Loading...
Please wait, while we are loading the content...
Similar Documents
Generalised Prioritisation: A New Way of Combining Similarity Metrics (1999)
| Content Provider | CiteSeerX |
|---|---|
| Author | Bridge, Derek Ferguson, Alex |
| Abstract | We describe similarity metrics, which are a generalisation of similarity measures having any partial order as their result type. We then present the main result of this paper by describing generalised prioritisation, a new way of combining similarity metrics. Generalised prioritisation uses an indifference relation on the degrees of similarity. We show that the use of the indifference relation gives a (very rough) analogue of the use of weights in more traditional, numeric-valued similarity measures. Similarity Metrics Similarity measures are used, e.g., in the retrieval phase of case-based reasoning systems and in forming `clusters' of objects in machine learning systems. More recently, they have been used in searching product catalogues in ecommerce systems. In the main, similarity measures have been binary operators that, when applied to two objects of type ff, return a number, usually a real from [0; 1], denoting their degree of similarity. That is, their type is most usually ff !... |
| File Format | |
| Publisher Date | 1999-01-01 |
| Access Restriction | Open |
| Subject Keyword | Indifference Relation Partial Order Generalised Prioritisation Binary Operator Ecommerce System Numeric-valued Similarity Measure Similarity Metric Similarity Measure New Way Retrieval Phase Product Catalogue Similarity Metric Type Ff Similarity Measure Combining Similarity Metric Result Type |
| Content Type | Text |
| Resource Type | Article |
