Efficient algorithm for Context Sensitive Aggregation in Natural Language generation
Hemanth Sagar Bayyarapu · 2011
Aggregation is a sub-task of Natural Language Generation (NLG) that improves the conciseness and readability of the text outputted by NLG systems. Till date, approaches towards the aggregation task have been predominantly manual (manual analysis of domain specific corpus and development of rules). In this paper, a new algorithm for aggregation in NLG is proposed, that learns context sensitive aggregation rules from a parallel corpus of multi-sentential texts and their underlying semantic representations. Additionally, the algorithm accepts external constraints and interacts with the surface realizer to generate the best output. Experiments show that the proposed context sensitive probablistic aggregation algorithm performs better than the deterministic hand crafted aggregation rules. • Sentence-Planning: Generates referring expressions, combines multiple propositions, selects appropriate lexical items and syntactic structures for each (aggregated) proposition and adds cohesion devices (eg, discourse markers) to make the text flow smoothly. • Surface-Realizer: Converts the lexicalized linguistic structure into a linearized string while ensuring grammaticality, proper punctuation, correct morphology. 1