Distance-based Positive and Unlabeled Learning for Ranking
This addresses ranking tasks in domains like vertex nomination in graphs, where supervision is light, but it appears incremental as it builds on existing ranking methods with a specific supervision approach.
The paper tackles the problem of learning to rank items with minimal supervision, using only a target item and a small set of similar items, and demonstrates that combining representations via an integer linear program is effective in simulations and real data.
Learning to rank -- producing a ranked list of items specific to a query and with respect to a set of supervisory items -- is a problem of general interest. The setting we consider is one in which no analytic description of what constitutes a good ranking is available. Instead, we have a collection of representations and supervisory information consisting of a (target item, interesting items set) pair. We demonstrate analytically, in simulation, and in real data examples that learning to rank via combining representations using an integer linear program is effective when the supervision is as light as "these few items are similar to your item of interest." While this nomination task is quite general, for specificity we present our methodology from the perspective of vertex nomination in graphs. The methodology described herein is model agnostic.