Automatic summarising and the CLASP system

This dissertation discusses summarisers and summarising in general, and presents a new summarising system, clasp. In chapters 1−3, I present a framework for thinking about summarisers in terms of context factors and the three stages of analysis, condensation and synthesis. I look at previous research in automatic summarising and identify four main directions that have been taken. I consider how summarising systems may be and have been evaluated. clasp, described in chapters 4−7, takes a new approach based on a shallow semantic representation of the source text as a predication cohesion graph. Nodes in the graph are simple predications corresponding to events, states and entities mentioned in the text; edges indicate related or similar nodes. Summary content is chosen by selecting some of these predications according to criteria of importance, representativeness and cohesiveness. These criteria are expressed as functions on the nodes of a weighted graph. Summary text is produced either by extracting whole sentences from the source text, or by generating short, indicative summary phrases from the selected predications. clasp uses linguistic processing but no domain knowledge, and therefore does not restrict the subject matter of the source text. It is intended to deal robustly with complex texts that it cannot analyse completely accurately or in full. Chapter 8 describes experiments in summarising stories from the Wall Street Journal. The results suggest that there may be a benefit in identifying important material in a semantic representation rather than a surface one, but that, despite the robustness of the source representation, inaccuracies in clasp’s linguistic analysis can dramatically affect the readability of its summaries. In chapter 9, I suggest ways in which clasp could be modified to overcome this and other problems.