Feature
| Federated(PKP Harvester) | Built-in |
Implied boolean operator between words
| AND | OR (default, configurable)
|
Full-text search
| No (unless full-text is included in descriptive metadata)
| First 10000 terms in document(configurable)
|
Search
multiple sites simultaneously | yes | no |
Indexed fields
| Believed to be all dublin core fields in current implementation...verifying.
| dc.contributor.*,
dc.creator.*,
dc.title.*,
dc.subject.*,
dc.description.abstract,
dc.description.statementofresponsibility,
dc.relation.ispartofseries,
dc.description.tableofcontents,
dc.format.mimetype,
dc.description.sponsorship,
dc.identifier.*,
dc.language.iso
(default, configurable - changed on main DRC site so all dc.language variants are indexed to give results for example 'French' search; John is in the process of adding dc.description on main drc site)
|
Thumbnails in results
| yes(customization, not currently functional on all repositories) | yes |
Able to search repositories other than DSpace
| yes (prototyped with CONTENTdm, Fedora repositories - may require significant maintenance work)
| no |
Results display
| title, creator,thumbnail or logo, approximately first 200 characters of description,repository name (configurable). Theme is currently OhioLINK generic from all sites.
| title, contributors, insertion date (default, configurable). Themable to host repository.
|
Default results ordering
| researching...(appears to count total number of occurrences of search terms in record, which biases search results in favor of records with long abstracts(ETDs, for example)
| researching... |
Result sort options
| none enabled
| relevance/title/submit date/issue date (ascending/descending)
|
Maximum size of result set
| Limit to 4000 results per search term, result sets are then merged( configurable)
| Not limited
|
Time required for addition of new items to the index
| Configurable - current suggestion is to harvest new metadata every 24 hours
| immediate |
Language Support
| researching... | researching... |
Metadata Schema
| dublin core , mods, marc built-in. Can be extended and crosswalked.
| Metadata schema that can be added to DSpace - so any flat (non-hierarchical) schema, qualifiers are allowed
|
Automatic Spelling Corrections
| no | no |
Highlight search terms in results
| Not currently, but this should be simple to implement
| no |
Suggest controlled vocabulary term to searcher
| no | no |
Maintenance costs
| | |
| Performance | | |
Scope Limits
| Limit to a single archive
| Limit to a single collection
|
Results per page
| 10 (configurable)
| 10
|
Post-search faceting
| no | no |
| Wildcards | * to match sequence of characters
| * to match sequence characters
|
Advanced search
| specify any dublin core field, archive. "Language" field populated with current choices, but choices are not controlled.
| specify one of 8 indices or keyword, conjunction, sort order
|
Allow user to revise search
| Advanced search only
| no |
Grouping of words
| Use parentheses
| no (or broken)
|
Exact Phrase Search
| Use double quotation marks
| Use double quotation marks
|
Availability of statistics
| | |
word stemming (sail v. sails v. sailing)
| no | yes |
stop words
| yes - researching list
"the"
| yes - researching list
"the","and","or"
|
case-sensitive
| no | no |
