Completing missing views for multiple sources of web media

Abstract

Combining multiple data sources, each with its own features, to achieve optimal inference has received a lot of attention in recent years. In inference from multiple data sources, each source can be thought of as providing one view of the underlying object. In general, different views may provide complementary information for the inference task. However, often not all the views are available all the time for the available instances in an application. In this paper, we propose a view completion approach based on canonical correlation analysis that heuristically predicts the missing views and further ranks all within-view features, through learning the intrinsic correlation among the views from training set. We evaluate our approach and compare it with existing approaches in the literature, using web page classification and photo tag recommendation as case studies. Experiments demonstrate the improved performance of the proposed approach. The results suggest that the work has great potential for inference problems with multiple information sources.

BioinQA: metadata-based multi-document QA system for addressing the issues in biomedical domain

“…Collection and integration of information from multiple data sources has recently attracted a lot of attention. Subramanya et al (2008) discuss a view completion approach for multiple sources of web media based on canonical correlation analysis that heuristically predicts the missing views and also ranks all within-view features. They have shown the suitability of such an approach through the experiments on the web page classification photo tag recommendation.…”