TUNA is a research project funded by the UK's Engineering and Physical Sciences Research Council (EPSRC). Natural Language Generation programs generate text from an underlying Knowledge Base. It can be difficult to find a mapping from the information in the Knowledge Base to the words in a sentence, for example, when the Knowledge Base uses ‘names’ such as ‘#Jones083’ that a hearer/reader does not understand, or has concepts which do not have their own names. (e.g., a specific tree or a chair). In all such cases, the program has to “invent” a description that enables the reader to identify the referent.
Existing algorithms tend to focus on one particular class of referring expressions, for example conjunctions of atomic or relational properties (e.g., ‘the black dog’, ‘the book on the table’). Our research is aimed at designing and implementing a new algorithm generates appropriate descriptions in a far greater variety of situations. The algorithm will be more complete, and generate expressions that are more appropriate because it will be based on empirical studies involving corpora and controlled experiments. Thus the project combines (psycho)linguistic, computational and logical challenges.