|
This is version 21.
It is not the current version, and thus it cannot be edited.
[Back to current version]
[Restore this version]
Notes created by Shawn Bowers
- 8:30 - 9:00
- Jim Beach on today's schedule; politics in Kansas primer
- Guest from NatureServe (Gwynn ...)
- Mike Frame from USGS
- Liz Blood from NSF (not here yet ...)
- NCSA ...
- New Folks
- Laura Downey (unm, usability engineer)
- Jianting Zhang (postdoc, unm)
- Steve Teckell (seek web-developer)
- Mark Serilla (LTER, network information system)
- 9:00 - 12:30
- Overviews, etc.
- Matt Jones
- Kepler
- Goal is to: create powerful analytical tools that is useful across disciplines; ecology, biology, engineering, geology, physics, ...
- Model the way scientists work w/ their data now: mentally coordinate export and import of data among software systems
- Workflows emphasize data flow
- metadata-driven data ingestion
- output generation includes creating appropriate metadata
- Kepler Today
- EcoGrid support: EML data support; DarwinCore data support; EcoGrid registry to discover data sources
- Ontology-based browsing
- Demonstration workflows
- Ecological niche models
- Biodiversity analysis tools
- EcoGrid queries ...
- EcoGrid: EML Data Access
- Right-click data box to get EML metadata display
- Query builder
- QBE style query
- Query execution: EML Metadata + Data -> generate DDL to create temp dbs -> HSQL executes data -> result
- EcoGrid DarwinCore Access
- Actor Search and Browse: support multiple views (tuned to ecologist) ... switchable at runtime
- Kepler for Ecological Modeling
- Biodiversity indices (from Dan)
- R in Kepler
- ROADNet Realtime (Revelle ship)
- Kepler This Year:
- User Interface ... need a full eval and reworking of all UI for our target group
- Semantic annotation interface / ontology browsing interfaces
- Distributed computing / grid computing
- large jobs; lots of machines; detaiched execution
- "Smart" data discovery (as for actors)
- support for data sources that are distributed
- Actor repository / downloadable actors
- Integration and transformation using semantics
- generate required transformation for: data-actor integration; actor-actor integration
- handle "semantics" for R scripts, etc.
- file based buffering reduces our ability to reason about the data flow
- SMB: not exactly sure where the precise problem is here ...
- Finish EcoGrid Access
- full EML support
- support for "large" data and 3rd party transfer
- more data sources and types of data sources (e.g., JDBC, GEON data)
- Provenance and metadata propagation
- Shawn Bowers (me)
- Jessie Kennedy
- SEEK Taxon Scenario
- Goal is to develop the taxon name server
- High-level approach
- extract information from taxonomic concept providers; and convert to taxonomy transfer schema (TML)
- "Semantic Mediation System" used to do concept matching/expansion/weighted concepts; leveraged from EML
- Most focus on Taxonomy Transfer schema
- Names versus concepts problems
- use concepts instead of names .... decouple concepts from names
- taxon name: label/word/string
- used for communicating ideas about organisms or groups of organisms ... meaningless without a definition, but it is used as if it has a clear meaning
- full scientific name: sci name + author abbreviation +date?
- implies a concept: as intended and documented by the original author of then ame
- vernacular name: scrub history
- What is a concept?
- taxon concept: unique name plus definition
- full sci. name + "according to" (author + date)
- Taxon goals
- define use case; acceess current taxon working practice; design schema for object server; implement prototype; eval globally unique ids and their applicability
- investigate interface between SMS and taxon group
- end user tools
- progress:
- use cases finished, but still under review
- taxon working practice: developed a langauge definition, etc ...
- version 1.01 ... aimee and robert will demo
- globally unique ids: UUIDs, handle system, DOIs, LSIDs
- prototype ...
- interface with sms: session during the week to discuss
- case study, anticipating north american mammals example
- Deana Pennington
- Original goals
- case studies in ecology and biodiversity
- assemble a framework that interconnects ...
- First project: eco niche modeling
- mammal project
- 2-3 dispersal scenarios; 2 major evolutionary-computing algorithms (GA and NN); 21 GCM scenarios, including all IPCC scenarios
- eco niche modeling
- 2nd beam project, focusing on biodiversity
- use biodiversity and productivity field data
- link with enm project
- use biodiversity indices
- expand group of scientists
- nceas group: Waid, Scheiner, Willig, Cox, Chalcraft; biodiv/productivity scaling rels in semi-arid grasses
- pdtnet group: Suding, Weiher, Collins, Cleland, Clark; biodiv/productivity functional rels in semi-arid grasses
- ENM/biodiv integrated analysis
- remotely sensed imagery -> from Jim Quinn at Davis
- Samantha Romanello
- EOT
- Current activities: on-site trainings; conf. workshops and related activities; new website
- future activities: on-line modules; distributed graduate seminar
- on-site trainings
- new fac. and postdoc workshop
- how to incorporate into their teaching
- results of the workshop from last january
- Raja
- GOAL: EcoGrid as a standardized interface (using web and grid services)
- Grid-standardized interfaces: uniform interface to metacat, srb, digir, xanthoria; anyone can implement these interfaces; hides complexity of underlying systems
- metadata-mediated data access
- supports multiple metadata standard, EML, Darwin Core as foci
- computational services: pre-defined analytical services; on-the-fly analytical services
- all based on gwsdl service
- metacat and srb services implemented
- metadata: EML and darwin core
- authentication
- query sevice level 1: get/query services; web-based available
- registry services: implemented in 2 phases -- reimplemented -- list of avail services on web
- Next steps:
- design and impl. other ecogrid services
- extend existing services
- more ams integration - kepler
- more sms integration - growl (?)
- add new services: onto servers, taxon servers, LSID server
- user interface development
- documentation and training
- enable more apps and data ingestions
- increase user collaborations
- Laura Downey
- usability
- "the extent to which a product can be used by specified users to achieve specified goals with effictiveness, efficiency, and satisfaction in a specified context of users"
- "an approach that incorporates direct user feedback througout the development cycle to reduce costs and create products and tools that meet user needs."
- benefits o fusability: increased productivity, sales and revenues, customer satisfaction; decrease traning and support costs, etc.
- real world example: after move.com completed the redesign of two problem features, users ability to find a home increases from 62 to 98 percent, sales lead generation to real estate agents increased 150%.
- user interface + functionality = usability
- elegant or sophisticated internal algorithms don't make any diff inf the interface is bad ...
- software versus usability engineering: ui code 30 percent, now almost 50-60; maintainence, etc.
- typical software program is released with 40 usability design problems; over 70% of CIO's state on of biggest problems is communicating with users
- incorporate usability in seek
- collect demographic data
- task analysis (interviews and observations)
- apply good human factors and design principles
- iterative design
- gather user feedback and input: ind. interviews; group discussions; usability evals and testing; categorization exercises; self-reporting; satisfaction surveys
|
