Abstract | ||
---|---|---|
In natural language generation, the use of a lexicalized grammar formalism and incremental syntactic and semantic processing places strong and specific constraints on the form and mean- ing of grammatical entries. These principles restrict which grammatical representations are possible and suggest examples an analyst can consult to decide among possibilities. We dis- cuss and justify a number of such constraints, and describe how they inform the design of lexi- cal entries for motion verbs. Our entries allow a generator to match the lexical choices found in a target corpus of action descriptions by assessing how the interpretation of a verb in context contributes towards the hearer's identification of the intended action. |
Year | Venue | Keywords |
---|---|---|
2000 | TAG+ | semantic processing |
Field | DocType | Citations |
Natural language generation,Semantic memory,Verb,Grammar formalism,Computer science,Grammar,Natural language processing,Artificial intelligence,Linguistics,Syntax,restrict | Conference | 4 |
PageRank | References | Authors |
0.66 | 11 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Matthew Stone | 1 | 887 | 118.99 |
Tonia Bleam | 2 | 44 | 10.44 |
Christine Doran | 3 | 183 | 22.08 |
Martha Stone Palmer | 4 | 5566 | 511.07 |