One key challenge in multi-document summarization is to capture the relations among input documents that distinguish between single document summarization (SDS) and multi-document summarization (MDS). Few existing MDS works address this issue. One effective way is to encode document positional information to assist models in capturing cross-document relations. However, existing MDS models, such as Transformer-based models, only consider token-level positional information. Moreover, these models fail to capture sentences' linguistic structure, which inevitably causes confusions in the generated summaries. Therefore, in this paper, we propose document-aware positional encoding and linguistic-guided encoding that can be fused with Transformer architecture for MDS. For document-aware positional encoding, we introduce a general protocol to guide the selection of document encoding functions. For linguistic-guided encoding, we propose to embed syntactic dependency relations into the dependency relation mask with a simple but effective non-linear encoding learner for feature learning. Extensive experiments show the proposed model can generate summaries with high quality.

Equipped with a state-of-the-art thermal camera, the drone crisscrossed the field, scanning it for cool, soggy patches where a gopher may have chewed through the buried drip irrigation line and caused a leak. In the drought-prone West, where every drop of water counts, California farmers are in a constant search for ways to efficiently use the increasingly scarce resource. Pictured above, Danny Royer, vice president of technology at Bowles Farming Co., prepares to pilot a drone over a tomato field near Los Banos, Calif. Farmers say leak-detecting drones can help save massive amounts of water. The video camera is paired up with a smartphone or computer tablet, which is used to control the drone.