Search

This study attempts to investigate cross-strait variations on two typical synonymous loanwords in Chinese, i.e. xie2shang1 and tan2pan4, drawn on MARVS theory. Through a comparative analysis, the study found some distributional, eventual, and contextual similarities and differences across Taiwan and Mainland Mandarin. Compared with the underused tan2pan4, xie2shang1 is significantly overused in Taiwan Mandarin and vice versa in Mainland Mandarin. Additionally, though both words can refer to an inchoative process in Mainland and Taiwan Mandarin, the starting point for xie2shang1 in Mainland Mandarin is somewhat blurring compared with the usage in Taiwan Mandarin. Further on, in Taiwan Mandarin, tan2pan4 can be used in economic and diplomatic contexts, while xie2shang1 is used almost exclusively in political contexts. In Mainland Mandarin, however, the two words can be used in a hybrid manner within political contexts; moreover, tan2pan4 is prominently used in diplomatic contexts with less reference to economic activities, while xie2sahng1 can be found in both political and legal contexts, emphasizing a role of mediation., To appear in PACLIC 2022. 10 pages, 5 figures, 5 tables

Supplemental material, sj-pdf-1-sgo-10.1177_21582440211031556 for From Contact Prevention to Social Distancing: The Co-Evolution of Bilingual Neologisms and Public Health Campaigns in Two Cities in the Time of COVID-19 by Xiaowen Wang and Chu-Ren Huang in SAGE Open

Supplemental material, sj-pdf-1-sgo-10.1177_21582440211031556 for From Contact Prevention to Social Distancing: The Co-Evolution of Bilingual Neologisms and Public Health Campaigns in Two Cities in the Time of COVID-19 by Xiaowen Wang and Chu-Ren Huang in SAGE Open

This paper examines gender and age salience and (stereo)typicality in British English talk with the aim to predict gender and age categories based on lexical, phrasal and turn-taking features. We examine the SpokenBNC, a corpus of around 11.4 million words of British English conversations and identify behavioural differences between speakers that are labelled for gender and age categories. We explore differences in language use and turn-taking dynamics and identify a range of characteristics that set the categories apart. We find that female speakers tend to produce more and slightly longer turns, while turns by male speakers feature a higher type-token ratio and a distinct range of minimal particles such as "eh", "uh" and "em". Across age groups, we observe, for instance, that swear words and laughter characterize young speakers' talk, while old speakers tend to produce more truncated words. We then use the observed characteristics to predict gender and age labels of speakers per conversation and per turn as a classification task, showing that non-lexical utterances such as minimal particles that are usually left out of dialog data can contribute to setting the categories apart., Comment: 10 pages

English research articles (RAs) are an essential genre in academia, so the attempts to employ NLP to assist the development of academic writing ability have received considerable attention in the last two decades. However, there has been no study employing feature engineering techniques to investigate the linguistic features of RAs of different academic impacts (i.e., the papers of high/moderate citation times published in the journals of high/moderate impact factors). This study attempts to extract micro-level linguistic features in high- and moderate-impact journal RAs, using feature engineering methods. We extracted 25 highly relevant features from the Corpus of English Journal Articles through feature selection methods. All papers in the corpus deal with COVID-19 medical empirical studies. The selected features were then validated of the classification performance in terms of consistency and accuracy through supervised machine learning methods. Results showed that 24 linguistic features such as the overlapping of content words between adjacent sentences, the use of third-person pronouns, auxiliary verbs, tense, emotional words provide consistent and accurate predictions for journal articles with different academic impacts. Lastly, the random forest model is shown to be the best model to fit the relationship between these 24 features and journal articles with high and moderate impacts. These findings can be used to inform academic writing courses and lay the foundation for developing automatic evaluation systems for L2 graduate students.

International audience; In linguistics and cognitive science, Logical metonymies are defined as type clashes between an event-selecting verb and an entitydenoting noun (e.g. The editor finished the article), which are typically interpreted by inferring a hidden event (e.g. reading) on the basis of contextual cues. This paper tackles the problem of logical metonymy interpretation, that is, the retrieval of the covert event via computational methods. We compare different types of models, including the probabilistic and the distributional ones previously introduced in the literature on the topic. For the first time, we also tested on this task some of the recent Transformer-based models, such as BERT, RoBERTa, XLNet, and GPT-2. Our results show a complex scenario, in which the best Transformer-based models and some traditional distributional models perform very similarly. However, the low performance on some of the testing datasets suggests that logical metonymy is still a challenging phenomenon for computational modeling.

The AMR (Abstract Meaning Representation) formalism for representing meaning of natural language sentences puts emphasis on predicate-argument structure and was not designed to deal with scope and quantifiers. By extending AMR with indices for contexts and formulating constraints on these contexts, a formalism is derived that makes correct predictions for inferences involving negation and bound variables. The attractive core predicate-argument structure of AMR is preserved. The resulting framework is similar to the meaning representations of Discourse Representation Theory employed in the Parallel Meaning Bank.

