Shengyi Jiang

Xu-Hui Liu, Feng Xu, Xinyu Zhang et al.

Imitation learning aims to mimic the behavior of experts without explicit reward signals. Passive imitation learning methods which use static expert datasets typically suffer from compounding error, low sample efficiency, and high hyper-parameter sensitivity. In contrast, active imitation learning methods solicit expert interventions to address the limitations. However, recent active imitation learning methods are designed based on human intuitions or empirical experience without theoretical guarantee. In this paper, we propose a novel active imitation learning framework based on a teacher-student interaction model, in which the teacher's goal is to identify the best teaching behavior and actively affect the student's learning process. By solving the optimization objective of this framework, we propose a practical implementation, naming it AdapMen. Theoretical analysis shows that AdapMen can improve the error bound and avoid compounding error under mild conditions. Experiments on the MetaDrive benchmark and Atari 2600 games validate our theoretical analysis and show that our method achieves near-expert performance with much less expert involvement and total sampling steps than previous methods. The code is available at https://github.com/liuxhym/AdapMen.

Nankai Lin, Yingwen Fu, Xiaotian Lin et al.

As an extensive research in the field of natural language processing (NLP), aspect-based sentiment analysis (ABSA) is the task of predicting the sentiment expressed in a text relative to the corresponding aspect. Unfortunately, most languages lack sufficient annotation resources, thus more and more recent researchers focus on cross-lingual aspect-based sentiment analysis (XABSA). However, most recent researches only concentrate on cross-lingual data alignment instead of model alignment. To this end, we propose a novel framework, CL-XABSA: Contrastive Learning for Cross-lingual Aspect-Based Sentiment Analysis. Based on contrastive learning, we close the distance between samples with the same label in different semantic spaces, thus achieving a convergence of semantic spaces of different languages. Specifically, we design two contrastive strategies, token level contrastive learning of token embeddings (TL-CTE) and sentiment level contrastive learning of token embeddings (SL-CTE), to regularize the semantic space of source and target language to be more uniform. Since our framework can receive datasets in multiple languages during training, our framework can be adapted not only for XABSA task but also for multilingual aspect-based sentiment analysis (MABSA). To further improve the performance of our model, we perform knowledge distillation technology leveraging data from unlabeled target language. In the distillation XABSA task, we further explore the comparative effectiveness of different data (source dataset, translated dataset, and code-switched dataset). The results demonstrate that the proposed method has a certain improvement in the three tasks of XABSA, distillation XABSA and MABSA. For reproducibility, our code for this paper is available at https://github.com/GKLMIP/CL-XABSA.

Yingwen Fu, Nankai Lin, Ziyu Yang et al.

Cross-lingual Named Entity Recognition (NER) has recently become a research hotspot because it can alleviate the data-hungry problem for low-resource languages. However, few researches have focused on the scenario where the source-language labeled data is also limited in some specific domains. A common approach for this scenario is to generate more training data through translation or generation-based data augmentation method. Unfortunately, we find that simply combining source-language data and the corresponding translation cannot fully exploit the translated data and the improvements obtained are somewhat limited. In this paper, we describe our novel dual-contrastive framework ConCNER for cross-lingual NER under the scenario of limited source-language labeled data. Specifically, based on the source-language samples and their translations, we design two contrastive objectives for cross-language NER at different grammatical levels, namely Translation Contrastive Learning (TCL) to close sentence representations between translated sentence pairs and Label Contrastive Learning (LCL) to close token representations within the same labels. Furthermore, we utilize knowledge distillation method where the NER model trained above is used as the teacher to train a student model on unlabeled target-language data to better fit the target language. We conduct extensive experiments on a wide variety of target languages, and the results demonstrate that ConCNER tends to outperform multiple baseline methods. For reproducibility, our code for this paper is available at https://github.com/GKLMIP/ConCNER.

Nankai Lin, Hongbin Zhang, Menglan Shen et al.

