Neuro-Symbolic Approaches for Text-Based Policy Learning

Text-Based Games (TBGs) have emerged as important testbeds for reinforcement learning (RL) in the natural language domain. Previous methods using LSTM-based action policies are uninterpretable and often overfit the training games showing poor performance to unseen test games. We present SymboLic Action policy for Textual Environments (SLATE), that learns interpretable action policy rules from symbolic abstractions of textual observations for improved generalization. We outline a method for end-to-end differentiable symbolic rule learning and show that such symbolic policies outperform previous state-of-the-art methods in text-based RL for the coin collector environment from 5-10x fewer training games. Additionally, our method provides human-understandable policy rules that can be readily verified for their logical consistency and can be easily debugged.

Different Strokes for Different Folks: Investigating Appropriate Further Pre-training Approaches for Diverse Dialogue Tasks

Loading models pre-trained on the large-scale corpus in the general domain and fine-tuning them on specific downstream tasks is gradually becoming a paradigm in Natural Language Processing. Previous investigations prove that introducing a further pre-training phase between pre-training and fine-tuning phases to adapt the model on the domain-specific unlabeled data can bring positive effects. However, most of these further pre-training works just keep running the conventional pre-training task, e.g., masked language model, which can be regarded as the domain adaptation to bridge the data distribution gap. After observing diverse downstream tasks, we suggest that different tasks may also need a further pre-training phase with appropriate training tasks to bridge the task formulation gap. To investigate this, we carry out a study for improving multiple task-oriented dialogue downstream tasks through designing various tasks at the further pre-training phase. The experiment shows that different downstream tasks prefer different further pre-training tasks, which have intrinsic correlation and most further pre-training tasks significantly improve certain target tasks rather than all. Our investigation indicates that it is of great importance and effectiveness to design appropriate further pre-training tasks modeling specific information that benefit downstream tasks. Besides, we present multiple constructive empirical conclusions for enhancing task-oriented dialogues.

LIORI at SemEval-2021 Task 2: Span Prediction and Binary Classification approaches to Word-in-Context Disambiguation

This paper presents our approaches to SemEval-2021 Task 2: Multilingual and Cross-lingual Word-in-Context Disambiguation task. The first approach attempted to reformulate the task as a question answering problem, while the second one framed it as a binary classification problem. Our best system, which is an ensemble of XLM-R based binary classifiers trained with data augmentation, is among the 3 best-performing systems for Russian, French and Arabic in the multilingual subtask. In the post-evaluation period, we experimented with batch normalization, subword pooling and target word occurrence aggregation methods, resulting in further performance improvements.

A Comparison of Different NMT Approaches to Low-Resource Dutch-Albanian Machine Translation

Low-resource languages can be understood as languages that are more scarce, less studied, less privileged, less commonly taught and for which there are less resources available (Singh, 2008; Cieri et al., 2016; Magueresse et al., 2020). Natural Language Processing (NLP) research and technology mainly focuses on those languages for which there are large data sets available. To illustrate differences in data availability: there are 6 million Wikipedia articles available for English, 2 million for Dutch, and merely 82 thousand for Albanian. The scarce data issue becomes increasingly apparent when large parallel data sets are required for applications such as Neural Machine Translation (NMT). In this work, we investigate to what extent translation between Albanian (SQ) and Dutch (NL) is possible comparing a one-to-one (SQ↔AL) model, a low-resource pivot-based approach (English (EN) as pivot) and a zero-shot translation (ZST) (Johnson et al., 2016; Mattoni et al., 2017) system. From our experiments, it results that the EN-pivot-model outperforms both the direct one-to-one and the ZST model. Since often, small amounts of parallel data are available for low-resource languages or settings, experiments were conducted using small sets of parallel NL↔SQ data. The ZST appeared to be the worst performing models. Even when the available parallel data (NL↔SQ) was added, i.e. in a few-shot setting (FST), it remained the worst performing system according to the automatic (BLEU and TER) and human evaluation.

