|
|
|
|
|
||
|
|
|
Semantic searches have been posited to significantly improve the capabilities of scientists to locate ecological and environmental data that are relvant to their particular purposes. This page provides some example searches that would be useful to be able to express and execute in the SEEK SMS system.
Current metadata searchesCurrent metadata-driven search systems allow searches against EML to retrieve relevant documents. Such searches include simple keyword matching against the full text of the metadata document, against particular metadata fields (e.g., to find the Creator of a data set), and compound searches. The compound searches use the structure of the metadata document to create a search based on space, time, biological taxonomy, or other criteria that are provided in the metadata content. Some example searches one can conduct include:
Semantic searchesThe above searches are good, but tend to not be particularly precise nor very complete. A search for 'soil' in the KNB produces over 3000 data sets, many of which may or may not be relevant to a study on, for example, soil chemistry. On the other hand, a search on 'soil chemistry' might only match a few data sets, because that term only occurs in a few metadata records. Instead, relevant meatdata records might include terms such as 'acidity', 'salt content', 'Nitrogen', 'Potassium', 'K', 'N', etc. A semantic search system that allowed one to recognize that 'N' and 'Nitrogen' are the same concept, and that they are both related to 'SoilChemistry' would be a huge improvement. In addition, scientists are frequently searching for data with particular measurements have been made, such as 'Density estimates for abalone' or 'Solar radiation at 1m above ground'. These searches are not possible with the current metadata-based search system, and instead require a far more semantically-aware search system. Some of the searches that would be useful include (an example dataset that should match the query is included in parentheses):
Hopefully these represent a good discussion starter on semantic searches.
|
| This material is based upon work supported by the National Science Foundation under award 0225676. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF). Copyright 2004 Partnership for Biodiversity Informatics, University of New Mexico, The Regents of the University of California, and University of Kansas |
