Wikidata:Requests for permissions/Bot/The Anomebot 3 2
- The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
- Approved--Ymblanter (talk) 19:29, 19 June 2020 (UTC)[reply]
The Anomebot 3 (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: The Anome (talk • contribs • logs)
Task/s: Addition of roughly 2900 entries corresponding to the knots listed in the Knotinfo database.
Code: Python bot using the quickstatements interface.
Function details:
- Create around 2900 items based on the entries in the knotinfo database, downloaded from https://knotinfo.math.indiana.edu/index.php
- For each entry, set the following properties:
- Set the name to "X knot", where "X" is the name given in knotinfo for that knot
- Where different, set alias names that are used in other naming systems: DT name, Conway name, tetrahedral census name, etc.
- instance of (P31) knot (Q1188853)
- Alexander–Briggs notation (P6432)
- Alexander polynomial (P5350)
- Conway polynomial (P5351)
- Jones polynomial (P5352)
with all of the above having knotinfo listed as the reference.
Finally, I will manually merge some of the autogenerated entries with the equivalent knot items that are in Wikidata already, as listed here:
- Sounds good. ChristianKl ❪✉❫ 22:17, 14 June 2020 (UTC)[reply]
# All instances of the class "Knot" (Q1188853), recursively, removing duplicates
SELECT DISTINCT ?item ?itemDescription ?itemLabel ?alexanderbriggs ?alexanderpoly ?conwaypoly ?jonespoly
WHERE {
?item wdt:P31 ?sub0 .
?sub0 (wdt:P279)* wd:Q1188853 .
OPTIONAL{?item wdt:P6432 ?alexanderbriggs .} # Alexander-Briggs notation
OPTIONAL{?item wdt:P5350 ?alexanderpoly .} # Alexander polynomial
OPTIONAL{?item wdt:P5351 ?conwaypoly .} # Conway polynomial
OPTIONAL{?item wdt:P5352 ?jonespoly .} # Jones polynomial
SERVICE wikibase:label { bd:serviceParam wikibase:language "en,en" }
}
ORDER BY ASC(?item)
LIMIT 10000
After this initial phase to create the items and their basic properties, I can then make further bot passes that will distribute the knots into finer subclasses, such as prime knot (Q1014491), torus knot (Q1892897), chiral knot (Q4064646), invertible knot (Q6060558), amphichiral knot (Q20873948). etc.
As and when other knot properties are created (to be proposed later, in their own separate property proposals), I can also add the values for these. For example, the Dowker-Thistlethwaite notation, which is a particularly useful representation of a knot, as it enables its diagram to be generated directly from the notation.
--The Anome (talk) 11:29, 14 June 2020 (UTC)[reply]
- Would there be no license issues? Is info licensed cc-0? Lymantria (talk) 05:52, 15 June 2020 (UTC)[reply]
- I don't believe there are any copyrightable aspects to this information, and the authors of knotinfo don't seem to assert any copyright: see [1]. Here is the situation as I understand it:
- the knots themselves are immutable facts of nature, like the polyhedra or the chemical elements. All of the properties are also facts of nature, and can be derived from the knots themselves algorithmically.
- There is, I believe, a general principle that there is no copyright in mathematical facts, regardless of how much effort was put into generating them [2],[3]
- There has been some minimal element of creativity involved in the assignment of names to the knots, but long-standing copyright doctrine holds that there is no copyright in names.[4]
- The tables are also exhaustive, listing all knots up to a crossing number of 12, so there is no creative element in selecting which items to list. Their ordering in the database is also arbitrary: you could sort on any field and get the results out in a different order for each one: we can sort them by lexicographic order of one of the computed properties (for example, the Dowker-Thistlethwaite notation) to determine the order for entry into Wikidata.
- In any case, database rights do not exist in the United States, where both Wikipedia and Knotinfo are based. See Feist v. Rural.