Grammatical error correction (GEC) is a challenging task of natural language processing techniques. While more attempts are being made in this approach for universal languages like English or Chinese, relatively little work has been done for low-resource languages for the lack of large annotated corpora. In low-resource languages, the current unsupervised GEC based on language model scoring performs well. However, the pre-trained language model is still to be explored in this context. This study proposes a BERT-based unsupervised GEC framework, where GEC is viewed as multi-class classification task. The framework contains three modules: data flow construction module, sentence perplexity scoring module, and error detecting and correcting module. We propose a novel scoring method for pseudo-perplexity to evaluate a sentence's probable correctness and construct a Tagalog corpus for Tagalog GEC research. It obtains competitive performance on the Tagalog corpus we construct and open-source Indonesian corpus and it demonstrates that our framework is complementary to baseline method for low-resource GEC task.

Xu-Hui Liu, Tian-Shuo Liu, Shengyi Jiang et al.

Combining offline and online reinforcement learning (RL) techniques is indeed crucial for achieving efficient and safe learning where data acquisition is expensive. Existing methods replay offline data directly in the online phase, resulting in a significant challenge of data distribution shift and subsequently causing inefficiency in online fine-tuning. To address this issue, we introduce an innovative approach, \textbf{E}nergy-guided \textbf{DI}ffusion \textbf{S}ampling (EDIS), which utilizes a diffusion model to extract prior knowledge from the offline dataset and employs energy functions to distill this knowledge for enhanced data generation in the online phase. The theoretical analysis demonstrates that EDIS exhibits reduced suboptimality compared to solely utilizing online data or directly reusing offline data. EDIS is a plug-in approach and can be combined with existing methods in offline-to-online RL setting. By implementing EDIS to off-the-shelf methods Cal-QL and IQL, we observe a notable 20% average improvement in empirical performance on MuJoCo, AntMaze, and Adroit environments. Code is available at \url{https://github.com/liuxhym/EDIS}.

Xue-Kun Jin, Xu-Hui Liu, Shengyi Jiang et al.

Value function estimation is an indispensable subroutine in reinforcement learning, which becomes more challenging in the offline setting. In this paper, we propose Hybrid Value Estimation (HVE) to reduce value estimation error, which trades off bias and variance by balancing between the value estimation from offline data and the learned model. Theoretical analysis discloses that HVE enjoys a better error bound than the direct methods. HVE can be leveraged in both off-policy evaluation and offline reinforcement learning settings. We, therefore, provide two concrete algorithms Off-policy HVE (OPHVE) and Model-based Offline HVE (MOHVE), respectively. Empirical evaluations on MuJoCo tasks corroborate the theoretical claim. OPHVE outperforms other off-policy evaluation methods in all three metrics measuring the estimation effectiveness, while MOHVE achieves better or comparable performance with state-of-the-art offline reinforcement learning algorithms. We hope that HVE could shed some light on further research on reinforcement learning from fixed data.

Xiaotian Lin, Nankai Lin, Yingwen Fu et al.

Deep learning-based text classification models need abundant labeled data to obtain competitive performance. Unfortunately, annotating large-size corpus is time-consuming and laborious. To tackle this, multiple researches try to use data augmentation to expand the corpus size. However, data augmentation may potentially produce some noisy augmented samples. There are currently no works exploring sample selection for augmented samples in nature language processing field. In this paper, we propose a novel self-training selection framework with two selectors to select the high-quality samples from data augmentation. Specifically, we firstly use an entropy-based strategy and the model prediction to select augmented samples. Considering some samples with high quality at the above step may be wrongly filtered, we propose to recall them from two perspectives of word overlap and semantic similarity. Experimental results show the effectiveness and simplicity of our framework.

Nankai Lin, Hongyan Wu, Sihui Fu et al.

Chinese spelling check is a task to detect and correct spelling mistakes in Chinese text. Existing research aims to enhance the text representation and use multi-source information to improve the detection and correction capabilities of models, but does not pay too much attention to improving their ability to distinguish between confusable words. Contrastive learning, whose aim is to minimize the distance in representation space between similar sample pairs, has recently become a dominant technique in natural language processing. Inspired by contrastive learning, we present a novel framework for Chinese spelling checking, which consists of three modules: language representation, spelling check and reverse contrastive learning. Specifically, we propose a reverse contrastive learning strategy, which explicitly forces the model to minimize the agreement between the similar examples, namely, the phonetically and visually confusable characters. Experimental results show that our framework is model-agnostic and could be combined with existing Chinese spelling check models to yield state-of-the-art performance.

