Thanks to the strong representation learning capability of deep learning, especially pre-training techniques with language model loss, dependency parsing has achieved great performance boost in the in-domain scenario with abundant labeled training da
ta for target domains. However, the parsing community has to face the more realistic setting where the parsing performance drops drastically when labeled data only exists for several fixed out-domains. In this work, we propose a novel model for multi-source cross-domain dependency parsing. The model consists of two components, i.e., a parameter generation network for distinguishing domain-specific features, and an adversarial network for learning domain-invariant representations. Experiments on a recently released NLPCC-2019 dataset for multi-domain dependency parsing show that our model can consistently improve cross-domain parsing performance by about 2 points in averaged labeled attachment accuracy (LAS) over strong BERT-enhanced baselines. Detailed analysis is conducted to gain more insights on contributions of the two components.
Most recent studies for relation extraction (RE) leverage the dependency tree of the input sentence to incorporate syntax-driven contextual information to improve model performance, with little attention paid to the limitation where high-quality depe
ndency parsers in most cases unavailable, especially for in-domain scenarios. To address this limitation, in this paper, we propose attentive graph convolutional networks (A-GCN) to improve neural RE methods with an unsupervised manner to build the context graph, without relying on the existence of a dependency parser. Specifically, we construct the graph from n-grams extracted from a lexicon built from pointwise mutual information (PMI) and apply attention over the graph. Therefore, different word pairs from the contexts within and across n-grams are weighted in the model and facilitate RE accordingly. Experimental results with further analyses on two English benchmark datasets for RE demonstrate the effectiveness of our approach, where state-of-the-art performance is observed on both datasets.
The current natural language processing is strongly focused on raising accuracy. The progress comes at a cost of super-heavy models with hundreds of millions or even billions of parameters. However, simple syntactic tasks such as part-of-speech (POS)
tagging, dependency parsing or named entity recognition (NER) do not require the largest models to achieve acceptable results. In line with this assumption we try to minimize the size of the model that jointly performs all three tasks. We introduce ComboNER: a lightweight tool, orders of magnitude smaller than state-of-the-art transformers. It is based on pre-trained subword embeddings and recurrent neural network architecture. ComboNER operates on Polish language data. The model has outputs for POS tagging, dependency parsing and NER. Our paper contains some insights from fine-tuning of the model and reports its overall results.
This paper deliberates on the process of building the first constituency-to-dependency conversion tool of Turkish. The starting point of this work is a previous study in which 10,000 phrase structure trees were manually transformed into Turkish from
the original PennTreebank corpus. Within the scope of this project, these Turkish phrase structure trees were automatically converted into UD-style dependency structures, using both a rule-based algorithm and a machine learning algorithm specific to the requirements of the Turkish language. The results of both algorithms were compared and the machine learning approach proved to be more accurate than the rule-based algorithm. The output was revised by a team of linguists. The refined versions were taken as gold standard annotations for the evaluation of the algorithms. In addition to its contribution to the UD Project with a large dataset of 10,000 Turkish dependency trees, this project also fulfills the important gap of a Turkish conversion tool, enabling the quick compilation of dependency corpora which can be used for the training of better dependency parsers.
Manually annotating a treebank is time-consuming and labor-intensive. We conduct delexicalized cross-lingual dependency parsing experiments, where we train the parser on one language and test on our target language. As our test case, we use Xibe, a s
everely under-resourced Tungusic language. We assume that choosing a closely related language as the source language will provide better results than more distant relatives. However, it is not clear how to determine those closely related languages. We investigate three different methods: choosing the typologically closest language, using LangRank, and choosing the most similar language based on perplexity. We train parsing models on the selected languages using UDify and test on different genres of Xibe data. The results show that languages selected based on typology and perplexity scores outperform those predicted by LangRank; Japanese is the optimal source language. In determining the source language, proximity to the target language is more important than large training sizes. Parsing is also influenced by genre differences, but they have little influence as long as the training data is at least as complex as the target.
