18176
Reporter: feedback bot
Type: Feedback
Summary: Problems with data search
Status: Open
Created: 2016-01-25 16:25:19.39
Updated: 2017-10-05 15:21:00.668
Description: 1. The "configure" and " filter" popdowns allow "Collector" or "Collector name" but the data result display shows "Recorded by." Please standardize your labels! They should all be "Collector
2. There is no filter nor configuration possibility for "Collector number," nor does it show in the result display. That needs to be included everywhere!!!
3. None of the collector names ("Recorded by"), for example "Hammel, Barry," Grayum, Michael, H.," etc. can be searched on successfully; they don't show up in the possibilities when the surname is entered in the "Collector" field of add filter, and all of the the likely *possible* searches give zero results! I.e., searches ("filters") on Collector name do not work, even though the option is allowed.]]>
Author: cgendreau
Created: 2016-01-25 17:04:19.844
Updated: 2016-01-25 17:04:19.844
1. The correct DarwinCore field is "Recorded by" but I would guess people are most familiar with "Collector".
2. DarwinCore does not define any fields such as "Collector number" but maybe other protocol could use it. see http://dev.gbif.org/issues/browse/POR-1550
3. I am able to search on Recorded By. e.x. http://www.gbif.org/occurrence/search?TAXON_KEY=2761715&COUNTRY=CR&RECORDED_BY=Hammel%2C+Barry&INSTITUTION_CODE=INB
Of course when surname is used it is different since it is a "different" name. This would require so index the field differently.
Author: mdoering@gbif.org
Created: 2016-01-25 23:44:19.592
Updated: 2016-01-25 23:44:19.592
DwC uses recordNumber instead of collector number:
http://rs.tdwg.org/dwc/terms/index.htm#recordNumber
You can search on it, for example:
http://www.gbif.org/occurrence/search?RECORD_NUMBER=M781
Dwc replaced the notion of a collecting with the broader idea of a recording occurrences. Collector mostly applies to specimens, while recording better fits the majority of observation records.
Author: cgendreau
Created: 2016-01-26 10:03:06.442
Updated: 2016-01-26 10:03:06.442
To handle multiple values inside recorded_by field we could give Solr RegularExpressionPatternTokenizer a try.
The problem is that even if DarwinCore recommended best practice is to use vertical bar (' | ') to separate different names it is not always used. Maybe we could start with, "|" and ";".
Author: hoefft
Comment: naming is normalized on the site. The table display part is captured here: https://github.com/gbif/portal16/issues/578 .
Created: 2017-10-05 15:21:00.668
Updated: 2017-10-05 15:21:00.668