Simple Attention-Based Representation Learning for Ranking Short Social Media Posts
This addresses the problem of efficient and accurate ranking for social media search, though it appears incremental as it builds on existing neural methods with attention enhancements.
The paper tackles ranking short social media posts for user queries by proposing a simple word-level Siamese architecture with attention mechanisms, achieving better effectiveness and faster performance than more complex existing approaches on TREC Microblog Tracks datasets.
This paper explores the problem of ranking short social media posts with respect to user queries using neural networks. Instead of starting with a complex architecture, we proceed from the bottom up and examine the effectiveness of a simple, word-level Siamese architecture augmented with attention-based mechanisms for capturing semantic "soft" matches between query and post tokens. Extensive experiments on datasets from the TREC Microblog Tracks show that our simple models not only achieve better effectiveness than existing approaches that are far more complex or exploit a more diverse set of relevance signals, but are also much faster. Implementations of our samCNN (Simple Attention-based Matching CNN) models are shared with the community to support future work.