Databasing the world
Download as PPTX, PDF1 like329 views
This document discusses key aspects of building databases to catalog global biodiversity in the 2000s, including standards, technology, data sharing challenges, and classification methods. It covers how database infrastructure requires stable standards and technology to ensure data accessibility over time. Issues around data ownership, privacy, and ensuring data can be shared and reused across disciplines are also addressed. Classification systems are evolving from paper-based to digital formats using tools like cladistics and computer programs to help organize the vast amounts of data being collected through worldwide biodiversity projects.
Databasing the world
- 1. Databasing the World:Biodiversity and the 2000sWritten by Bowker, G. C. Presented by Chen Zhang (Mike)
- 2. Four Key AspectsDatabase InfrastructureStandards—flexible, stableTechnology—stable CommunicationData SharingOwnershipDisarticulationData collection
- 3. Four Key AspectsDistributed Collective PracticeCollaborate workNew Knowledge EconomyAccounting for lifeDevelopment of ClassificationCladisticsThe Future
- 4. Database Infrastructure
- 5. StandardsWhy do we need standardsExample of air-conditioner industryDiameter Match between screw and the hole on the panelReasons for databaseNeed ‘handshake’ among various mediaMIME<Multipurpose Internet Mail Extensions>protocol Each layer of infrastructure requires its own set of standardsNeed standardized categories.
- 6. StandardsStandards will not always winSome best-known standardsQWERTY keyboard
- 7. StandardsStandards will not always winSome best-known standardsVHS (Video Home System) standard
- 8. StandardsStandards will not always winSome best-known standardsDOS computing system
- 9. StandardsStandards will not always winWhy?The best standard maybe doesn’t have best marketStandards setting is a key site of political workThe inferior standard may be respected by the political agency. ( Such as standards-setting bodies)
- 10. StandardsInteroperabilityContinuum of strategies for standards settingOne Standard Fits AllLet A Thousand standards bloom
- 11. StandardsInteroperabilitySome Related Standards1. ANSI/NISO Z39.50ANSI/NISO Z39.50 is the American National Standard Information Retrieval Application Service Definition and Protocol Specification for Open Systems Interconnection. IT makes it easier to use large information databases by standardizing the procedures and features for searching and retrieving information.
- 13. StandardsInteroperabilitySome Related Standards1. ANSI/NISO Z39.50 A single enquiry over multiple databases. widely adopter in the library world.
- 14. StandardsInteroperabilitySome Related Standards2. XMLExtensible Markup Language (XML) is a set of rules for encoding documents in machine-readable form. Two extremes: a. Colonial modelb. Democratic model (win out) People’s established computing environment
- 15. TechnologyTechnology must be stableNothing to guarantee the stability of vast data setsFailure of Paul Otlet’s well catalogued microfichesDevelopment of computer memoryHard to retrieve information
- 16. TechnologyTechnology must stableData accessible and usableInfrastructure will require a continued maintenance effortReasons a. Data is passed from one medium to anotherb. Data is analyzed by one generation of database technology to the next.
