- From: Deborah Dahl <Dahl@conversational-Technologies.com>
- Date: Wed, 8 Sep 2021 12:08:57 -0400
- To: <public-voiceinteraction@w3.org>
https://www.w3.org/2021/09/08-voiceinteraction-minutes.html and below as text [1]W3C [1] https://www.w3.org/ - DRAFT - Voice Interaction 08 September 2021 [2]IRC log. [2] https://www.w3.org/2021/09/08-voiceinteraction-irc Attendees Present bev, debbie, dirk, jon, tobias Regrets - Chair debbie Scribe ddahl Contents 1. [3]publicity update 2. [4]meeting at W3C Technical Plenary 3. [5]planning for interfaces work 4. [6]emma properties Meeting minutes <ddahl> agenda: Agenda 1. publication publicity See marketing plan at [7]https://github.com/w3c/voiceinteraction/blob/master/ marketing-plan.md 2. Should we meet at TPAC with any other groups or propose a breakout session? Currently scheduled group meetings: [8]https://www.w3.org/wiki/TPAC/2021/GroupMeetings Last time we discussed: Web Real-Time Communications Working Group Breakout groups [9]https://www.w3.org/wiki/TPAC/2021/ SessionIdeas 3. planning [CUT] [7] https://github.com/w3c/voiceinteraction/blob/master/marketing-plan.md [8] https://www.w3.org/wiki/TPAC/2021/GroupMeetings [9] https://www.w3.org/wiki/TPAC/2021/SessionIdeas publicity update debbie: contacted Speech Technology Magazine and they were interested . online newsletter could come out anytime jon: OVON newsletter will go out soon . Bradley and Martin several thousand subscribers . Brett podcast in October . invited Debbie and Tobias tobias: invited to VUX World podcast . on interoperability meeting at W3C Technical Plenary debbie: last time talked about meeting with WebRTC . but breakout session would be better . standards for voice agent interoperability debbie: will set up breakout session at TPAC . send link to our session planning for interfaces work debbie: questions from last time about items 4 and 5 in walkthrough The audio is sent along with all augmenting data to the IPA Service. . what is augmenting data? dirk: not defined, one example would be GPS debbie: we can define anything we need debbie: what is the "received data" in step 5? dirk: just whatever we got in step 4 debbie: we take the same data and send it down both paths dirk: timestamps could be included, also media dirk: but not speech recognition debbie: sent list of MMI properties list of properties at [10]https://lists.w3.org/Archives/Public/ public-voiceinteraction/2021Sep/0000.html [10] https://lists.w3.org/Archives/Public/public-voiceinteraction/2021Sep/0000.html jon: should OVON start from here? debbie: I think so also EMMA tobias: we should talk about required or optional debbie: we have this in all the specification tobias: there's no intent debbie: left out semantics of NLP tobias: it would be in interpretation of EMMA debbie: left that out originally, but things have changed . we could standardize what we use for the name of "intent" or "slot" . could use a registry emma properties bev: "dialog turn" debbie: turn of a dialog debbie: EmotionML has finished goes through ASR and NL properties of EMMA tobias: how would we move forward? debbie: just move forward . with our documents . could be a de facto standard debbie: next time should we take up the interfaces document and define a high priority interface dirk: that makes sense dirk: still needs to align interfaces document with architecture . diagrams
Received on Wednesday, 8 September 2021 16:09:17 UTC