BioGuide - Project - Preference Criteria

BioGuide Project





Related Projects





Use BioBrowsing


Use BioGuideSRS


Use BioGuide


Restricted area

User documentation


Querying process
Configuration file








The conducted interviews reveal that many criteria determine preferences.
We give below the list of the main preference criteria collected and (when ambiguous) the definition given by the interviewee as well as the percentage of interviewees (30 persons) who mention the criteria.
Our aim is to emphasize the variety of criteria determining preferences among users. We do not aim to provide a classification of these criteria or give a consensual definition of each of them either.

Please, consult the user manual to see how to express preferences in BioGuide and BioGuideSRS.

General Criteria

For some users, the data collected within sources have to be
  • Reliable, Accurate (93%)
  • Complete (90%)
  • Precise (53%)
  • Free (available without having to pay for accessing them) (15%)
  • Non-redundant (60%)
  • Value-added (beneficial and provide advantage from their use) (7.5%)
  • Diverse (60%)
  • Objective (unbiased, unprejudiced and not directly associated to expertise) (6.6%)

Criteria related to the usability of data, Source Capability

For some users, a given source is used rather than another because of it is:

  • Easy-to-use, Easy-to-manipulate (90%)
  • Unambiguous (93%)
  • Easy-to-question (60%)
  • Well-documented (83%)
  • Has a secure access (queries sent to the source are not known by the administrator of the resource) (83%)
  • Has a capacity of providing results fast (responsiveness) (60%)

Criteria related to representation of data

For some users, a given source is used rather than another because it is:

  • Concise
  • (6.6%)
  • Well-formatted (6.6%)
  • Well-presented (7.5%)
  • Understandable, Interpretable (provided in appropriate language, and units and data definition are clear) (7.5%)

Criteria related to Traceability and Time

For some users, a given source is used rather than another because it is:
  • Easily traced (90%)
  • Verifiable (90%)
  • Often Updated (60%)

Several "sub-criteria" can be used to determine these points such as:

  • History (the history of manipulation applied to data is provided)
  • Currency (the time interval between the last update and the time it is used)
  • Volatility (frequency of data changes: measure of information instability)

Laboratoire de Recherche en Informatique, University of Pennsylvania

Contact us