Bidirectional Encoder Representations from Transformers (BERT) has achieved state-of-the-art performances on several text classification tasks, such as GLUE and sentiment analysis. Recent work in the legal domain started to use BERT on tasks, such as
legal judgement prediction and violation prediction. A common practise in using BERT is to fine-tune a pre-trained model on a target task and truncate the input texts to the size of the BERT input (e.g. at most 512 tokens). However, due to the unique characteristics of legal documents, it is not clear how to effectively adapt BERT in the legal domain. In this work, we investigate how to deal with long documents, and how is the importance of pre-training on documents from the same domain as the target task. We conduct experiments on the two recent datasets: ECHR Violation Dataset and the Overruling Task Dataset, which are multi-label and binary classification tasks, respectively. Importantly, on average the number of tokens in a document from the ECHR Violation Dataset is more than 1,600. While the documents in the Overruling Task Dataset are shorter (the maximum number of tokens is 204). We thoroughly compare several techniques for adapting BERT on long documents and compare different models pre-trained on the legal and other domains. Our experimental results show that we need to explicitly adapt BERT to handle long documents, as the truncation leads to less effective performance. We also found that pre-training on the documents that are similar to the target task would result in more effective performance on several scenario.
In this work, we propose a novel framework, Gradient Aligned Mutual Learning BERT (GAML-BERT), for improving the early exiting of BERT. GAML-BERT's contributions are two-fold. We conduct a set of pilot experiments, which shows that mutual knowledge d
istillation between a shallow exit and a deep exit leads to better performances for both. From this observation, we use mutual learning to improve BERT's early exiting performances, that is, we ask each exit of a multi-exit BERT to distill knowledge from each other. Second, we propose GA, a novel training method that aligns the gradients from knowledge distillation to cross-entropy losses. Extensive experiments are conducted on the GLUE benchmark, which shows that our GAML-BERT can significantly outperform the state-of-the-art (SOTA) BERT early exiting methods.
Transformer-based models have become the de facto standard in the field of Natural Language Processing (NLP). By leveraging large unlabeled text corpora, they enable efficient transfer learning leading to state-of-the-art results on numerous NLP task
s. Nevertheless, for low resource languages and highly specialized tasks, transformer models tend to lag behind more classical approaches (e.g. SVM, LSTM) due to the lack of aforementioned corpora. In this paper we focus on the legal domain and we introduce a Romanian BERT model pre-trained on a large specialized corpus. Our model outperforms several strong baselines for legal judgement prediction on two different corpora consisting of cases from trials involving banks in Romania.
Passage retrieval and ranking is a key task in open-domain question answering and information retrieval. Current effective approaches mostly rely on pre-trained deep language model-based retrievers and rankers. These methods have been shown to effect
ively model the semantic matching between queries and passages, also in presence of keyword mismatch, i.e. passages that are relevant to a query but do not contain important query keywords. In this paper we consider the Dense Retriever (DR), a passage retrieval method, and the BERT re-ranker, a popular passage re-ranking method. In this context, we formally investigate how these models respond and adapt to a specific type of keyword mismatch -- that caused by keyword typos occurring in queries. Through empirical investigation, we find that typos can lead to a significant drop in retrieval and ranking effectiveness. We then propose a simple typos-aware training framework for DR and BERT re-ranker to address this issue. Our experimental results on the MS MARCO passage ranking dataset show that, with our proposed typos-aware training, DR and BERT re-ranker can become robust to typos in queries, resulting in significantly improved effectiveness compared to models trained without appropriately accounting for typos.
