Recently, the emergence of pre-trained models (PTMs) has brought natural language processing (NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly ...introduce language representation learning and its research progress. Then we systematically categorize existing PTMs based on a taxonomy from four different perspectives. Next, we describe how to adapt the knowledge of PTMs to downstream tasks. Finally, we outline some potential directions of PTMs for future research. This survey is purposed to be a hands-on guide for understanding, using, and developing PTMs for various NLP tasks.
MarIA: Spanish Language Models Gutiérrez-Fandiño, Asier; Armengol-Estapé, Jordi; Pàmies, Marc ...
Procesamiento del Lenguaje Natural,
03/2022, Letnik:
68
Journal Article
Odprti dostop
This work presents MarIA, a family of Spanish language models and associated resources made available to the industry and the research community. Currently, MarIA includes RoBERTa-base, ...RoBERTa-large, GPT2 and GPT2-large Spanish language models, which can arguably be presented as the largest and most proficient language models in Spanish. The models were pretrained using a massive corpus of 570GB of clean and deduplicated texts with 135 billion words extracted from the Spanish Web Archive crawled by the National Library of Spain between 2009 and 2019. We assessed the performance of the models with nine existing evaluation datasets and with a novel extractive Question Answering dataset created ex novo. Overall, MarIA models outperform the existing Spanish models across a variety of NLU tasks and training settings.
To examine the effect of a communication intervention package on expressive communication and visual attention in individuals with Rett syndrome.
A modified withdrawal (A-B1-A1-B2-A2) single case ...experimental design with a direct inter-subject replication across three participants was applied. Three women with Rett syndrome participated. The study took place over a six-week period and comprised 32 sessions with each participant. All sessions were video recorded. During the intervention the communication partner used aided language modelling on a gaze-controlled device in combination with using responsive partner strategies. Expressive communication was assessed as synthesised words per minute and unique synthesised words per minute. Visual attention was assessed as rate of focused gazes (1 s or longer) in interaction.
An intervention effect was found on the rate of unique words for all participants. The rate of words increased for two participants when the intervention was introduced but no withdrawal effect could be seen. An intervention effect on visual attention could be seen for one participant. The intervention appeared to have social validity as reported by caregivers.
Aided language modelling (ALM), while using responsive partner strategies and a gaze-controlled device may be used with adult individuals with Rett syndrome to increase their rate of expressive communication. Detailed observational measures revealed individual learning patterns, which may provide clinically valuable insights.
Implications for Rehabilitation
Adults with Rett syndrome may benefit from access to gaze-controlled devices in combination with responsive partner strategies.
Responsive partner communication may be effective for some individuals with Rett syndrome to increase their rate of synthesised utterances.
Rate of focused gazes may be considered as an outcome measure for individuals with oculomotor difficulties when introducing aided language modelling.
The aim of this paper is to investigate the benefit of combining both language and acoustic modelling for speaker diarization. Although conventional systems only use acoustic features, in some ...scenarios linguistic data contain high discriminative speaker information, even more reliable than the acoustic ones. In this study we analyze how an appropriate fusion of both kind of features is able to obtain good results in these cases. The proposed system is based on an iterative algorithm where a LSTM network is used as a speaker classifier. The network is fed with character-level word embeddings and a GMM based acoustic score created with the output labels from previous iterations. The presented algorithm has been evaluated in a Call-Center database, which is composed of telephone interview audios. The combination of acoustic features and linguistic content shows a 84.29% improvement in terms of a word-level DER as compared to a HMM/VB baseline system. The results of this study confirms that linguistic content can be efficiently used for some speaker recognition tasks.
•Linguistic content can be efficiently combined with acoustic features for the speaker diarization task.•Given a specific scenario, speaker diarization can be solved with only linguistic content.•Acoustic features lead to a better diarizatcon performance in comparison with linguistic content in large speaker segments.•Linguistic content is more discriminative than acoustic features to identify speakers in short speech segments.
ABSTRACT This research aimed to evaluate the performance of the A Lite BERT (ALBERT), efficiently learning an encoder that classifies token replacements accurately (ELECTRA) and a robust optimized ...BERT pretraining approach (RoBERTa) models to support the development of the Indonesian language question and answer system model. The two problems above, namely sorting candidate documents and validating answers have been handled by several methods such as the application of long-short term memory-recurrent neural network (LSTM-RNN) 12, template convolutional recurrent neural network (T-CRNN) 13, CNN-BiLSTM 14, dynamic co-attention networks (DCN) 15. ...section 4 presents the conclusion of the paper. Based on the proposed method in Figure 1, the article about coronavirus disease 2019 (COVID'19) news (we got it from crawling results on Indonesian Wikipedia, Jakarta News, Okezone, Antara, Kumparan, Tribune, and Open Super-large Crawled ALMAnaCH coRpus (OSCAR)) which is the input data for our study in preprocessing and converting the format to be used as input data for our study as a knowledge base system.
