Parsing and interpreting ambiguous structures in spatial hypermedia