- I don't believe there are any copyrightable aspects to this information, and the authors of knotinfo don't seem to assert any copyright: see [1]. Here is the situation as I understand it:
- Update: All of this information, apart from the names, can also be generated automatically by software, for example, Mathematica, or knotkit, or many others. The names (to the best of my knowledge) are derived from tables which were published in books and papers first: while books and papers are copyrightable, the mathematical facts and names in them are not (see above), and they are not databases. For example, Conway's tables, and their correspondence to the Rolfsen names, are published in "An Enumeration of Knots and Links, and Some of Their Algebraic Properties.", collected in Computational Problems in Abstract Algebra., pages 329–358, published in 1970, which (a) isn't a database, (b) predates the creation of EU database rights in 1996, and (c) is more than 15 years ago, so any EU database rights, even if applied retroactively, would have expired by now. Simlarly, Rolfsen's Knots and Links, which created the Rolfsen names, was first published no later than 1976. As ever, not a lawyer, not legal advice, etc. -- The Anome (talk) 10:36, 15 June 2020 (UTC)[reply]
- Thanks for the clarifications, The Anome. Lymantria (talk) 05:27, 17 June 2020 (UTC)[reply]
- Update: All of this information, apart from the names, can also be generated automatically by software, for example, Mathematica, or knotkit, or many others. The names (to the best of my knowledge) are derived from tables which were published in books and papers first: while books and papers are copyrightable, the mathematical facts and names in them are not (see above), and they are not databases. For example, Conway's tables, and their correspondence to the Rolfsen names, are published in "An Enumeration of Knots and Links, and Some of Their Algebraic Properties.", collected in Computational Problems in Abstract Algebra., pages 329–358, published in 1970, which (a) isn't a database, (b) predates the creation of EU database rights in 1996, and (c) is more than 15 years ago, so any EU database rights, even if applied retroactively, would have expired by now. Simlarly, Rolfsen's Knots and Links, which created the Rolfsen names, was first published no later than 1976. As ever, not a lawyer, not legal advice, etc. -- The Anome (talk) 10:36, 15 June 2020 (UTC)[reply]
- I am not a bot approver here, but I Support this proposal.
- With respect to copyright issues, I agree with The Anome's argument that the facts being contemplated for importation are not copyrightable. There would be significant concerns about importing all the knot images from the knotinfo site, but that is not being proposed. —Scs (talk) 10:56, 15 June 2020 (UTC)[reply]
- Agreed. We should generate our own knot images from scratch programmatically, preferably with a statement by the author of the knot drawing program that they are willing to license the images as CC-0 to eliminate any doubt. This will also benefit Commons, and potentially Wikipedia as well. Ideally, the images would be SVG, to maximize reuse potential. -- The Anome (talk) 11:01, 15 June 2020 (UTC)[reply]
- Note that Knotilus may be a possible source of images: see this example of a knot computed from a Gauss code. -- The Anome (talk) 17:04, 16 June 2020 (UTC)[reply]
- Please note: I've also created a property proposal, Wikidata:Property proposal/Dowker-Thistlethwaite notation, to add one more field for the bot to add. Unlike the other properties, it can be used to synthesize knot diagrams, and (up to isomorphism) uniquely identify knots. -- The Anome (talk) 12:25, 15 June 2020 (UTC)[reply]
- Agreed. We should generate our own knot images from scratch programmatically, preferably with a statement by the author of the knot drawing program that they are willing to license the images as CC-0 to eliminate any doubt. This will also benefit Commons, and potentially Wikipedia as well. Ideally, the images would be SVG, to maximize reuse potential. -- The Anome (talk) 11:01, 15 June 2020 (UTC)[reply]
Tobias1984
Arthur Rubin
Cuvwb
TomT0m
Physikerwelt
Lymantria
Bigbossfarin
Infovarius
Helder
PhilMINT
Malore
Lore.mazza51
Wikisaurus
The Anome
The-erinaceous-one
Daniel Mietchen
Haansn08
Xenmorpha
John Samuel
Jeremy Dover
Toni 001
Bocardodarapti
Duckmather
HTinC23
fgnievinski
Paul-Olivier Dehaye
uni
Dexxor