Graph-to-Tree Neural Networks for Learning Structured Input-Output Translation with Applications to Semantic Parsing and Math Word Problem

Li, Shucheng; Wu, Lingfei; Feng, Shiwei; Xu, Fangli; Xu, Fengyuan

doi:10.48550/arxiv.2004.13781

Cited by 11 publications

(10 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) Constituency Graph: This is one of the more popular graph-construction methods that captures phrase-based syntactic relationships in sentences. It follows the relationships of phrase structure syntax rather than dependencies and dependent syntax [45].…”

Section: ) Evaluation Methods and Metricsmentioning

confidence: 99%

Automatic Requirements Classification Based on Graph Attention Network

Zheng

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Requirements classification is a significant task for requirements engineering, which is timeconsuming and challenging. The traditional requirements classification models usually rely on manual preprocessing and have poor generalization capability. Moreover, these traditional models ignore the sentence structure and syntactic information in requirements. To address these problems, we propose an automatic requirements classification based BERT and graph attention network (GAT), called DBGAT. We construct dependency parse trees and then utilize the GAT for mining the implicit structure feature and syntactic feature of requirements. In addition, we introduce BERT to improve the generalization ability of the model. Experimental results of the PROMISE datasets demonstrate that our proposed DBGAT significantly outperforms existing state-of-the-art methods. Moreover, we investigate the impact of graph construction methods on non-functional requirements classification. DBGAT achieved the best classification results on both seen (F1-scores of up to 91%) and unseen projects (F1-scores of up to 88%), further demonstrating the strong generalization ability.

show abstract

Section: ) Evaluation Methods and Metricsmentioning

confidence: 99%

Automatic Requirements Classification Based on Graph Attention Network

Zheng

et al. 2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…stack) to align the source and target based on Seq2Seq model. [Li et al, 2019] brought in attention mechanism and adopted a two-stage way to generate the sequential expressions.…”

Section: Related Workmentioning

confidence: 99%

“…[Zhang et al, 2020a] utilized knowledge distillation and multiple decoders to generate diverse expressions. [Wu et al, 2020] introduced common sense knowledge into Seq2Tree based generator. And [Wu et al, 2021] explicitly encoded the numeric value of numbers into the Seq2Tree model.…”

Section: Related Workmentioning

confidence: 99%

“…Previous works mainly formalize MWPs as a generation task and adopt Sequence-to-Sequence (Seq2Seq) [Wang et al, 2017;Wang et al, 2018] or Sequence-to-Tree (Seq2Tree) [Xie and Sun, 2019;Zhang et al, 2020b;Wu et al, 2020;Wu et al, 2021] -: r 99 76 Table 1: An example of MWPs. To normalize, we reorder the operands by the appearance order.…”

Section: Introductionmentioning

confidence: 99%

“…However, previous methods generating the expression structure in a top-down manner (Seq2Tree) or generating the expression sequentially (Seq2Seq) with a global representation of the input math problem is hard to notice the subtle local variations of the source problem. Therefore, they are insensitive to subtle local variations [Li et al, 2022] and can easily produce incorrect solutions. Actually, many of the above-mentioned methods have been proven to learn shallow heuristics purely [Patel et al, 2021].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Tackling Math Word Problems with Fine-to-Coarse Abstracting and Reasoning

Li¹,

Jiang²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Math Word Problems (MWP) is an important task that requires the ability of understanding and reasoning over mathematical text. Existing approaches mostly formalize it as a generation task by adopting Seq2Seq or Seq2Tree models to encode an input math problem in natural language as a global representation and generate the output mathematical expression. Such approaches only learn shallow heuristics and fail to capture fine-grained variations in inputs. In this paper, we propose to model a math word problem in a fine-to-coarse manner to capture both the local fine-grained information and the global logical structure of it. Instead of generating a complete equation sequence or expression tree from the global features, we iteratively combine low-level operands to predict a higher-level operator, abstracting the problem and reasoning about the solving operators from bottom to up. Our model is naturally more sensitive to local variations and can better generalize to unseen problem types. Extensive evaluations on Math23k and SVAMP datasets demonstrate the accuracy and robustness of our method. 1

show abstract