Yingwen Fu, Jinyi Chen, Nankai Lin et al.

The Yunshan Cup 2020 track focused on creating a framework for evaluating different methods of part-of-speech (POS). There were two tasks for this track: (1) POS tagging for the Indonesian language, and (2) POS tagging for the Lao tagging. The Indonesian dataset is comprised of 10000 sentences from Indonesian news within 29 tags. And the Lao dataset consists of 8000 sentences within 27 tags. 25 teams registered for the task. The methods of participants ranged from feature-based to neural networks using either classical machine learning techniques or ensemble methods. The best performing results achieve an accuracy of 95.82% for Indonesian and 93.03%, showing that neural sequence labeling models significantly outperform classic feature-based methods and rule-based methods.

Nankai Lin, Nankai Lin, Xiaotian Lin et al.

As a fundamental task in natural language processing, Chinese Grammatical Error Correction (CGEC) has gradually received widespread attention and become a research hotspot. However, one obvious deficiency for the existing CGEC evaluation system is that the evaluation values are significantly influenced by the Chinese word segmentation results or different language models. The evaluation values of the same error correction model can vary considerably under different word segmentation systems or different language models. However, it is expected that these metrics should be independent of the word segmentation results and language models, as they may lead to a lack of uniqueness and comparability in the evaluation of different methods. To this end, we propose three novel evaluation metrics for CGEC in two dimensions: reference-based and reference-less. In terms of the reference-based metric, we introduce sentence-level accuracy and char-level BLEU to evaluate the corrected sentences. Besides, in terms of the reference-less metric, we adopt char-level meaning preservation to measure the semantic preservation degree of the corrected sentences. We deeply evaluate and analyze the reasonableness and validity of the three proposed metrics, and we expect them to become a new standard for CGEC.

Nankai Lin, Meiyu Zeng, Wentao Huang et al.

Currently, the majority of research in grammatical error correction (GEC) is concentrated on universal languages, such as English and Chinese. Many low-resource languages lack accessible evaluation corpora. How to efficiently construct high-quality evaluation corpora for GEC in low-resource languages has become a significant challenge. To fill these gaps, in this paper, we present a framework for constructing GEC corpora. Specifically, we focus on Indonesian as our research language and construct an evaluation corpus for Indonesian GEC using the proposed framework, addressing the limitations of existing evaluation corpora in Indonesian. Furthermore, we investigate the feasibility of utilizing existing large language models (LLMs), such as GPT-3.5-Turbo and GPT-4, to streamline corpus annotation efforts in GEC tasks. The results demonstrate significant potential for enhancing the performance of LLMs in low-resource language settings. Our code and corpus can be obtained from https://github.com/GKLMIP/GEC-Construction-Framework.

Nankai Lin, Sihui Fu, Hongyan Wu et al.

Chinese features prominently in the Chinese communities located in the nations of Malay Archipelago. In these countries, Chinese has undergone the process of adjustment to the local languages and cultures, which leads to the occurrence of a Chinese variant in each country. In this paper, we conducted a quantitative analysis on Chinese news texts collected from five Malay Archipelago nations, namely Indonesia, Malaysia, Singapore, Philippines and Brunei, trying to figure out their differences with the texts written in modern standard Chinese from a lexical and syntactic perspective. The statistical results show that the Chinese variants used in these five nations are quite different, diverging from their modern Chinese mainland counterpart. Meanwhile, we managed to extract and classify several featured Chinese words used in each nation. All these discrepancies reflect how Chinese evolves overseas, and demonstrate the profound impact rom local societies and cultures on the development of Chinese.

Zhongren Chen, Xinyue Xu, Shengyi Jiang et al.

Small subgraphs (graphlets) are important features to describe fundamental units of a large network. The calculation of the subgraph frequency distributions has a wide application in multiple domains including biology and engineering. Unfortunately due to the inherent complexity of this task, most of the existing methods are computationally intensive and inefficient. In this work, we propose GNNS, a novel representational learning framework that utilizes graph neural networks to sample subgraphs efficiently for estimating their frequency distribution. Our framework includes an inference model and a generative model that learns hierarchical embeddings of nodes, subgraphs, and graph types. With the learned model and embeddings, subgraphs are sampled in a highly scalable and parallel way and the frequency distribution estimation is then performed based on these sampled subgraphs. Eventually, our methods achieve comparable accuracy and a significant speedup by three orders of magnitude compared to existing methods.