- 17. Issues of CommunicationProblem of reliable metadataMetadata—data about dataThe blue lines are metadata
- 18. Issues of CommunicationProblem of reliable metadataThe standard name of certain kinds of dataSearchable—easy to search over multiple databaseIssue—how detail does the name of data should be?Lack of details— the information of data is uselessToo many details— longer time, more work
- 19. Issues of CommunicationDublin codeThe Dublin Core set ofmetadata elements provides a small and fundamental group of text elements through which most resources can be described and cataloged.The Simple Dublin Core Metadata Element Set (DCMES) consists of 15 metadata elements:LanguageRelationCoverageRightsTitle Creator Subject Description Publisher ContributorDateTypeFormatIdentifierSource
- 20. Data Sharing
- 21. OwnershipControl of knowledgeMid-nineteenth century: only professionally trained scientists and doctors New information economy: from many peopleExample: patients group
- 22. OwnershipPrivacyKeep data private is difficult : Example: data is complied by third-company to generate a new, marketable form of knowledgeNew Patterns of ownershipScience has frequently been analyzed as a “public good”Increasing privatization of knowledge : It is unclear to what extent the vaunted openness of the scientific community will last
- 23. DisarticulationIdeal databaseShould according to most practitioners be theory-neutral, but should serve as a common basis for a number of scientific disciplines to progress.Example: genome databank new kind of science genome construct arguments about the genetic causation ≠ the process of mapping the genomeData must be reusable by scientists
- 24. The data in a database should be easily manipulated by other scientists.Data CollectionBiodiversityLarge-scale databases are being developed for a diverse array of animal and plant groupsWorldwide effortIUBSCODATAIUMSDeal with old dataData was rolled into a theory should rememberAll its own dataPotentially data that had not yet been collected
- 25. Data CollectionDeal with old dataDifficultiesScientific paper don’t in general offer enough information to allow an experiment or procedure to be repeated.The distributed database is becoming a new model form of scientific publication in its own rightIssues of UpdateNo automatic update from one field to a cognate oneScientist are not able to share information across discipline divides
- 26. Data CollectionInternational TechnosciencePurpose: Narrow the gaps between countriesIssues:People do not have equal knowledgeAccess is never really equalGovernment have doubts of the usefulness of opening the database onto internet.
- 28. Collaborative WorkManagement structures in universities and industry still tend to support the heroic myth of the individual researcher.What kind of value the large publishing houses add to journal production.Great attention must be paid to the social and organizational setting of technoscientific work
- 29. New Knowledge EconomyThree central issuesThe development of flexible, stable data standardThe generation of protocols for data sharingThe restructuring of scientific careers
- 31. Development of ClassificationIntroduction: PANDORA taxonomic database
- 32. Development of ClassificationImportance of classification18th-19th centuries : botanist must know all genera, and commit their names to memory, but cannot be expected to remember all specific names. ( A.J. Cain, 1958)Later part of 19th century: new information technologies developed which permitted the easy storage and coding of larger amounts of data than could previously be easily manipulated. (Chandler,1977),(Yates,1989)
- 33. Development of ClassificationExample of classificationPaper-based archival practice.Issues: hard to reclassifiedType specimen had to be relocated physicallySo do Series of articles or books
- 34. Development of ClassificationExample of classificationMultifaceted classification systemImprove: Enabling the classifications to be ordered in multiple ways, rather than in a singleExample: A collection of books might be classified using an author facet, a subject facet, a date facet
- 35. Development of ClassificationExample of classificationHierarchical classification (for reading the past)E.F. Codd In the early 1970sSplit physical storage of data in the computer and the representation of that data.Disadvantage: becomes awkward to introduce other levels of taxonomic category as an afterthought.Improve method: one record for every name, regardless of its taxonomic level
- 36. CladisticsDefinitionIt is a method of classifying species of organisms into groups called clades, which consist of 1) all the descendants of an ancestral organism and 2) the ancestor itself.Features : Give a more regular algorithm for determining phylogenyFocusing attention on shared, derived characteristics of set organismsUsing ‘outgroup’ comparisons to develop the classification system
- 37. CladisticsTree of lifeCladists use cladograms, diagrams which show ancestral relations between taxa, to represent the evolutionary tree of lifeCharles Darwin (1809–1882) was the first to produce an evolutionary tree of life
- 39. CladisticsComputer programs in cladisticsUndertaken using Swofford’s (1985) package PAUP version 2.4installed on a Cyber mainframe computer and version 2.4.1 on an amstrad 1512 PCDavid Swofford’s PAUP is a software package for inference of evolutionary treesPurpose: follow a given algorithm for generating and testing cladograms
- 40. CladisticsComputer programs in cladistics
- 41. CladisticsComputer programs in cladisticsIssues:The packages produce variable results and cannot possibly look at all the possibilities, since there is NP-complete problem.Algorithm issues
- 42. The FutureStore the lifeLife is described as itself a program, with DNA being code.IF everything is information, then life can equally well be “stored”
- 43. THANK YOU !