Aggregation gisting

There are often multiple approaches to aggregating entity data. This paper presents an approach to creating a composite, or gisted, aggregation from these multiple aggregation techniques. Approaches are presented for merging aggregation results that provide either crisp or fuzzy aggregations. The approach is based on determining a strength metric for the linkage between any two entities in the composite aggregation by comparing the relationship patterns that the entities have with other entities. The more frequently two entities have the same pattern of relationships; the more similar they are thought to be and the more strongly they are linked. In addition, an approach to de-cluttering the presentation of the composite aggregation is presented. This de-cluttering allows the user to focus on different aspects of the composite aggregation.