pScaffoldNetwork
Experiments in building a [Scaffolding Network
In pLinkSuggest we did find common substrings between a "given sentence" and a "large number of pre-parsed documents" This means that in Scrapbox, given a new page, it is equivalent to finding links between that page and other pages
On the other hand, it could also be used as a fuzzy cross search within the collection, and I left it wondering which way to go.
The hypothesis is that it is better to create a "scaffolding network" for "a set of pages that do not yet have links" rather than to find "links between one page and a set of other pages".
pScaffoldNetwork 2022-04-01
A prototype was created.
Import an average of 1000 characters and 300 articles as plain text input.
does not look bad
This is not good.
Because they don't understand "links that already exist."
Chapter headings, etc., added by hand for navigation, are naturally substrings of the carryout, making them links within links.
Need process to combine link notation into one token
Creating a scaffolding network from books, done!
2022-04-25
Using MeCab's Constrained Analysis
Created a development environment for Deno.
I was able to parse the Scrapbox notation and output the format for MeCab's constrained analysis
Scrapbox notation combined into one token
Flagged not to include in keywords.
The original data contains nearly identical sentences in different versions, and the long matches are treated as keywords.
Ability to show what will be updated in a dry-run instead of a sudden destructive update (in progress).
TODO Support for block notation currently ignored
TODO notation sway absorption is quite intense and identical, so it would be more satisfactory to loosen it a little more.
memo
---
This page is auto-translated from /nishio/pScaffoldNetwork using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.