The evolution of Large Language Models (LLMs) has led to significant advancements, with models like Claude and Gemini capable of processing contexts up to 1 million tokens. However, efficiently handling long sequences remains challenging, particularly during the prefilling stage when input lengths exceed GPU memory capacity. Traditional methods often segment sequence into chunks and compress them iteratively with fixed-size memory. However, our empirical analysis shows that the fixed-size memory results in wasted computational and GPU memory resources. Therefore, we introduces Incremental Memory (IM), a method that starts with a small memory size and gradually increases it, optimizing computational efficiency. Additionally, we propose Decremental Chunk based on Incremental Memory (IMDC), which reduces chunk size while increasing memory size, ensuring stable and lower GPU memory usage. Our experiments demonstrate that IMDC is consistently faster (1.45x) and reduces GPU memory consumption by 23.3% compared to fixed-size memory, achieving comparable performance on the LongBench Benchmark.

Incorporating lexical knowledge into deep learning models has been proved to be very effective for sequence labeling tasks. However, previous works commonly have difficulty dealing with large-scale dynamic lexicons which often cause excessive matching noise and problems of frequent updates. In this paper, we propose DyLex, a plug-in lexicon incorporation approach for BERT based sequence labeling tasks. Instead of leveraging embeddings of words in the lexicon as in conventional methods, we adopt word-agnostic tag embeddings to avoid re-training the representation while updating the lexicon. Moreover, we employ an effective supervised lexical knowledge denoising method to smooth out matching noise. Finally, we introduce a col-wise attention based knowledge fusion mechanism to guarantee the pluggability of the proposed framework. Experiments on ten datasets of three tasks show that the proposed framework achieves new SOTA, even with very large scale lexicons.

Language model pre-training, such as BERT, has significantly improved the performances of many natural language processing tasks. However, pre-trained language models are usually computationally expensive, so it is difficult to efficiently execute them on resource-restricted devices. To accelerate inference and reduce model size while maintaining accuracy, we first propose a novel Transformer distillation method that is specially designed for knowledge distillation (KD) of the Transformer-based models. By leveraging this new KD method, the plenty of knowledge encoded in a large “teacher” BERT can be effectively transferred to a small “student” TinyBERT. Then, we introduce a new two-stage learning framework for TinyBERT, which performs Transformer distillation at both the pre-training and task-specific learning stages. This framework ensures that TinyBERT can capture the general-domain as well as the task-specific knowledge in BERT. TinyBERT4 with 4 layers is empirically effective and achieves more than 96.8% the performance of its teacher BERT-Base on GLUE benchmark, while being 7.5x smaller and 9.4x faster on inference. TinyBERT4 is also significantly better than 4-layer state-of-the-art baselines on BERT distillation, with only ~28% parameters and ~31% inference time of them. Moreover, TinyBERT6 with 6 layers performs on-par with its teacher BERT-Base.

This paper describes DM-NLP’s system for toponym resolution task at Semeval 2019. Our system was developed for toponym detection, disambiguation and end-to-end resolution which is a pipeline of the former two. For toponym detection, we utilized the state-of-the-art sequence labeling model, namely, BiLSTM-CRF model as backbone. A lot of strategies are adopted for further improvement, such as pre-training, model ensemble, model averaging and data augment. For toponym disambiguation, we adopted the widely used searching and ranking framework. For ranking, we proposed several effective features for measuring the consistency between the detected toponym and toponyms in GeoNames. Eventually, our system achieved the best performance among all the submitted results in each sub task.

Gazetteers were shown to be useful resources for named entity recognition (NER). Many existing approaches to incorporating gazetteers into machine learning based NER systems rely on manually defined selection strategies or handcrafted templates, which may not always lead to optimal effectiveness, especially when multiple gazetteers are involved. This is especially the case for the task of Chinese NER, where the words are not naturally tokenized, leading to additional ambiguities. To automatically learn how to incorporate multiple gazetteers into an NER system, we propose a novel approach based on graph neural networks with a multi-digraph structure that captures the information that the gazetteers offer. Experiments on various datasets show that our model is effective in incorporating rich gazetteer information while resolving ambiguities, outperforming previous approaches.

Supervised approaches to named entity recognition (NER) are largely developed based on the assumption that the training data is fully annotated with named entity information. However, in practice, annotated data can often be imperfect with one typical issue being the training data may contain incomplete annotations. We highlight several pitfalls associated with learning under such a setup in the context of NER and identify limitations associated with existing approaches, proposing a novel yet easy-to-implement approach for recognizing named entities with incomplete data annotations. We demonstrate the effectiveness of our approach through extensive experiments.