SAPPHIRE: Approaches for Enhanced Concept-to-Text Generation

We motivate and propose a suite of simple but effective improvements for concept-to-text generation called SAPPHIRE: Set Augmentation and Post-hoc PHrase Infilling and REcombination. We demonstrate their effectiveness on generative commonsense reasoning, a.k.a. the CommonGen task, through experiments using both BART and T5 models. Through extensive automatic and human evaluation, we show that SAPPHIRE noticeably improves model performance. An in-depth qualitative analysis illustrates that SAPPHIRE effectively addresses many issues of the baseline model generations, including lack of commonsense, insufficient specificity, and poor fluency.

Classification of COVID19 tweets using Machine Learning Approaches

The reported work is a description of our participation in the Classification of COVID19 tweets containing symptoms'' shared task, organized by the Social Media Mining for Health Applications (SMM4H)'' workshop. The literature describes two machine learning approaches that were used to build a three class classification system, that categorizes tweets related to COVID19, into three classes, viz., self-reports, non-personal reports, and literature/news mentions. The steps for pre-processing tweets, feature extraction, and the development of the machine learning models, are described extensively in the documentation. Both the developed learning models, when evaluated by the organizers, garnered F1 scores of 0.93 and 0.92 respectively.

Bridging Resolution: Making Sense of the State of the Art

While Yu and Poesio (2020) have recently demonstrated the superiority of their neural multi-task learning (MTL) model to rule-based approaches for bridging anaphora resolution, there is little understanding of (1) how it is better than the rule-based approaches (e.g., are the two approaches making similar or complementary mistakes?) and (2) what should be improved. To shed light on these issues, we (1) propose a hybrid rule-based and MTL approach that would enable a better understanding of their comparative strengths and weaknesses; and (2) perform a manual analysis of the errors made by the MTL model.

IITP-CUNI@3C: Supervised Approaches for Citation Classification (Task A) and Citation Significance Detection (Task B)

Citations are crucial to a scientific discourse. Besides providing additional contexts to research papers, citations act as trackers of the direction of research in a field and as an important measure in understanding the impact of a research publication. With the rapid growth in research publications, automated solutions for identifying the purpose and influence of citations are becoming very important. The 3C Citation Context Classification Task organized as part of the Second Workshop on Scholarly Document Processing @ NAACL 2021 is a shared task to address the aforementioned problems. In this paper, we present our team, IITP-CUNI@3C's submission to the 3C shared tasks. For Task A, citation context purpose classification, we propose a neural multi-task learning framework that harnesses the structural information of the research papers and the relation between the citation context and the cited paper for citation classification. For Task B, citation context influence classification, we use a set of simple features to classify citations based on their perceived significance. We achieve comparable performance with respect to the best performing systems in Task A and superseded the majority baseline in Task B with very simple features.

Nearest neighbour approaches for Emotion Detection in Tweets

Emotion detection is an important task that can be applied to social media data to discover new knowledge. While the use of deep learning methods for this task has been prevalent, they are black-box models, making their decisions hard to interpret for a human operator. Therefore, in this paper, we propose an approach using weighted k Nearest Neighbours (kNN), a simple, easy to implement, and explainable machine learning model. These qualities can help to enhance results' reliability and guide error analysis. In particular, we apply the weighted kNN model to the shared emotion detection task in tweets from SemEval-2018. Tweets are represented using different text embedding methods and emotion lexicon vocabulary scores, and classification is done by an ensemble of weighted kNN models. Our best approaches obtain results competitive with state-of-the-art solutions and open up a promising alternative path to neural network methods.

Comparing Approaches to Dravidian Language Identification

This paper describes the submissions by team HWR to the Dravidian Language Identification (DLI) shared task organized at VarDial 2021 workshop. The DLI training set includes 16,674 YouTube comments written in Roman script containing code-mixed text with English and one of the three South Dravidian languages: Kannada, Malayalam, and Tamil. We submitted results generated using two models, a Naive Bayes classifier with adaptive language models, which has shown to obtain competitive performance in many language and dialect identification tasks, and a transformer-based model which is widely regarded as the state-of-the-art in a number of NLP tasks. Our first submission was sent in the closed submission track using only the training set provided by the shared task organisers, whereas the second submission is considered to be open as it used a pretrained model trained with external data. Our team attained shared second position in the shared task with the submission based on Naive Bayes. Our results reinforce the idea that deep learning methods are not as competitive in language identification related tasks as they are in many other text classification tasks.