Jing-Cheng Pang, Kaiyuan Li, Yidi Wang et al.

A central challenge in reinforcement learning (RL) is its dependence on extensive real-world interaction data to learn task-specific policies. While recent work demonstrates that large language models (LLMs) can mitigate this limitation by generating synthetic experience (noted as imaginary rollouts) for mastering novel tasks, progress in this emerging field is hindered due to the lack of a standard benchmark. To bridge this gap, we introduce ImagineBench, the first comprehensive benchmark for evaluating offline RL algorithms that leverage both real rollouts and LLM-imaginary rollouts. The key features of ImagineBench include: (1) datasets comprising environment-collected and LLM-imaginary rollouts; (2) diverse domains of environments covering locomotion, robotic manipulation, and navigation tasks; and (3) natural language task instructions with varying complexity levels to facilitate language-conditioned policy learning. Through systematic evaluation of state-of-the-art offline RL algorithms, we observe that simply applying existing offline RL algorithms leads to suboptimal performance on unseen tasks, achieving 35.44% success rate in hard tasks in contrast to 64.37% of method training on real rollouts for hard tasks. This result highlights the need for algorithm advancements to better leverage LLM-imaginary rollouts. Additionally, we identify key opportunities for future research: including better utilization of imaginary rollouts, fast online adaptation and continual learning, and extension to multi-modal tasks. Our code is publicly available at https://github.com/LAMDA-RL/ImagineBench.

Feng Xu, Yan Yin, Xinyu Zhang et al.

Alphas are pivotal in providing signals for quantitative trading. The industry highly values the discovery of formulaic alphas for their interpretability and ease of analysis, compared with the expressive yet overfitting-prone black-box alphas. In this work, we focus on discovering formulaic alphas. Prior studies on automatically generating a collection of formulaic alphas were mostly based on genetic programming (GP), which is known to suffer from the problems of being sensitive to the initial population, converting to local optima, and slow computation speed. Recent efforts employing deep reinforcement learning (DRL) for alpha discovery have not fully addressed key practical considerations such as alpha correlations and validity, which are crucial for their effectiveness. In this work, we propose a novel framework for alpha discovery using DRL by formulating the alpha discovery process as program construction. Our agent, $\text{Alpha}^2$, assembles an alpha program optimized for an evaluation metric. A search algorithm guided by DRL navigates through the search space based on value estimates for potential alpha outcomes. The evaluation metric encourages both the performance and the diversity of alphas for a better final trading strategy. Our formulation of searching alphas also brings the advantage of pre-calculation dimensional analysis, ensuring the logical soundness of alphas, and pruning the vast search space to a large extent. Empirical experiments on real-world stock markets demonstrates $\text{Alpha}^2$'s capability to identify a diverse set of logical and effective alphas, which significantly improves the performance of the final trading strategy. The code of our method is available at https://github.com/x35f/alpha2.

Hongyan Wu, Peijian Zeng, Weixiong Zheng et al.

Cross-modal text-molecule retrieval task bridges molecule structures and natural language descriptions. Existing methods predominantly focus on aligning text modality and molecule modality, yet they overlook adaptively adjusting the learning states at different training stages and enhancing training efficiency. To tackle these challenges, this paper proposes a Curriculum Learning-bAsed croSS-modal text-molecule training framework (CLASS), which can be integrated with any backbone to yield promising performance improvement. Specifically, we quantify the sample difficulty considering both text modality and molecule modality, and design a sample scheduler to introduce training samples via an easy-to-difficult paradigm as the training advances, remarkably reducing the scale of training samples at the early stage of training and improving training efficiency. Moreover, we introduce adaptive intensity learning to increase the training intensity as the training progresses, which adaptively controls the learning intensity across all curriculum stages. Experimental results on the ChEBI-20 dataset demonstrate that our proposed method gains superior performance, simultaneously achieving prominent time savings.