We propose using a multilabel probing task to assess the morphosyntactic representations of multilingual word embeddings. This tweak on canonical probing makes it easy to explore morphosyntactic representations, both holistically and at the level of
individual features (e.g., gender, number, case), and leads more naturally to the study of how language models handle co-occurring features (e.g., agreement phenomena). We demonstrate this task with multilingual BERT (Devlin et al., 2018), training probes for seven typologically diverse languages: Afrikaans, Croatian, Finnish, Hebrew, Korean, Spanish, and Turkish. Through this simple but robust paradigm, we verify that multilingual BERT renders many morphosyntactic features simultaneously extractable. We further evaluate the probes on six held-out languages: Arabic, Chinese, Marathi, Slovenian, Tagalog, and Yoruba. This zero-shot style of probing has the added benefit of revealing which cross-linguistic properties a language model recognizes as being shared by multiple languages.
Understanding robustness and sensitivity of BERT models predicting Alzheimer's disease from text is important for both developing better classification models and for understanding their capabilities and limitations. In this paper, we analyze how a c
ontrolled amount of desired and undesired text alterations impacts performance of BERT. We show that BERT is robust to natural linguistic variations in text. On the other hand, we show that BERT is not sensitive to removing clinically important information from text.
Rumor detection on social media puts pre-trained language models (LMs), such as BERT, and auxiliary features, such as comments, into use. However, on the one hand, rumor detection datasets in Chinese companies with comments are rare; on the other han
d, intensive interaction of attention on Transformer-based models like BERT may hinder performance improvement. To alleviate these problems, we build a new Chinese microblog dataset named Weibo20 by collecting posts and associated comments from Sina Weibo and propose a new ensemble named STANKER (Stacking neTwork bAsed-on atteNtion-masKed BERT). STANKER adopts two level-grained attention-masked BERT (LGAM-BERT) models as base encoders. Unlike the original BERT, our new LGAM-BERT model takes comments as important auxiliary features and masks co-attention between posts and comments on lower-layers. Experiments on Weibo20 and three existing social media datasets showed that STANKER outperformed all compared models, especially beating the old state-of-the-art on Weibo dataset.
This paper describes the HEL-LJU submissions to the MultiLexNorm shared task on multilingual lexical normalization. Our system is based on a BERT token classification preprocessing step, where for each token the type of the necessary transformation i
s predicted (none, uppercase, lowercase, capitalize, modify), and a character-level SMT step where the text is translated from original to normalized given the BERT-predicted transformation constraints. For some languages, depending on the results on development data, the training data was extended by back-translating OpenSubtitles data. In the final ordering of the ten participating teams, the HEL-LJU team has taken the second place, scoring better than the previous state-of-the-art.
This paper presents a pilot study to automatic linguistic preprocessing of Ancient and Byzantine Greek, and morphological analysis more specifically. To this end, a novel subword-based BERT language model was trained on the basis of a varied corpus o
f Modern, Ancient and Post-classical Greek texts. Consequently, the obtained BERT embeddings were incorporated to train a fine-grained Part-of-Speech tagger for Ancient and Byzantine Greek. In addition, a corpus of Greek Epigrams was manually annotated and the resulting gold standard was used to evaluate the performance of the morphological analyser on Byzantine Greek. The experimental results show very good perplexity scores (4.9) for the BERT language model and state-of-the-art performance for the fine-grained Part-of-Speech tagger for in-domain data (treebanks containing a mixture of Classical and Medieval Greek), as well as for the newly created Byzantine Greek gold standard data set. The language models and associated code are made available for use at https://github.com/pranaydeeps/Ancient-Greek-BERT
Current language models are usually trained using a self-supervised scheme, where the main focus is learning representations at the word or sentence level. However, there has been limited progress in generating useful discourse-level representations.
In this work, we propose to use ideas from predictive coding theory to augment BERT-style language models with a mechanism that allows them to learn suitable discourse-level representations. As a result, our proposed approach is able to predict future sentences using explicit top-down connections that operate at the intermediate layers of the network. By experimenting with benchmarks designed to evaluate discourse-related knowledge using pre-trained sentence representations, we demonstrate that our approach improves performance in 6 out of 11 tasks by excelling in discourse relationship detection.
