The Experiments with the Linear Combination Data Fusion Method in Information Retrieval

In data fusion, the linear combination method is a very flexible method since different weights can be assigned to different systems. However, it remains an open question that which weighting schema is good. In many cases, a simple weighting schema was used: for a system, its weight is assigned as its average performance over a group of training queries. In this paper, we empirically investigate the weighting issue. We find that, a series of power functions of average performance, which can be implemented as efficiently as the simple weighting schema, is more effective than the simple weighting schema for data fusion. We also investigate combined weights which concern both performance of component results and dissimilarity among component results. Further performance improvement on data fusion is achievable by using the combined weights.