basisOfRecord from observations

[tucotuco]

If a researcher walks around making observations in the field and takes photos to accompany some of those observations now and then, what should be the basisOfRecord?

[tucotuco]

Following the consensus reached in the discussion about the basisOfRecord for Occurrences found in literature (#62), the basisOfRecord should always reflect the evidence upon which the Occurrence record was based. As such, the researcher's record that have photographs should have basisOfRecord=MachineObservation, and the rest should have basisOfRecord=HumanObservation.

[MattBlissett]

I thought MachineObservation was only where the machine (in some automated way) gathered the observations: a camera trap, remote sensing, DNA analysis of a sample etc.

Searching for the definition, this is the third result: gbif/portal-feedback#681 (comment) (!).

[tucotuco]

It's exactly because of this duality that I wanted to have something concrete and unambiguous. The literature discussion seemed to solidify that - the basisOfRecord should reflect the best evidence it left behind. For the photo, by camera trap or otherwise, the evidence is captured by the machine. Without the media, we have only the HumanObservation. It seems useful to be able to distinguish between having some media to show the evidence, and not having that. The intention really didn't matter so much as that. So, I think this posture is defensible, whereas the one in gbif/portal-feedback#681 (comment) is not.

More comments welcome.

[Jegelewicz]

the basisOfRecord should reflect the best evidence it left behind. For the photo, by camera trap or otherwise, the evidence is captured by the machine. Without the media, we have only the HumanObservation.

This makes perfect sense to me and is definitely less ambiguous.

[dagendresen]

You may also flip the argument and say that "without a human presence and intent of making the species observation, all you have is the MachineObservation...". Maybe informative to ask: Whose intention was it to make the observation? The human or the machine? If the human reports her HumanObservation and later attach a photo as (supplementary) evidence does this "transform" the observation to a MachineObservation? If somebody else (machine or human, etc) post a photo of the human making her HumanObservation on the Internet - does this "transform" the HumanObservation to a MachineObservation? If the photo is out of focus and blurred is the photo still always of a "higher level" of evidence than the observation reported by the human? Maybe informative to ask: Who would you credit for the observation (dwciri:recordedBy) - The identification number of the machine? or the ORCID of the human?