Comments on: What kinds of invited speakers could we have?

By: Zeerak Waseem

Zeerak Waseem — Fri, 29 Sep 2017 08:05:26 +0000

I’d like to see people from the social sciences who research algorithmic bias and discrimination and/or research the impact of algorithms have had on our lives from another point of view than the typical NLP/ML crowd.

Some suggestions could be people like Liz Losh, Nishant Shah, or Wendy Chun.

By: Dave Howcroft

Dave Howcroft — Tue, 19 Sep 2017 11:54:18 +0000

I like Chris’ idea of the literature/storytelling/artistic area and would add a general digital humanities suggestion. IIRC, this might have been a recent keynote topic at ACL or another major conf, though. I think this is an area where there could be more active interdisciplinary collaboration in general.

Like Cassandra, I’m selfish: I think it’d be nice to have someone talk about natural language generation, but there was a talk from our community at NAACL last year (by Ehud Reiter). I guess a more general version of this would be to highlight one of the smaller CompLing/NLP communities.

Following on themes from SIGDIAL and INLG this year, it could be interesting to have a talk about combining symbolic and statistical or neural approaches to AI/CompLing. There are a lot of cases where being able to fix a rule is helpful or can leverage linguistic knowledge but we’d like to combine those benefits with the ability to learn from more data to avoid building anything too brittle.

By: Chris Brew

Chris Brew — Tue, 19 Sep 2017 01:27:04 +0000

I like people doing applications of NL that don’t immediately come to mind for
ACL focused grad students. Want to get their minds off of network architectures and onto language tasks as things that live in and affect the world.

Two suggestions

– people doing NL applications in psychiatry (preferably not the core NLers who do that)
– people doing NL work on literature, storytelling or artistic endeavors (again not core NLers)

All other things being equal, not too senior. Thought of Molly Ireland, who is very creative.

Anti-recommendation: no chatbots for now. maybe one day, if they ever work.

By: Cassandra Jacobs

Cassandra Jacobs — Mon, 18 Sep 2017 23:30:52 +0000

Selfishly, models informed by cognitive processes, or computational social science more broadly.

By: Leon Derczynski

Leon Derczynski — Fri, 15 Sep 2017 08:27:12 +0000

Indeed. A broad consultation is absolutely our goal, spanning multiple groups and bodies. Though the ICCL is not left out; we have a representative as PC co-chair, making the ICCL very influential in the decision making process. And of course, ICCL members are not barred from submitting their recommendations.

By: Miles Osborne

Miles Osborne — Fri, 15 Sep 2017 05:58:11 +0000

interesting areas:

* privacy / ethics
* disaster management
* crowd sourcing

As mentioned on FB, I would avoid having well known names from the CL community talk about their work. (You can always hear about them anyway).

By: Yorick wilks

Yorick wilks — Thu, 14 Sep 2017 21:56:17 +0000

It seems traditional to consult ICCL. THeyve been around a long time and seen lots of fads come and go.

By: Ted Pedersen

Ted Pedersen — Thu, 14 Sep 2017 17:01:12 +0000

I always enjoy when there is some contrast among the invited speakers – insider / outsider, young / old, linguistics / computer science, etc. I guess at this particular moment in time it might be interesting to have a linguistics / deep learning contrast – I don’t have particular names in mind, but it might be nice to invite people who we might not see at our events otherwise. Another thought would be a sort of meta topic, not specifically nlp or cl related but perhaps someone talking more generally about issues of diversity in scientific communities (and how to improve and cultivate that) or issues of scientific practice (reproducibility, fraud, etc). Again, no specific names in mind, just thinking big picture here.

By: Sowmya

Sowmya — Thu, 14 Sep 2017 00:28:49 +0000

I’d like to hear from linguists (for a change). Last year in NAACL, there was a panel full of deep learning and computational linguistics researchers and I wondered if there is really any role for linguistics in contemporary NLP. I would love to listen to linguists of the kind who can understand CL/NLP and its relevance (e.g., a typologist who can see UD as a practical version with similar goals). in COLING 2012, there was Paul Kiparsky giving a keynote – although it is not about NLP per se, that was one inspiring talk I heard in my life as a NLP-er

By: Kyle Gorman

Kyle Gorman — Wed, 13 Sep 2017 15:59:06 +0000

I’d like to hear from at speakers who studies formal learnability of languages/grammars.

I’d like to hear from speakers who work in industry but specifically on applications (and not just in the pure research element). In particular applications in healthcare?

I feel that there is an *extremely small* pool of speakers invited to NLP conferences and I’d like to suggest the following: the PC *not* invite anyone who gave a keynote at an international NLP conference in 2017 or 2016.