An impact ordering approach for indexing fuzzy sets

We propose an approach for indexing fuzzy data based on inverted files that speeds up retrieval considerably by stopping the traversal of postings lists early. This is possible because the entries in the postings lists are organized in a way that guarantees that there are no matching items beyond a certain point in a list. Consequently, we can reduce the number of false positives significantly, leading to an increase in retrieval performance. We have implemented our approach and evaluated it experimentally, including a test on skewed and real-world data, comparing it to an approach that has previously been shown to be superior to other methods.