Abstract
Previous research has demonstrated that language use can vary depending on the context of situation. The present paper extends this finding by comparing word predictability differences ...between 14 speech registers ranging from highly informal conversations to read-aloud books. We trained 14 statistical language models to compute register-specific word predictability and trained a register classifier on the perplexity score vector of the language models. The classifier distinguishes perfectly between samples from all speech registers and this result generalizes to unseen materials. We show that differences in vocabulary and sentence length cannot explain the speech register classifier’s performance. The combined results show that speech registers differ in word predictability.
Detecting plagiarism poses a significant challenge for academic institutions, research centers, and content-centric organizations, especially in cases involving subtle paraphrasing and content ...manipulation where conventional methods often prove inadequate. Our paper proposes FTLM (Fuzzy TOPSIS Language Modeling), a novel method for detecting plagiarism within decision science. FTLM integrates language models with fuzzy sorting techniques to assess plagiarism severity by evaluating the similarity of potential solutions to a reference. The method involves two stages: leveraging language modeling to define criteria and alternatives and implementing enhanced fuzzy TOPSIS. Word usage patterns, grammatical structures, and semantic coherence represent fuzzy membership functions. Moreover, pre-trained language models enhance semantic similarity analysis. This approach highlights the benefits of combining fuzzy logic's tolerance for imprecision with the semantic evaluation capabilities of advanced language models, thereby offering a comprehensive and contextually aware method for analyzing plagiarism severity. The experimental results on the benchmark dataset demonstrate effective features that enhance performance on the user-defined severity ranking order.
The amount of archaeological literature is growing rapidly. Until recently, these data were only accessible through metadata search. We implemented a text retrieval engine for a large archaeological ...text collection (~658 million words). In archaeological IR, domain-specific entities such as locations, time periods and artefacts play a central role. This motivated the development of a named entity recognition (NER) model to annotate the full collection with archaeological named entities. In this article, we present ArcheoBERTje, a BERT (Bidirectional Encoder Representations from Transformers) model pre-trained on Dutch archaeological texts. We compare the model’s quality and output on an NER task to a generic multilingual model and a generic Dutch model. We also investigate ensemble methods for combining multiple BERT models, and combining the best BERT model with a domain thesaurus using conditional random fields. We find that ArcheoBERTje outperforms both the multilingual and Dutch model significantly with a smaller standard deviation between runs, reaching an average F1 score of 0.735. The model also outperforms ensemble methods combining the three models. Combining ArcheoBERTje predictions and explicit domain knowledge from the thesaurus did not increase the F1 score. We quantitatively and qualitatively analyse the differences between the vocabulary and output of the BERT models on the full collection and provide some valuable insights in the effect of fine-tuning for specific domains. Our results indicate that for a highly specific text domain such as archaeology, further pre-training on domain-specific data increases the model’s quality on NER by a much larger margin than shown for other domains in the literature, and that domain-specific pre-training makes the addition of domain knowledge from a thesaurus unnecessary.
Continual learning systems will interact with humans, with each other, and with the physical world through time - and continue to learn and adapt as they do. An important open problem for continual ...learning is a large-scale benchmark which enables realistic evaluation of algorithms. In this paper, we study a natural setting for continual learning on a massive scale. We introduce the problem of personalized online language learning (POLL), which involves fitting personalized language models to a population of users that evolves over time. To facilitate research on POLL, we collect massive datasets of Twitter posts. These datasets, Firehose10 M and Firehose100 M, comprise 100 million tweets, posted by one million users over six years. Enabled by the Firehose datasets, we present a rigorous evaluation of continual learning algorithms on an unprecedented scale. Based on this analysis, we develop a simple algorithm for continual gradient descent (ConGraD) that outperforms prior continual learning methods on the Firehose datasets as well as earlier benchmarks. Collectively, the POLL problem setting, the Firehose datasets, and the ConGraD algorithm enable a complete benchmark for reproducible research on web-scale continual learning.