Hongyan Wu, Zhengming Chen, Zijian Li et al.

Hate speech frequently appears on social media platforms and urgently needs to be effectively controlled. Alleviating the bias caused by hate speech can help resolve various ethical issues. Although existing research has constructed several datasets for hate speech detection, these datasets seldom consider the diversity and variability of bias, making them far from real-world scenarios. To fill this gap, we propose a benchmark HateDebias to analyze the fairness of models under dynamically evolving environments. Specifically, to meet the diversity of biases, we collect hate speech data with different types of biases from real-world scenarios. To further simulate the variability in the real-world scenarios(i.e., the changing of bias attributes in datasets), we construct a dataset to follow the continuous learning setting and evaluate the detection accuracy of models on the HateDebias, where performance degradation indicates a significant bias toward a specific attribute. To provide a potential direction, we further propose a continual debiasing framework tailored to dynamic bias in real-world scenarios, integrating memory replay and bias information regularization to ensure the fairness of the model. Experiment results on the HateDebias benchmark reveal that our methods achieve improved performance in mitigating dynamic biases in real-world scenarios, highlighting the practicality in real-world applications.

Xiaotian Lin, Nankai Lin, Kanoksak Wattanachote et al.

As the fourth largest language family in the world, the Dravidian languages have become a research hotspot in natural language processing (NLP). Although the Dravidian languages contain a large number of languages, there are relatively few public available resources. Besides, text classification task, as a basic task of natural language processing, how to combine it to multiple languages in the Dravidian languages, is still a major difficulty in Dravidian Natural Language Processing. Hence, to address these problems, we proposed a multilingual text classification framework for the Dravidian languages. On the one hand, the framework used the LaBSE pre-trained model as the base model. Aiming at the problem of text information bias in multi-task learning, we propose to use the MLM strategy to select language-specific words, and used adversarial training to perturb them. On the other hand, in view of the problem that the model cannot well recognize and utilize the correlation among languages, we further proposed a language-specific representation module to enrich semantic information for the model. The experimental results demonstrated that the framework we proposed has a significant performance in multilingual text classification tasks with each strategy achieving certain improvements.

Ru Peng, Nankai Lin, Yi Fang et al.

Syntax knowledge contributes its powerful strength in Neural machine translation (NMT) tasks. Early NMT works supposed that syntax details can be automatically learned from numerous texts via attention networks. However, succeeding researches pointed out that limited by the uncontrolled nature of attention computation, the NMT model requires an external syntax to capture the deep syntactic awareness. Although existing syntax-aware NMT methods have born great fruits in combining syntax, the additional workloads they introduced render the model heavy and slow. Particularly, these efforts scarcely involve the Transformer-based NMT and modify its core self-attention network (SAN). To this end, we propose a parameter-free, Dependency-scaled Self-Attention Network (Deps-SAN) for syntax-aware Transformer-based NMT. A quantified matrix of dependency closeness between tokens is constructed to impose explicit syntactic constraints into the SAN for learning syntactic details and dispelling the dispersion of attention distributions. Two knowledge sparsing techniques are further integrated to avoid the model overfitting the dependency noises introduced by the external parser. Experiments and analyses on IWSLT14 German-to-English and WMT16 German-to-English benchmark NMT tasks verify the effectiveness of our approach.

Nankai Lin, Yingwen Fu, Chuwei Chen et al.

Trained on the large corpus, pre-trained language models (PLMs) can capture different levels of concepts in context and hence generate universal language representations. They can benefit multiple downstream natural language processing (NLP) tasks. Although PTMs have been widely used in most NLP applications, especially for high-resource languages such as English, it is under-represented in Lao NLP research. Previous work on Lao has been hampered by the lack of annotated datasets and the sparsity of language resources. In this work, we construct a text classification dataset to alleviate the resource-scare situation of the Lao language. We additionally present the first transformer-based PTMs for Lao with four versions: BERT-small, BERT-base, ELECTRA-small and ELECTRA-base, and evaluate it over two downstream tasks: part-of-speech tagging and text classification. Experiments demonstrate the effectiveness of our Lao models. We will release our models and datasets to the community, hoping to facilitate the future development of Lao NLP applications.

