{"id":"intent-with-incomplete-maps","title":"Interpreting intent when the map is incomplete","regions":["computation","language-and-experience"],"question":"Can a geographic interface translate intent where labels and place knowledge are incomplete?","formulation":"A practical interface is imagined that interprets abbreviated place descriptions and handles places known to users but absent from the map.","representations":[{"name":"Literal label matching","affords":"Keeps the interpretation close to available labels.","loses":"Misses abbreviations, colloquial names, and unmapped places.","epistemic_kind":"reconstruction"},{"name":"Intent interpretation with candidate locations","affords":"Offers alternatives beyond exact labels.","loses":"An inferred coordinate may appear more certain than the evidence warrants.","epistemic_kind":"reconstruction"}],"tension":"A useful guess can become a false location when uncertainty is hidden.","challenge":"How should the interface distinguish an abbreviation expansion from an unsupported geographic inference?","development":"independent_branch","epistemic_kind":"reconstruction","evidence":["basis-intent-with-incomplete-maps"],"next_move":{"text":"Create fictional incomplete maps with ambiguous descriptions and evaluate candidate ranking, uncertainty, and correction.","epistemic_kind":"model_proposed"},"open_questions":["Can a geographic interface translate intent where labels and place knowledge are incomplete?","How should the interface distinguish an abbreviation expansion from an unsupported geographic inference?"],"formal_status":"open_problem","version":1,"content_hash":"6c0e74fe40720ba9da3320aefffc2dc2494e21a32c8900a23e18c65ad5370c68"}