The advent of networking, information and imaging technologies has enabled us to better understand the complex world we live in and the interconnected society we work in. Large data sets, either structured or unstructured, are being generated dynamically from monitoring sensors and devices; heterogeneous sources of opinions, judgments, values, beliefs, and facts; and diverse systems including those involving information retrieval, environmental scanning, social networks, multimedia, modeling and simulation, and other scoring and ranking systems. For these huge amounts of data to be useful, they have to be analyzed. Significant patterns have to be identified, information from various sources and systems has to be fused, and meaningful and useful knowledge must be extracted for action and decision making.
Information fusion plays a crucial role in the pipeline of scientific discovery through processes of data acquisition, information integration, and knowledge discovery. Although methods for fusion of data or decision such as Borda count and plurality voting have been used since the 1770s, it remains a challenging problem to understand when, what, and how to best fuse information. In particular, we are interested in the following two types of problems:
(I) Given two (scoring) systems A and B, when and how to best combine A and B in order to improve the performance?
(II) Given a complex problem and many possible (scoring) systems for its solution, what is the optimal number of systems for fusion and how to fuse them?
The solution to the above problems (I) and (II) relies on parameters like the number of systems, the performance of systems, and the diversity between systems.
It can be shown in some contexts that the fused system C(A,B) from systems A and B is better than each of the individual systems only if each of these individual systems performs well and they are diverse. However, to understand when this might happen, the concept of diversity has to be well defined in particular for decision fusion and variable (feature, attribute, cue, indicator, and parameter) combination. Another issue of great importance is the performance variation between score combination and rank combination.
The Workshop WAIF will provide a forum for researchers and practitioners to mingle with and learn from each other on the design, analysis, and implementation of algorithms for information fusion, whatever the type of information to be fused: logical, ranking, scoring, etc. This gathering of multidisciplinary experts and researchers will be expected to conceive new ideas and create novel solutions. WAIF will address, among others, the following topics and directions: