Just a random thought to see if we have many/any takers. Fresh off attending my monthly short story bookclub, wondering if anyone would be interested in a monthly Zoom meetup for tinkering with some “NLP tool”.
The idea would be to showcase some tool that people were interested in working with and like a book club we:
Vote on the next month’s tool
The session leader (I expect this to be mostly me) does some ground work here and there during the month, like create a Colab with the tool and right dependencies installed
We spend maybe an hour or so tinkering with the Colab with screen share (perhaps in breakout rooms if there’s too many people!)
I plan for this to be a mainly social and super-low stakes gathering from which we might even just build:
a catalogue of tools and
minimal working examples for people to get started with
Hoping this would partly help with the issue of “so much tooling has been developed for low resource NLP but so little is accessible” problem @SarahRMoeller and @cbowern.
Great idea @fauxneticien! I think one on setting up the morphological parser in FLEx might be appreciated. I’m glad to volunteer my sketchy knowledge there.
Also, I’m teaching a project based class this spring where students have an option to prepare a tutorial for some technical skill. So let me know if you want volunteer leaders later this spring.
Hi all — happy New Year! Glad to hear that there’s some interest!
To kick things off let’s kick off with some polling for:
a date/time for the first meeting
a tool to demo
For times for 1, I think for a synchronous Zoom, we might just be able to make the timezones work for @cbowern , @SarahRMoeller (US Eastern), me (US Pacific), and @skalyan (AU Brisbane). (Assuming somewhere in the US for @jrrabbit based on profile/intro post).
I think we might be looking at something like: 6 pm US Eastern, 3 pm US Pacific, next day 9 am AU Brisbane: The World Clock Meeting Planner - Details (arbitrarily picked February 14/15 as a date).
Action item: can each of you respond with whether this time configuration might work and if so a list of US dates (Siva it will be +1 for you, e.g. February 15 for US 14).
Action item: can each of you suggest a tool or add your votes to another tool already suggested?
To keep things simple since there’s only 5 of us so far, here’s a (publicly-)editable Google Sheet:
Okay since vote for date is tied, let’s do February 21st 3 pm Pacific (6 pm Eastern) (February 22, 9 am AU).
Looks like the tool choice is also tied so I’m going to pick pympi. I’ll send out a calendar invite with a Zoom link to everyone’s e-mails and post a public link in this thread closer to the date.
For a future session with pyfoma I could also chip in as I recently built a morphological analyzer with it and I could show you how to support rhythmic templates with reduplication
Another interest of mine which would be cool to look at would be tools for extracting information from PDFs, how to deal with idiosyncratic font encodings, processing tables… things that can be very handy when working with existing language documentation.
Extracting tables and figures with docling (yes … they stole our name … and please ignore “gen AI” in the description of this tool, it is also very good at converting PDF to HTML )
Visualizing PDF objects and metadata in a Colab notebook
If you will be using DoReCo for this–perhaps for other apps that may be particularly useful to those working on polysynthetic or contact languages, please consider including data from Warlpiri and Light Warlpiri.
Kihchi-marsii! Thank you!
Ah, I’m so sorry I missed it! I was recovering from a major deadline. Hope it went well, and I’d be interested to read a recap if anyone cares to share
@lgessler — no worries. I think it went super well in that we kind of just hung out and chatted for perhaps an hour and a half and kind of didn’t get to doing the pympi demo. Hope you can join next time!