Telecon 11.09.2013

From RDF Stream Processing Community Group

Participants

  • Jacopo Urbani
  • Andy Seaborne
  • Esko Nuutila
  • Jean-Paul Calbimonte
  • Oscar Corcho (scribe)
  • Danh Le Phuoc
  • Emanuele Della-Valle
  • Marco Balduini
  • Daniele Dell'Aglio
  • Eva Blomqvist
  • Lorenz Fischer
  • Robin KeskisŠärkkŠä
  • Stephen Allen
  • Jim Smith
  • Alessandro Margara
  • Mikko Rinne

Agenda

Minutes

Members short presentation

Already on the public mailing list.

RSP CG introduction & goals

Jean Paul: explaining reason for having a Community Group instead of an Incubator Group (they do not exist anymore)

Jean Paul: even if it is a more "informal" way of running a group, it can still focus on what the aims of incubator groups were (that is, propose joint documents for potential standardisation, etc.)

Emanuele: nothing more to add to what Jean Paul has been commenting.

Perspectives on RDF stream models queries and services

Jean Paul: several types of background profiles in this group. Given this, maybe the work can be organised according to the following topics:

  • 1. RDF Stream models. If we do not know what the model is, then it will be difficult to know what to expect when sending queries, etc.
  • 2. RDF Stream querying: understanding the query languages, extensions to SPARQL, etc.
  • 3. RDF Stream publishing services: how can we publish RDF streams, how do we adhere to REST, etc.

Andy Seaborne: Agree with point 1 - the model of data moving is essential.

Jacopo: are we going to consider anything about reasoning?

Jean Paul: this could be also considered as part of the objective 2.

Emanuele: before starting talking about reasoning it would be probably better to focus first on models and query evaluation Andy Seaborne: The other questions is one of ambition. Is the objective consolidation of current research or enable new (sem)web capabilities and uses?

Emanuele: where do we put the protocol part of query execution and results? In group 2 or 3?

Jean Paul: it looks better to have it in group 3

Jim: would it make sense to talk about producers (the ones who produce data streams), consumers (those who consume it), and intermediate processors (those that process an incoming stream and may produce another one)

Jean Paul: decide on what the next steps should be.

Jean Paul: proposes to start collecting in the following 2 weeks links to documentation on how people have been working on this area so far

Jean Paul/Emanuele: so that we can come out to the f2f session with enough background about everybody else

Oscar: I agree (sorry that I am silent - I am in a very busy place and do not want to disturb you)

ACTION: collect RDF Stream models by September 25th

ACTION: collect RDF Stream querying by October 9th

ACTION: collect RDF Stream publishing/processing by f2f meeting

Set up periodic calls (bi-weekly?)

ACTION: next calls, as per Lorenz/Avi request, will be sent out to doodle.

F2F meeting in Sydney

Oscar: Agenda item: f2f meeting

Lorenz: could it be also doodled?

Emanuele/Jean Paul: that's difficult, as it was really difficult to get sufficient slots available around the conference

Emanuele: f2f meeting will be Tuesday 22nd, during the afternoon, in the second half of the Ordring workshop, and then we can continue afterwards http://streamreasoning.org/events/ordring2013#program

Oscar: Who from here will be present in that f2f?

Jacopo: I'll be at the meeting

Jacopo: I'll be at the meeting in Sydney

Alessandro Margara: I won't be there, unfortunately..

Eva Blomqvist: Oscar: from LIU/CNR both me and Robin will be in the f2f

Robin KeskisŠärkkŠä: I will be at the meeting in Sydney

Lorenz Fischer: I will be there

Use cases

Andy: which are the current users of such streaming formats or who consider themselves potential users?

Jim: we are trying to see whether our Open Annotation work can reuse this

Andy: I am looking for other motivating examples: Is it only sensors or is there any other type of application?

Emanuele: should we consider building up a use case document?

Andy: looking for this yes.

Emanuele: we are using it for social analytics.

Danh: using it in the IoT domain

Mikko (Aalto): We are processing sensor-generated streams related to traffic and smart spaces.

Emanuele, Marco, Daniele (Polimi): shall we have a page where we collect use cases?

Oscar (sorry, not said on the phone): we are using it in transport

Oscar (again): and in the area of geomarketing

Lorenz: use cases are important before even going into any models.

Jean Paul: agree, let's also start collecting them on the wiki

ACTION: collect use cases in the wiki (after Jean Paul prepares a bit the wiki to make this collection of use cases - through a template)

Emanuele: template: which are the data sources, volume, velocity, any additional background knowledge?, etc.

ACTION (Emanuele): come up with a template

ACTION: collect use cases in the meantime, until the f2f meeting

Other concerns

Jacopo: it is very important to start getting industrial partners in order to make the results of the group more attractive

Emanuele: this is an important topic for all of us, trying to involve more people from industry

Andy: rather than just industrial partners, it is more about having a really active community. This is what the W3C is looking for (industry partners are an indication of that)

Oscar: yes, I agree

Jean Paul: AOB?

Thanks for organising this, Jean Paul, and Danh for hosting

Oscar:I will save the chat and upload it on the wiki

Oscar:bye

Eva Blomqvist: bye

Robin KeskisŠärkkäŠ: bye

Andy Seaborne: Thanks all.

Lorenz Fischer: thank you

Alessandro Margara: bye

Actions

  • ACTION: (JP) create wiki for collecting RDF stream models and query processors
  • ACTION: (all) collect RDF Stream models by September 25th
  • ACTION: (all) collect RDF Stream querying by October 9th
  • ACTION: (all) collect RDF Stream publishing/processing by f2f meeting
  • ACTION: (JP) next calls will be sent out to doodle.
  • ACTION: (JP) wiki for the use cases
  • ACTION (Emanuele): come up with a template
  • ACTION: (all) collect use cases in the meantime, until the f2f meeting