Planned Transformation of the Identification System
Summary
The transformation of the izeltlabuak.hu identification system has been in progress for several years. The current system only allows experienced identifiers to provide opinions on observations, but the planned new system will enable all users to participate in the identification process. This will be similar to the iNaturalist approach, but will prioritize the reliability of identifications. During development, the observation display interfaces have been redesigned, and changes are also planned in the taxonomic system database structure. The gradual introduction of the new system is expected once the dependencies are resolved.
Current Identification Process
Currently on izeltlabuak.hu, users can make suggested identifications for recorded observations. The identifications are performed by experienced users with identifier privileges. A separate interface is available for identifiers on observation displays. Here they can approve the suggested identification or make a new identification, where in addition to specifying the taxon, they can provide other data such as developmental stage, sex, or comments about the identification.
Identification functions
According to the current system, one valid identification belongs to one observation. This also means that an identifier can correct a previously submitted identification. To avoid conflicts, I have formulated the following rule: "In case of disagreement with another identifier, identifiers must handle the situation constructively through cooperation. It may happen that different identifiers give different identifications for a finding. In such cases, they should set the narrowest common taxonomic group and justify the decision in the description. When overriding, the decision should also be justified whenever possible."
Identification interface
Planned Transformations
The current identification system has limitations. I consider it better to have an approach where every user can give their opinion on what they think is visible in the image. The current identifications will be transformed accordingly, and multiple identifications will be possible for one observation. The experience level of identifiers will be taken into account during identifications, and final identifications will be calculated accordingly.
This approach is similar to the iNaturalist identification system in that anyone can provide an identification, but the planned identification system will differ in several ways, for example, by prioritizing the reliability of identifications.
Several developments are planned for the new system. It is conceivable that these will not all be implemented, or not in this form. The plans include, among other things, the following elements:
- currently, taxonomic identification, developmental stage, and sex can be submitted simultaneously during identification. The system would handle these separately, and their submission could be initiated separately.
- the identifier could specify their confidence level regarding the submitted identification during the identification process.
- I would introduce for identifiers that their expertise level applies to the specific taxonomic group in which they are considered experienced.
- it would be possible to indicate if a submitted identification is incorrect.
- there would be multiple aggregated identifications for one finding. One expert and one community identification. The two may differ from each other.
- submitted identifications could be reviewed in an aggregated view.
- users would be considered experienced identifiers according to a rethought system.
Development Process
This transformation requires fundamental changes to the izeltlabuak.hu system. Its planning and preparation have been ongoing for several years. The plans for the new identification system are largely complete. As part of eliminating the main dependencies of the transformations, I need to implement the general display interfaces for observations so that they can uniformly handle the submission and display of new identifications.
Most of the generalization was implemented during 2023. The following were redesigned:
- taxonomic overview pages
- pages that help review findings
- pages for recording and editing new findings
- observation pages, e.g., Varga Nimród's finding: Curculio propinquus
- finding group listings, e.g., Fráter Szabolcs's finding groups
- and finding group subpages, e.g., Fráter Szabolcs's moth finding group.
- Event pages were introduced, e.g., moths recorded during European Moth Nights 2023.
Currently, the renewal of user-specific taxon lists and observation overview subpages is underway, followed by user notification pages about identifications and identifier identification overview pages.
I also plan transformations in the taxonomic system database structure. The current solution is extremely flexible. It can handle synonyms and name changes. The Hungarian species lists can also be easily modified with it. Published sources are assigned to species names. These are easily traceable and help verify the authenticity of names. Modifying these sources is also simply solved. Nevertheless, it has its own disadvantages. With a planned transformation, I could make it even more flexible.
Once these are completed, the implementation and gradual introduction of the new identification system will follow.
