Tools needs filtering

[gbif-portal]

Tools need filtering

---

Github user: @dschigel User: See in registry System: Chrome 81.0.4044 / Windows 10.0.0 Referer: https://www.gbif.org/resource/search?contentType=tool Window size: width 1730 - height 838 API log&_a=(columns:!(_source),index:'prod-varnish-',interval:auto,query:(query_string:(analyze_wildcard:!t,query:'response:%3E499')),sort:!('@timestamp',desc))) Site log&_a=(columns:!(_source),index:'prod-portal-',interval:auto,query:(query_string:(analyze_wildcard:!t,query:'response:%3E499')),sort:!('@timestamp',desc))) System health at time of feedback: CRITICAL

[dschigel]

As discussed @andrewrodrigues

[MortenHofft]

@dnoesgaard could you please give recommendations on what filters are meaningful given the data managed in Contentful?

[dnoesgaard]

None, to be honest. These aren't curated at all. At best, they have a title, a description and a link...

[dschigel]

Please consult with @andrewrodrigues regarding filter and sorting needs, as some of the tools will be useful for data use training, others not. Also with @laurarussell re: possible tools relevant for data publishing, there will be some overlap. In addition to what-they-do filters, we can think about the environment (R, Python) and file format filters (xls, FASTA).

[dnoesgaard]

Ok, as long as we're clear that we currently don't have any means of filtering these. Someone would have to go through all 100+ tools and populate relevant fields for the suggested filters. I'm not saying we can't do this, I'm just saying it's quite a task.

One field that we could add that is already in use in other content type is "purpose" with these options:

• Data curation & quality
• Data publishing
• Data analysis
• Capacity enhancement
• Data digitization
• Data access

So perhaps we start there?

[dschigel]

I am not sure if Andrew and Laura agree, but I would actually move away from internal GBIF jargon, and get something that it clear and less abstract in categories. For example:

Action: cleaning, georeferencing, identification, validation Data type: occurrence, sequence, checklist, name File format: FASTA, csv, txt

Choosing one or more of such filter values would reconstruct user needs. Nobody I know comes to look for a tool for "data analysis" or "capacity enhancement", this seems to be too general, but categories be can useful in addition to "action filters" I am suggesting: the purpose here is not to systematize what we have, but to give users efficient mean to find the tool that matches the need. The needs are typically very practical: identify-sequence-FASTA, clean-occurrence-csv, georeference-occurrence-csv, validate-name-txt etc. Inputs from Laura's and Andrew's course experiences, also from helpdesk @ahahn-gbif would be welcome.

I suggest we try to make this filters to respond to real user needs and to mimic actual UX, to me this is more important than consistency with internal GBIF terminology. Search is already doing some of this job, providing that you know what you are looking for (try "Python"). Filters show the options and work in more inviting way, I think. Because going through a 100 tools will indeed take time, I suggest we do general categories and action filers at the same time.

[dnoesgaard]

I am not sure if Andrew and Laura agree, but I would actually move away from internal GBIF jargon, and get something that it clear and less abstract in categories.

Ok, sounds good to me. I hear all your points and I agree 100% that it could be useful.

How about I make a shared sheet with all tools in the catalogue and then we can all collaborate on populating the suggested categories based on our knowledge and needs?

[dschigel]

@dnoesgaard sounds like a perfect start to me. I am happy to take part in the action. A hakathon at the lawn, with laptops and... inspiration?

[dnoesgaard]

https://docs.google.com/spreadsheets/d/1YsKBbDCfRtpNjhVUTMM1VZVWRVwxk6wfubBhTmM6TvI/edit?usp=sharing

Anyone with a GBIF.org G suite account can access. Otherwise, just request access...

[dschigel]

Thanks Daniel, I started experimenting with the first three lines with action filters, leaving broader categories for later