Paper: Using Integer Linear Programming for Content Selection, Lexicalization, and Aggregation to Produce Compact Texts from OWL Ontologies

ACL ID W13-2106
Title Using Integer Linear Programming for Content Selection, Lexicalization, and Aggregation to Produce Compact Texts from OWL Ontologies
Venue European Workshop on Natural Language Generation
Session
Year 2013
Authors

We present an Integer Linear Program- ming model of content selection, lexical- ization, and aggregation that we devel- oped for a system that generates texts from OWL ontologies. Unlike pipeline archi- tectures, our model jointly considers the available choices in these three text gen- eration stages, to avoid greedy decisions and produce more compact texts. Experi- ments with two ontologies confirm that it leads to more compact texts, compared to a pipeline with the same components, with no deterioration in the perceived quality of the generated texts. We also present an ap- proximation of our model, which allows longer texts to be generated efficiently.