Yingwen Fu, Nankai Lin, Zhihe Yang et al.

Named entity recognition (NER) is a fundamental task of natural language processing (NLP). However, most state-of-the-art research is mainly oriented to high-resource languages such as English and has not been widely applied to low-resource languages. In Malay language, relevant NER resources are limited. In this work, we propose a dataset construction framework, which is based on labeled datasets of homologous languages and iterative optimization, to build a Malay NER dataset (MYNER) comprising 28,991 sentences (over 384 thousand tokens). Additionally, to better integrate boundary information for NER, we propose a multi-task (MT) model with a bidirectional revision (Bi-revision) mechanism for Malay NER task. Specifically, an auxiliary task, boundary detection, is introduced to improve NER training in both explicit and implicit ways. Furthermore, a gated ignoring mechanism is proposed to conduct conditional label transfer and alleviate error propagation by the auxiliary task. Experimental results demonstrate that our model achieves comparable results over baselines on MYNER. The dataset and the model in this paper would be publicly released as a benchmark dataset.

Jing-Cheng Pang, Tian Xu, Shengyi Jiang et al.

Reinforcement learning (RL) has demonstrated impressive performance in decision-making tasks like embodied control, autonomous driving and financial trading. In many decision-making tasks, the agents often encounter the problem of executing actions under limited budgets. However, classic RL methods typically overlook the challenges posed by such sparse-executing actions. They operate under the assumption that all actions can be taken for a unlimited number of times, both in the formulation of the problem and in the development of effective algorithms. To tackle the issue of limited action execution in RL, this paper first formalizes the problem as a Sparse Action Markov Decision Process (SA-MDP), in which specific actions in the action space can only be executed for a limited time. Then, we propose a policy optimization algorithm, Action Sparsity REgularization (ASRE), which adaptively handles each action with a distinct preference. ASRE operates through two steps: First, ASRE evaluates action sparsity by constrained action sampling. Following this, ASRE incorporates the sparsity evaluation into policy learning by way of an action distribution regularization. We provide theoretical identification that validates the convergence of ASRE to a regularized optimal value function. Experiments on tasks with known sparse-executing actions, where classical RL algorithms struggle to train policy efficiently, ASRE effectively constrains the action sampling and outperforms baselines. Moreover, we present that ASRE can generally improve the performance in Atari games, demonstrating its broad applicability.

Xu-Hui Liu, Zhenghai Xue, Jing-Cheng Pang et al.

In reinforcement learning, experience replay stores past samples for further reuse. Prioritized sampling is a promising technique to better utilize these samples. Previous criteria of prioritization include TD error, recentness and corrective feedback, which are mostly heuristically designed. In this work, we start from the regret minimization objective, and obtain an optimal prioritization strategy for Bellman update that can directly maximize the return of the policy. The theory suggests that data with higher hindsight TD error, better on-policiness and more accurate Q value should be assigned with higher weights during sampling. Thus most previous criteria only consider this strategy partially. We not only provide theoretical justifications for previous criteria, but also propose two new methods to compute the prioritization weight, namely ReMERN and ReMERT. ReMERN learns an error network, while ReMERT exploits the temporal ordering of states. Both methods outperform previous prioritized sampling algorithms in challenging RL benchmarks, including MuJoCo, Atari and Meta-World.

Zhenyu Xuan, Rui Bao, Shengyi Jiang

Chinese NER is a challenging task. As pictographs, Chinese characters contain latent glyph information, which is often overlooked. In this paper, we propose the FGN, Fusion Glyph Network for Chinese NER. Except for adding glyph information, this method may also add extra interactive information with the fusion mechanism. The major innovations of FGN include: (1) a novel CNN structure called CGS-CNN is proposed to capture both glyph information and interactive information between glyphs from neighboring characters. (2) we provide a method with sliding window and Slice-Attention to fuse the BERT representation and glyph representation for a character, which may capture potential interactive knowledge between context and glyph. Experiments are conducted on four NER datasets, showing that FGN with LSTM-CRF as tagger achieves new state-of-the-arts performance for Chinese NER. Further, more experiments are conducted to investigate the influences of various components and settings in FGN.