Wikidata:Requests for permissions/Bot/DL2204bot
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) 15:55, 14 February 2023 (UTC)
DL2204bot[edit]
DL2204bot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: DL2204 (talk • contribs • logs)
Task/s: Add LiLa Linking Latin URI (P11033) ExternalID values to Latin lexemes, add Latin lexemes
Code: I am using WBI 12.0; see code for writing a first batch of LiLa Linking Latin URI (P11033) values here.
Function details: We propose to add LiLa Linking Latin URI (P11033) values to Latin (Q397) lexemes, and create Latin lexemes from data contained in the LiLa Lemma Bank (see P11033 property proposal), in cooperation with LiLa project team at CIRCSE (Q89883181). LiLa Lemma bank contains about 200.000 lexemes; in LiLa, lexical resources such as dictionaries are linked on lemma level to LiLa lexemes. LiLa databases also contain corpora and information about lexical forms. The linking of WD Latin lexemes to the LiLa LemmaBank will allow further enrichment from or federation with content in LiLa databases. The span of the present request is the alignment at lexeme level using LiLa Linking Latin URI (P11033). The linking of the about 34.000 existing Latin lexemes to LiLa will be manually validated; new Latin lexemes (eventually some 160.000 more) will have a POS and a LiLa Linking Latin URI (P11033) value. Alignment of existing lexemes will be first. Before creating new lexemes, we will make announcements for collecting feedback in Lexicographical Data Telegram channel and on P11033 talk page.
--DL2204 (talk) 11:57, 4 September 2022 (UTC)
Support Great things will come from this bot!
- Theklan (talk) 12:09, 4 September 2022 (UTC)
Comment It's really cool! There should be a specific request for permission for each task. You should redirect the talk page of your bot to your own talk page, so you will not miss any feedback. And don't forget the test runs. Cheers, — Envlh (talk) 15:23, 4 September 2022 (UTC)
- I think the externalID tasks are not critical regarding permissions, because the tasks will be performed in cooperation with the mantainers of the databases they redirect to. The Basque task is backed by the Basque WD community. Regarding Quechua, we plan to use Qichwabase (Wikibase) for curation until we get a noise-free dataset to upload, validated by natives. DL2204 (talk) 16:11, 4 September 2022 (UTC)
- I think the externalID tasks are not critical regarding permissions, because the tasks will be performed in cooperation with the mantainers of the databases they redirect to. The last time it happened (a week ago...), it didn't end well at all (the bot owner left the import half-finished, angry because the community reported issues with the import). One common flaw with newcomers in Wikidata is to try to do too big tasks too fast too early. Importing data into Wikidata is not the same as to do it in your own database/Wikibase that you manage. You should really start with one small easy task, complete it, and then continue with the next one, relying on the experience you gained with the first one. Cheers, — Envlh (talk) 16:46, 4 September 2022 (UTC)
- I fully understand your concern, and also understand your recommendation to restrict this request for permissions. So, should I leave just one task in this request? (The Latin task would go first.) And, for future additional tasks, I guess I won't have to request another bot account but only permissions for another task using the same account. Is that true? Thanks. DL2204 (talk) 21:09, 4 September 2022 (UTC)
- Yes, my advice is that you should put only one task in this request. For future tasks, you can proceed the same way, with the same bot account (so you don't have to setup everything again). Here is an example of the 9th request for a bot: Wikidata:Requests for permissions/Bot/MsynBot 9. Don't hesitate to ping me (on wiki or off wiki) if you need reviews for your requests. Cheers, — Envlh (talk) 21:01, 5 September 2022 (UTC)
- OK, yes, I have left the Latin task in the proposal. Thanks. DL2204 (talk) 19:24, 10 September 2022 (UTC)
- Yes, my advice is that you should put only one task in this request. For future tasks, you can proceed the same way, with the same bot account (so you don't have to setup everything again). Here is an example of the 9th request for a bot: Wikidata:Requests for permissions/Bot/MsynBot 9. Don't hesitate to ping me (on wiki or off wiki) if you need reviews for your requests. Cheers, — Envlh (talk) 21:01, 5 September 2022 (UTC)
- I fully understand your concern, and also understand your recommendation to restrict this request for permissions. So, should I leave just one task in this request? (The Latin task would go first.) And, for future additional tasks, I guess I won't have to request another bot account but only permissions for another task using the same account. Is that true? Thanks. DL2204 (talk) 21:09, 4 September 2022 (UTC)
- I think the externalID tasks are not critical regarding permissions, because the tasks will be performed in cooperation with the mantainers of the databases they redirect to. The last time it happened (a week ago...), it didn't end well at all (the bot owner left the import half-finished, angry because the community reported issues with the import). One common flaw with newcomers in Wikidata is to try to do too big tasks too fast too early. Importing data into Wikidata is not the same as to do it in your own database/Wikibase that you manage. You should really start with one small easy task, complete it, and then continue with the next one, relying on the experience you gained with the first one. Cheers, — Envlh (talk) 16:46, 4 September 2022 (UTC)
- I think the externalID tasks are not critical regarding permissions, because the tasks will be performed in cooperation with the mantainers of the databases they redirect to. The Basque task is backed by the Basque WD community. Regarding Quechua, we plan to use Qichwabase (Wikibase) for curation until we get a noise-free dataset to upload, validated by natives. DL2204 (talk) 16:11, 4 September 2022 (UTC)
- Comment The first manually validated batch of 27,485 values for LiLa Linking Latin URI (P11033) (LiLa Lemma Bank to Lexeme alignment) is now ready to be uploaded. Data to upload and upload scripts are accessible here. DL2204 (talk) 19:20, 24 January 2023 (UTC)
- Thanks for the request and work to align the lexemes. I feel very happy seeing you successfully wrote the code using WBI 😀. I suggest you make 50-100 test edits and link them here so we can see how it looks. —So9q (talk) 22:04, 24 January 2023 (UTC)
- Following your suggestion, I have written the first 50 values as normal user, without bot flag (done items log). DL2204 (talk) 10:58, 27 January 2023 (UTC)
- Support Great! the bot works exactly as expected, and does what it is supposed to do efficiently. Thanks! Francescomambrini 14:25, 27 January 2023 (UTC)
- Support I had a look at the data and it looks great. This is a very useful task. Sina.recherche 11:27, 28 January 2023 (UTC-5)
- Thanks for the request and work to align the lexemes. I feel very happy seeing you successfully wrote the code using WBI 😀. I suggest you make 50-100 test edits and link them here so we can see how it looks. —So9q (talk) 22:04, 24 January 2023 (UTC)
