Solving algebraic word problems has recently emerged as an important natural language processing task. To solve algebraic word problems, recent studies suggested neural models that generate solution equations by using 'Op (operator/operand)' tokens as a unit of input/output. However, such a neural model suffered two issues: expression fragmentation and operand-context separation. To address each of these two issues, we propose a pure neural model, Expression-Pointer Transformer (EPT), which uses (1) 'Expression' token and (2) operand-context pointers when generating solution equations. The performance of the EPT model is tested on three datasets: ALG514, DRAW-1K, and MAWPS. Compared to the state-of-the-art (SoTA) models, the EPT model achieved a comparable performance accuracy in each of the three datasets; 81.3% on ALG514, 59.5% on DRAW-1K, and 84.5% on MAWPS. The contribution of this paper is two-fold; (1) We propose a pure neural model, EPT, which can address the expression fragmentation and the operandcontext separation. (2) The fully automatic EPT model, which does not use hand-crafted features, yields comparable performance to existing models using hand-crafted features, and achieves better performance than existing pure neural models by at most 40%.

No abstract

No abstract

Math word problem solving is an emerging research topic in Natural Language Processing. Recently, to address the math word problem solving task, researchers have applied the encoderdecoder architecture, which is mainly used in machine translation tasks. The state-of-the-art neural models use hand-crafted features and are based on generation methods. In this paper, we propose the GEO (Generation of Equations by utilizing Operators) model that does not use handcrafted features and addresses two issues that are present in existing neural models: 1. missing domain-specific knowledge features and 2. losing encoder-level knowledge. To address missing domain-specific feature issue, we designed two auxiliary tasks: operation group difference prediction and implicit pair prediction. To address losing encoder-level knowledge issue, we added an Operation Feature Feed Forward (OP3F) layer. Experimental results showed that the GEO model outperformed existing state-of-the-art models on two datasets, 85.1% in MAWPS, and 62.5% in DRAW-1K, and reached comparable performance of 82.1% in ALG514 dataset.

This study proposes a method of using data augmentation to address the problem of data shortages in miscue detection tasks. Three main steps were taken. First, a phoneme classifier was developed to acquire force-aligned data, which would be used for miscue classification and data augmentation. In order to create the phoneme classifier, phonetic features of "Seoul Reading Speech" (SRS) corpus were extracted by using grapheme-to-phoneme (G2P) to train CNN-based models. Second, to obtain miscue labeled corpus, we performed data augmentation using the phoneme classifier output, which is artificially generated miscue corpus of SRS (modified-SRS). This miscue corpus was created by randomly deleting or modifying sound sections according to three miscue categories; extension (EXT), pause (PAU), and pre-correction (PRE). Third, the performance of the miscue classifier was tested after training three types of RNN based models (LSTM, BiLSTM, BiGRU) with the modified-SRS corpus. The results show that the BiGRU model performed best at 0.819 in F1-score on augmented data, while BiLSTM model performed best at 0.512 on real data.