar categories could be grouped together. h) This left me with 248 “similar topics” from the 4700 new topics. i) I got rid of duplicates leaving a nice set of new keywords. j) Went to google Scholar and Google to find out what it thinks those words have in common. k) I end up with almost the same kinds of topics I started with but instead of 2400+ Wikipedia pages, I have 248 new topics. I only did 13 of 248 so far as this is a lot of manual work and thinking but as I have it sorted by “largest number of times keyword was used in Wikipedia short descriptions grouped by similar keywords” (it’s what I did – there’s no automatic way for that just yet), I’m pretty sure this is in priority-for-me order. That is… …it’s reflective of what’s important to me, hidden in OTHER PEOPLE’S descriptions of Wikipedia pages, chopped up and recombined. I went full circle intentionally the long crazy chopped up and recombined way but I did not expect to end up in a similar (but generalized) place that matches up so darn well.

Pleasantly shocked! Shouldn’t surprise me but it does.
a) Took 2400+ of my Wikipedia searches from 2013-2021 (not all, just bookmarked)
b) Grabbed the short Wikipedia descriptions such as:
Adhesion
interaction in which dissimilar surfaces or particles stick together
c) Threw away the Topic word, leaving behind the meaning.
d) Put ALL of the descriptions into one big text file.
e) Did an AI topic classification of it, breaking it into 2 topics through 99 topics. (Used Wordstat 8)
f) Took all 4700+ new topics and split up the attached keywords.
g) Sorted the 2-99 Topics with split keywords so that similar categories could be grouped together.
h) This left me with 248 “similar topics” from the 4700 new topics.
i) I got rid of duplicates leaving a nice set of new keywords.
j) Went to google Scholar and Google to find out what it thinks those words have in common.
k) I end up with almost the same kinds of topics I started with but instead of 2400+ Wikipedia pages, I have 248 new topics.
I only did 13 of 248 so far as this is a lot of manual work and thinking but as I have it sorted by “largest number of times keyword was used in Wikipedia short descriptions grouped by similar keywords” (it’s what I did – there’s no automatic way for that just yet), I’m pretty sure this is in priority-for-me order. That is…
…it’s reflective of what’s important to me, hidden in OTHER PEOPLE’S descriptions of Wikipedia pages, chopped up and recombined.
I went full circle intentionally the long crazy chopped up and recombined way but I did not expect to end up in a similar (but generalized) place that matches up so darn well.

Leave a comment

Your email address will not be published. Required fields are marked *


− 2 = five

Leave a Reply