We review two features of mixture of experts (MoE) models which we call averaging and clustering effects in the context of graph-based dependency parsers learned in a supervised probabilistic framework. Averaging corresponds to the ensemble combinati
on of parsers and is responsible for variance reduction which helps stabilizing and improving parsing accuracy. Clustering describes the capacity of MoE models to give more credit to experts believed to be more accurate given an input. Although promising, this is difficult to achieve, especially without additional data. We design an experimental set-up to study the impact of these effects. Whereas averaging is always beneficial, clustering requires good initialization and stabilization techniques, but its advantages over mere averaging seem to eventually vanish when enough experts are present. As a by product, we show how this leads to state-of-the-art results on the PTB and the CoNLL09 Chinese treebank, with low variance across experiments.
We describe the NUIG solution for IWPT 2021 Shared Task of Enhanced Dependency (ED) parsing in multiple languages. For this shared task, we propose and evaluate an End-to-end Seq2seq mBERT-based ED parser which predicts the ED-parse tree of a given i
nput sentence as a relative head-position tag-sequence. Our proposed model is a multitasking neural-network which performs five key tasks simultaneously namely UPOS tagging, UFeat tagging, Lemmatization, Dependency-parsing and ED-parsing. Furthermore we utilise the linguistic typology available in the WALS database to improve the ability of our proposed end-to-end parser to transfer across languages. Results show that our proposed Seq2seq ED-parser performs on par with state-of-the-art ED-parser despite having a much simpler de- sign.
This paper describes a methodology for syntactic knowledge transfer between high-resource languages to extremely low-resource languages. The methodology consists in leveraging multilingual BERT self-attention model pretrained on large datasets to dev
elop a multilingual multi-task model that can predict Universal Dependencies annotations for three African low-resource languages. The UD annotations include universal part-of-speech, morphological features, lemmas, and dependency trees. In our experiments, we used multilingual word embeddings and a total of 11 Universal Dependencies treebanks drawn from three high-resource languages (English, French, Norwegian) and three low-resource languages (Bambara, Wolof and Yoruba). We developed various models to test specific language combinations involving contemporary contact languages or genetically related languages. The results of the experiments show that multilingual models that involve high-resource languages and low-resource languages with contemporary contact between each other can provide better results than combinations that only include unrelated languages. As far genetic relationships are concerned, we could not draw any conclusion regarding the impact of language combinations involving the selected low-resource languages, namely Wolof and Yoruba.
The introduction of pre-trained transformer-based contextualized word embeddings has led to considerable improvements in the accuracy of graph-based parsers for frameworks such as Universal Dependencies (UD). However, previous works differ in various
dimensions, including their choice of pre-trained language models and whether they use LSTM layers. With the aims of disentangling the effects of these choices and identifying a simple yet widely applicable architecture, we introduce STEPS, a new modular graph-based dependency parser. Using STEPS, we perform a series of analyses on the UD corpora of a diverse set of languages. We find that the choice of pre-trained embeddings has by far the greatest impact on parser performance and identify XLM-R as a robust choice across the languages in our study. Adding LSTM layers provides no benefits when using transformer-based embeddings. A multi-task training setup outputting additional UD features may contort results. Taking these insights together, we propose a simple but widely applicable parser architecture and configuration, achieving new state-of-the-art results (in terms of LAS) for 10 out of 12 diverse languages.
AMR (Abstract Meaning Representation) and EDS (Elementary Dependency Structures) are two popular meaning representations in NLP/NLU. AMR is more abstract and conceptual, while EDS is more low level, closer to the lexical structures of the given sente
nces. It is thus not surprising that EDS parsing is easier than AMR parsing. In this work, we consider using information from EDS parsing to help improve the performance of AMR parsing. We adopt a transition-based parser and propose to add EDS graphs as additional semantic features using a graph encoder composed of LSTM layer and GCN layer. Our experimental results show that the additional information from EDS parsing indeed gives a boost to the performance of the base AMR parser used in our experiments.