In sentiment analysis of product reviews, both user and product information are proven to be useful. Current works handle user profile and product information in a unified model which may not be able to learn salient features of users and products effectively. In this work, we propose a dual user and product memory network (DUPMN) model to learn user profiles and product information for reviews classification using separate memory networks. Then, the two representations are used jointly for sentiment analysis. The use of separate models aims to capture user profiles and product information more effectively. Comparing with state-of-the-art unified prediction models, evaluations on three benchmark datasets (IMDB, Yelp13, and Yelp14) show that our dual learning model gives performance gain of 0.6%, 1.2%, and 0.9%, respectively. The improvements are also deemed very significant measured by p-values.

A realistic Chinese word segmentation tool must adapt to textual variations with minimal training input and yet robust enough to yield reliable segmentation result for all variants. Various lexicon-driven approaches to Chinese segmentation, e.g. [1,16], achieve high f-scores yet require massive training for any variation. Text-driven approach, e.g. [12], can be easily adapted for domain and genre changes yet has difficulty matching the high f-scores of the lexicon-driven approaches. In this paper, we refine and implement an innovative text-driven word boundary decision (WBD) segmentation model proposed in [15]. The WBD model treats word segmentation simply and efficiently as a binary decision on whether to realize the natural textual break between two adjacent characters as a word boundary. The WBD model allows simple and quick training data preparation converting characters as contextual vectors for learning the word boundary decision. Machine learning experiments with four different classifiers show that training with 1,000 vectors and 1 million vectors achieve comparable and reliable results. In addition, when applied to SigHAN Bakeoff 3 competition data, the WBD model produces OOV recall rates that are higher than all published results. Unlike all previous work, our OOV recall rate is comparable to our own F-score. Both experiments support the claim that the WBD model is a realistic model for Chinese word segmentation as it can be easily adapted for new variants with the robust result. In conclusion, we will discuss linguistic ramifications as well as future implications for the WBD approach., Proceedings of the 20th Conference on Computational Linguistics and Speech Processing

The workshop aimed to bring together members of the language-resource (LR) landscape, focusing on complex linguistic knowledge that requires linguistic expertise, e.g. on dictionaries, ontologies and grammars. Such manually-built resources are key to the development of natural language processing (NLP) tools and applications. We intended to strengthen the cohesion of the scientific ’production chain’ spanning from the construction of LRs to their exploitation in hybrid or symbolic NLP. It is necessary to increase mutual awareness between researchers along this production chain, regarding their activities, skills and needs, in view of improving the building processes of the resources, their validation and their exploitation. Many linguists are comfortable with descriptive tasks such as checking lexical entries for a given feature, even if each entry requires analysing or pondering. On the other hand, computer scientists are familiar with formalization and, usually, with notions such as falsifiability or reproducibility, which are fundamental to sciences. Combining all these skills is likely to stimulate innovation. The workshop offered an opportunity of interaction which is required to overcome the compartmentalization between humanities and sciences, and to intensify co-operation between the two ends of the chain.

International audience; Wiktionary, a satellite of the Wikipedia initiative, can be seen as a potential resource for Natural Language Processing. It requires however to be processed before being used efficiently as an NLP resource. After describing the relevant aspects Wiktionary for our purposes, we focus on its structural properties. Then, we describe how we extracted synonymy networks from this resource. We provide an in-depth study of these synonymy networks and compare them to those extracted from traditional resources. Finally, we describe two methods for semiautomatically improving this network by adding missing relations: (i) using a kind of semantic proximity measure; (ii) using translation relations of Wiktionary itself.

PACLIC 20 / Wuhan, China / 1-3 November, 2006

This paper seeks to improve the performance of automatic speech recognition (ASR) systems operating on code-switched speech. Code-switching refers to the alternation of languages within a conversation, a phenomenon that is of increasing importance considering the rapid rise in the number of bilingual speakers in the world. It is particularly challenging for ASR owing to the relative scarcity of code-switching speech and text data, even when the individual languages are themselves well-resourced. This paper proposes to overcome this challenge by applying linguistic theories in order to generate more realistic code-switching text, necessary for language modelling in ASR. Working with English-Spanish code-switching, we find that Equivalence Constraint theory and part-of-speech labelling are particularly helpful for text generation, and bring 2% improvement to ASR performance.

We analyze the learning dynamics of neural language and translation models using Loss Change Allocation (LCA), an indicator that enables a fine-grained analysis of parameter updates when optimizing for the loss function. In other words, we can observe the contributions of different network components at training time. In this article, we systematically study masked language modeling, causal language modeling, and machine translation. We show that the choice of training objective leads to distinctive optimization procedures, even when performed on comparable Transformer architectures. We demonstrate how the various Transformer parameters are used during training, supporting that the feed-forward components of each layer are the main contributors to the optimization procedure. Finally, we find that the learning dynamics are not affected by data size and distribution but rather determined by the learning objective.

