AI RESEARCH
Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis
arXiv CS.CL
•
ArXi:2605.20998v1 Announce Type: new Aspect-Term Sentiment Analysis (ATSA) in multi-aspect sentences faces a fundamental tradeoff between efficiency and expressiveness. Existing models either re-encode the sentence for each aspect or rely on static use of deep representations, leading to redundant computation and limited adaptivity. We argue that Transformer depth is a costly, queryable resource, and propose DABS, a single-pass inference framework that encodes each sentence once to construct a reusable, depth-ordered substrate.