NOTE: slides etc. from working group reports available as attachments to this page (links near the bottom of the page)

Outstanding:
1. Debbie: discuss how to get linguists in various countries involved in standards development
2. Needed: more scientific way of determining relationships pertaining to classification of languages. Need to have a way to get access both to a proposed grouping and the evidence/support for proposing that grouping. Need ways of linking to the data used to support that grouping.




Planning for Sunday Afternoon Report

slide:
* approach: getting a start at organizing vast amount of information
* brief outline of four areas of concern: storage, retrieval, search, access/reuse
* approach: not attempting to define standards for all of linguistics

slide:
* knowledge changes: standards must be adaptable to changes
* case studies re: adaptability of standards

slide:
* importance for the individual linguist of being able to find the how-to as well as the what-to
* subfield-specific seed lists


somewhere:
* collection/dissemination of conventions used in currently-available data sets

-------------------------------------------------------

NOTE: other files attached... links at bottom of page

WS1: Planning [Friday afternoon, 2:30-5pm]

* Decide what we want to have accomplished by the end of the workshop, refine/adjust agenda

• What concrete goals can we realistically aim to accomplish?
• Are there other, more abstract goals we wish to accomplish?
• How can we best lay a groundwork for moving forward, facilitating future development in our area?
• At what level do we want to aim our efforts? We can consider this in (at least) two different ways:
  
• ..... granularity wrt linguistics and its subfields. Do we want to work on standards for individual subfields? At least two of our members (myself included) have done work in particular subfields. Or do we want to stay at a more general level, considering standards/solutions for the whole of linguistics?
• .... second is the question of granularity wrt areas of interest. Do we want to aim for more detailed recommendations for a few key areas related to storage/retrieval/search/use? Or rather focus on a broader understanding of how to support interoperability between these areas?

* According to these goals, decide how we want to set up resource pages on the wiki and who can take main responsibility (at least for the course of the workshop) for individual pages/areas.
* Talk about reports (two interim reports (~5-6 minutes each), one longer report at the closing panel (~20-25 minutes?)) and who's interested in presenting them
* Talk about the written report and how we'd like to handle that.
--------------------------------------------------------------------------------------------
NOTES FROM FIRST WORKING SESSION
--------------------------------------------------------------------------------------------


Alexis - notes the tedium and difficulty of linguistics without standards. Notes absence of labeled data. Combine machine learning and human annotator. Complex learning problem for machines (annotation and research happen in parallel).

Johanna: notes about research process. Understanding changes, recording and coding has to change over time. Notes that any tools and standards need to be tentative. Ontologies must be adaptable.

Eric: notes about the danger of standards inertia. Infrastructure and tools get built off of standards, making the standards harder to change.

Alicia: Notes the issue of version control issue. Keep enough information of why classifications and taxonomies are determined.

Alicia: Buy-in issue. Need simple but powerful, but it is hard to get colleagues to go with it. Cultural issue about tools and tool adoption.

Johanna: wants to have a basic inter-linearizing tool.
--------------------------------------------------------------------------------------------

WS2: Existing standards [Saturday morning, 10am-12pm]

* For each focus area, what are the existing standards? Which seem to be more widely accepted and/or adopted? By linguists/linguistics? If the standards are either general or oriented toward other disciplines, how well would they work for linguistics? With how much adaptation (if any)? In the areas targeted by the standards, is it realistic to expect adoption of a set of standards within linguistics? Best practices? How much flexibility is necessary in the standard? How flexible is the standard already? ... etc. ...

--------------------------------------------------------------------------------------------

WS3: Needed standards [Saturday afternoon, 3-6pm]

* For each focus area, where are standards most needed in order to build a successful cyberinfrastructure? Which are the top priorities? Any suggestions for development? Are there existing organizations which seem well-posed to aid/facilitate development of such a standard? Existing standards which could be modified? What are the potential problems/challenges? ... etc. ...

--------------------------------------------------------------------------------------------

WS4: Going from abstract to concrete, clarifying how to move forward such that real results are most likely to occur [Sunday morning, 10am-12pm]

Notes:
Language classification: [New version of my slideshow now attached. JN]




--------------------------------------------------------------------------------------------

Some more general questions to ponder:

1. What makes a standard a standard, as opposed to, e.g., a best practice recommendation or a commonly-followed convention?
2. For the field of linguistics, are there specific areas where flexibility in standards is needed? Where rigidity is needed?
3. What creates incentive for researchers to follow standards?
4. How can systems and standards be designed to accommodate a multitude of needs/interests (e.g. needs of specific subfields, practices and preferences of individual researchers, etc.)?
   
5. How do we help lower barriers to entry and reuse (especially unanticipated applications and ad hoc uses in multidisciplinary settings)? How do we design services that facilitate experimentation and "mashups"?

STORAGE: archiving, digitization, formats & standards, legacy data