Recombining known primitive concepts into larger novel combinations is a quintessentially human cognitive capability. Whether large neural models in NLP can acquire this ability while learning from data is an open question. In this paper, we investigate this problem from the perspective of formal languages. We use deterministic finite-state transducers to make an unbounded number of datasets with controllable properties governing compositionality. By randomly sampling over many transducers, we explore which of their properties contribute to learnability of a compositional relation by a neural network. We find that the models either learn the relations completely or not at all. The key is transition coverage, setting a soft learnability limit at 400 examples per transition., Proceedings of the 29th International Conference on Computational Linguistics

Prerecorded laughter accompanying dialog in comedy TV shows encourages the audience to laugh by clearly marking humorous moments in the show. We present an approach for automatically detecting humor in the Friends TV show using multimodal data. Our model is capable of recognizing whether an utterance is humorous or not and assess the intensity of it. We use the prerecorded laughter in the show as annotation as it marks humor and the length of the audience's laughter tells us how funny a given joke is. We evaluate the model on episodes the model has not been exposed to during the training phase. Our results show that the model is capable of correctly detecting whether an utterance is humorous 78% of the time and how long the audience's laughter reaction should last with a mean absolute error of 600 milliseconds., Comment: Outstanding paper award in COLING 2022

This paper presents M3L-Contrast -- a novel multimodal multilingual (M3L) neural topic model for comparable data that maps texts from multiple languages and images into a shared topic space. Our model is trained jointly on texts and images and takes advantage of pretrained document and image embeddings to abstract the complexities between different languages and modalities. As a multilingual topic model, it produces aligned language-specific topics and as multimodal model, it infers textual representations of semantic concepts in images. We demonstrate that our model is competitive with a zero-shot topic model in predicting topic distributions for comparable multilingual data and significantly outperforms a zero-shot model in predicting topic distributions for comparable texts and images. We also show that our model performs almost as well on unaligned embeddings as it does on aligned embeddings., Comment: Published in COLING 2022 Proceddings

The ability to generalize compositionally is key to understanding the potentially infinite number of sentences that can be constructed in a human language from only a finite number of words. Investigating whether NLP models possess this ability has been a topic of interest: SCAN (Lake and Baroni, 2018) is one task specifically proposed to test for this property. Previous work has achieved impressive empirical results using a group-equivariant neural network that naturally encodes a useful inductive bias for SCAN (Gordon et al., 2020). Inspired by this, we introduce a novel group-equivariant architecture that incorporates a group-invariant hard alignment mechanism. We find that our network's structure allows it to develop stronger equivariance properties than existing group-equivariant approaches. We additionally find that it outperforms previous group-equivariant networks empirically on the SCAN task. Our results suggest that integrating group-equivariance into a variety of neural architectures is a potentially fruitful avenue of research, and demonstrate the value of careful analysis of the theoretical properties of such architectures., Comment: Accepted at COLING 2022

Although Mandarin Chinese is shared by Chinese communities such as Mainland China, Taiwan, Hong Kong, and Singapore, linguistic differences are frequently found among regional uses, ranging from pronunciation, orthography, vocabulary, grammar, and discourse. Along with the increasingly recognized notion of “World Chineses” in recent years, the study of the regional variations has also become more linguistically, socially, and culturally significant. Such a study facilitates more efficient communication among speakers of different varieties, reflects the social and cultural differences of the Chinese speaking communities from a linguistic perspective, and contributes to the theoretical discussion of language variation and change. With specific examples of the linguistic features exhibited in Mainland China, Taiwan, Hong Kong, and Singapore Mandarin Chinese, this chapter is an overview of the current studies, methodologies, and motivations of variation.

This chapter describes the main types and usages of deictics and anaphors. The phenomenon of deixis involves the use of a word or phrase whose interpretation is determined through considerations of the physical properties in the situation of utterance, from the perspective of the speaker or the addressee who are engaged in the act of communication. When the need to make further structural distinctions is in order, a deictic consisting of one single word is referred to as a “deictic term,” while a larger deictic built around a deictic term is referred to as a “deictic expression.”

Texts comprise a large part of visual information that we process every day, so one of the tasks of language science is to make them more accessible. However, often the text design process is focused on the font size, but not on its type; which might be crucial especially for the people with reading disabilities. The current paper represents a study on text accessibility and the first attempt to create a research-based accessible font for Cyrillic letters. This resulted in the dyslexic-specific font, LexiaD. Its design rests on the reduction of interletter similarity of the Russian alphabet. In evaluation stage, dyslexic and non-dyslexic children were asked to read sentences from the Children version of the Russian Sentence Corpus. We tested the readability of LexiaD compared to PT Sans and PT Serif fonts. The results showed that all children had some advantage in letter feature extraction and information integration while reading in LexiaD, but lexical access was improved when sentences were rendered in PT Sans or PT Serif. Therefore, in several aspects, LexiaD proved to be faster to read and could be recommended to use by dyslexics who have visual deficiency or those who struggle with text understanding resulting in re-reading.