dev: use CiTO (ontology) for paper facts and citations

[geritwagner]

CiTO provides an ontology that standardizes facts and citation types. For example, citations between papers can be stated based on record IDs, they can be qualified, and analyzed for facts (like retract status). Since these fields are not standardized in BiBTex, a separate CiTO/rdf document should be introduced. Qualifying citations would also be useful for citation searches (selection) and analyses.

Explore data structures for citation relations (between records and beyond the sample), the retrieval of structured data, the codification and collaborative editing of graph-data; define standardized fields (e.g., cited_by: number or references? cited_by_file (cited_by (number) and cited_by (backward search) write to the same field), semantic relations like “corrects paper X”, “tests paper X”, or codings like “empirical”/”meta-analysis-ready”)

Open questions:

• Is CiTO the best option/what are the alternatives?
• If data is added to the ontology in the search/prep operations, how do we handle potential changes in record IDs (set_ids)?
• Should the ontology be part of curations/local_index? How can we implement provenance and correction paths?
• Should the ontology be included by default or optional (e.g., for citation analyses)?
• The facts could be stated based on origins.

Generally, we should avoid separate and redundant data structures. One possibility would be to add a dictionary/json to the bib file (REFERENCES/CITATIONS field).