In the real-world question answering scenarios, hybrid form combining both tabular and textual contents has attracted more and more attention, among which numerical reasoning problem is one of the most typical and challenging problems. Existing methods usually adopt encoder-decoder framework to represent hybrid contents and generate answers. However, it can not capture the rich relationship among numerical value, table schema, and text information on the encoder side. The decoder uses a simple predefined operator classifier which is not flexible enough to handle numerical reasoning processes with diverse expressions. To address these problems, this paper proposes a \textbf{Re}lational \textbf{G}raph enhanced \textbf{H}ybrid table-text \textbf{N}umerical reasoning model with \textbf{T}ree decoder (\textbf{RegHNT}). It models the numerical question answering over table-text hybrid contents as an expression tree generation task. Moreover, we propose a novel relational graph modeling method, which models alignment between questions, tables, and paragraphs. We validated our model on the publicly available table-text hybrid QA benchmark (TAT-QA). The proposed RegHNT significantly outperform the baseline model and achieve state-of-the-art results. We openly released the source code and data at https://github.com/lfy79001/RegHNT (2022-05-05).

翻译：在真实世界问题的解答情景中,将表格内容和文本内容相结合的混合形式吸引了越来越多的关注,其中数字推理问题是最典型和最具挑战性的问题之一。现有方法通常采用编码器-解码器框架来代表混合内容并生成答案。然而,它无法捕捉数字值、表 schema 和编码器文字信息之间的丰富关系。解码器使用一个简单的预定义操作器分类器,该分类器不够灵活,无法用多种表达式处理数字推理过程。为解决这些问题,本文件建议采用一种新型关系图解方法,用以模拟问题、表格和段落之间的对齐。我们用\ textbf{ table schema,N}umerical 推理模型用\ textb{T}rere decoder (\ textbf{tree decoder) (\ textbf{regHNT/regNT/regNT-rg-ral-ral-ral-ral-ral-leg-leg-leg-leg-ral-ral-ral-leg-rg-ral-rg-ral-le-rg-ral-rg-le-rest-le-rg-lemental-rmal-rg-rg-rg-rg-le-le-rg-le-le-le-lement-le-le-le-le-lemental-leg-rmal-le-le-reg-leg-s-leg-leg-reg-reg-s-le-leg-le-le-le-s-re-leg-s-leg-s-s-s-s-leg-le-le-le-le-le-le-le-le-le-le-le-le-lement-le-le-le-regal-leg-real-re-le-le-le-re-re-re-re-re-re-re-re-re-le-re-re-re-re-re-lement-reg-reg-reg-real-le-le-reg-le-le-le-le-