This paper introduces a new task – Chinese address parsing – the task of mapping Chinese addresses into semantically meaningful chunks. While it is possible to model this problem using a conventional sequence labelling approach, our observation is that there exist complex dependencies between labels that cannot be readily captured by a simple linear-chain structure. We investigate neural structured prediction models with latent variables to capture such rich structural information within Chinese addresses. We create and publicly release a new dataset consisting of 15K Chinese addresses, and conduct extensive experiments on the dataset to investigate the model effectiveness and robustness. We release our code and data at http://statnlp.org/research/sp.

This paper introduces the DM_NLP team’s system for NLPTEA 2018 shared task of Chinese Grammatical Error Diagnosis (CGED), which can be used to detect and correct grammatical errors in texts written by Chinese as a Foreign Language (CFL) learners. This task aims at not only detecting four types of grammatical errors including redundant words (R), missing words (M), bad word selection (S) and disordered words (W), but also recommending corrections for errors of M and S types. We proposed a hybrid system including four models for this task with two stages: the detection stage and the correction stage. In the detection stage, we first used a BiLSTM-CRF model to tag potential errors by sequence labeling, along with some handcraft features. Then we designed three Grammatical Error Correction (GEC) models to generate corrections, which could help to tune the detection result. In the correction stage, candidates were generated by the three GEC models and then merged to output the final corrections for M and S types. Our system reached the highest precision in the correction subtask, which was the most challenging part of this shared task, and got top 3 on F1 scores for position detection of errors.

Semantic role labeling (SRL) aims to recognize the predicate-argument structure of a sentence. Syntactic information has been paid a great attention over the role of enhancing SRL. However, the latest advance shows that syntax would not be so important for SRL with the emerging much smaller gap between syntax-aware and syntax-agnostic SRL. To comprehensively explore the role of syntax for SRL task, we extend existing models and propose a unified framework to investigate more effective and more diverse ways of incorporating syntax into sequential neural networks. Exploring the effect of syntactic input quality on SRL performance, we confirm that high-quality syntactic parse could still effectively enhance syntactically-driven SRL. Using empirically optimized integration strategy, we even enlarge the gap between syntax-aware and syntax-agnostic SRL. Our framework achieves state-of-the-art results on CoNLL-2009 benchmarks both for English and Chinese, substantially outperforming all previous models.

This paper describes our submissions for SemEval-2018 Task 8: Semantic Extraction from CybersecUrity REports using NLP. The DM_NLP participated in two subtasks: SubTask 1 classifies if a sentence is useful for inferring malware actions and capabilities, and SubTask 2 predicts token labels (“Action”, “Entity”, “Modifier” and “Others”) for a given malware-related sentence. Since we leverage results of Subtask 2 directly to infer the result of Subtask 1, the paper focus on the system solving Subtask 2. By taking Subtask 2 as a sequence labeling task, our system relies on a recurrent neural network named BiLSTM-CNN-CRF with rich linguistic features, such as POS tags, dependency parsing labels, chunking labels, NER labels, Brown clustering. Our system achieved the highest F1 score in both token level and phrase level.

Hypernym discovery aims to discover the hypernym word sets given a hyponym word and proper corpus. This paper proposes a simple but effective method for the discovery of hypernym sets based on word embedding, which can be used to measure the contextual similarities between words. Given a test hyponym word, we get its hypernym lists by computing the similarities between the hyponym word and words in the training data, and fill the test word’s hypernym lists with the hypernym list in the training set of the nearest similarity distance to the test word. In SemEval 2018 task9, our results, achieve 1st on Spanish, 2nd on Italian, 6th on English in the metric of MAP.

This paper introduces Alibaba NLP team system on IJCNLP 2017 shared task No. 1 Chinese Grammatical Error Diagnosis (CGED). The task is to diagnose four types of grammatical errors which are redundant words (R), missing words (M), bad word selection (S) and disordered words (W). We treat the task as a sequence tagging problem and design some handcraft features to solve it. Our system is mainly based on the LSTM-CRF model and 3 ensemble strategies are applied to improve the performance. At the identification level and the position level our system gets the highest F1 scores. At the position level, which is the most difficult level, we perform best on all metrics.

Idioms and other figuratively used expressions pose considerable problems to natural language processing applications because they are very frequent and often behave idiosyncratically. Consequently, there has been much research on the automatic detection and extraction of idiomatic expressions. Most studies focus on type-based idiom detection, i.e., distinguishing whether a given expression can (potentially) be used idiomatically. However, many expressions such as ""break the ice"" can have both literal and non-literal readings and need to be disambiguated in a given context (token-based detection). So far relatively few approaches have attempted context-based idiom detection. One reason for this may be that few annotated resources are available that disambiguate expressions in context. With the IDIX corpus, we aim to address this. IDIX is available as an add-on to the BNC and disambiguates different usages of a subset of idioms. We believe that this resource will be useful both for linguistic and computational linguistic studies.