Skip to content

FeforProcessors

Jump to bottom
AnnCopestake edited this page May 29, 2006 · 34 revisions

Processing session

See also FeforRmrs

Warning: this page is in a highly unstable state: best not to edit unless you're one of the organisers! Thanks, Ann.

Suggested timetable

Morning of the 13th

  1. Plenary meeting to discuss organisation, working groups etc.
  2. Initial working group meetings over coffee
  3. Session on testing

Afternoon of the 13th

  1. Interfaces, preprocessors and standoff annotation.
  2. Working group meetings etc

Morning of the 14th

  1. Tools and lexicon building

Testing (technology and practice for developers)

Interfaces, preprocessors and standoff annotation

  • Standoff Annotation Interface / SMAF (BenWaldron and UlrichSchaefer)

  • Preprocessors (FrancisBond)

  • PET API (proposal, discussion) (UlrichSchaefer)

  • Grammar interfaces: clear interface for robustness-related grammar elements (paths/types/pos-tags/expected tokenization/punctuation etc.) e.g. specified in the inline documentation in such a way that it could be processed/extracted automatically from grammar sources. (UlrichSchaefer)

Tools

  • Presentation/Demonstration: Trollet: tool for multilingual grammar engineering - a wrapper around the LKB (PavelMihaylov)

  • Presentation/Demonstration: NL1 project by Martin Magnusson, integrated LKB with the theorem prover SNARK and Temporal Action Logic. LarsAhrenberg

  • Discussion: LKB/PET compatibility issues: shared configuration, reducing and documenting the globals (FrancisBond)

  • Multilingual grammar engineering issues (NTNU) [ not currently sure what this means, have asked - AAC ]

Lexicon building

There will be at least two talks about this in the plenary session.

  • Lexical acquisition, esp. for immature grammars (EmilyBender)

  • large lexica and LexDB (NTNU)

  • large lexicons (FrancisBond)

Working groups

The aim of a working group is to get something working! Working groups are intended for topics that involve a small group that needs to sort out something specific/technical. For instance: there's some tool that the leader developed, the participants have an active interest in and have used at a novice level, but they want it to get it to work on a new grammar etc. We will try to identify possible group leaders, who will sort out the topic and group composition before Fefor. Scheduling is done by the group on an ad-hoc basis.

Installation parties

Suggested by UlrichSchaefer. Leaders have expertise in some resource. The objective is to get particular resource running on participants' computers AND to document on the Wiki on an ongoing basis. i.e, rather than take notes on pieces of paper, the installers write notes on the Wiki. In an ideal world, later installers have a go at installation from the Wiki instructions alone while still at Fefor so we get successive stages of refinement. Leaders need to circulate info about hardware/software requirements in advance. We will schedule some time for these events and identify potential leaders.

Tours of resources

Short descriptions of particular resources. Possibly including:

  • DanFlickinger: Semantic test suite (for people who haven't seen it, in advance of discussion in FeforRmrs)

  • AnnCopestake: Algebra checking tool (for people who haven't seen it, preferably in advance of discussion of algebra in FeforRmrs)

  • UlrichSchaefer: Defining SDL cascades in HoG e.g. taking the recently integrated RMRS-merging stylesheets (developed by AnetteFrank)

Also the resources listed above under working groups or installation parties, for people who just want a introduction.

Home | Forum | Discussions | Events

Clone this wiki locally