scholarly journals Efficient Contextual Representation Learning With Continuous Outputs

2019 ◽  
Vol 7 ◽  
pp. 611-624 ◽  
Author(s):  
Liunian Harold Li ◽  
Patrick H. Chen ◽  
Cho-Jui Hsieh ◽  
Kai-Wei Chang
Keyword(s):  
Target Word ◽  
Language Processing ◽  
Language Model ◽  
Representation Learning ◽  
Great Success ◽  
Training Procedure ◽  
Vocabulary Size ◽  
Speed Advantage ◽  
Contextual Representation ◽  
High Computational Complexity

Contextual representation models have achieved great success in improving various downstream natural language processing tasks. However, these language-model-based encoders are difficult to train due to their large parameter size and high computational complexity. By carefully examining the training procedure, we observe that the softmax layer, which predicts a distribution of the target word, often induces significant overhead, especially when the vocabulary size is large. Therefore, we revisit the design of the output layer and consider directly predicting the pre-trained embedding of the target word for a given context. When applied to ELMo, the proposed approach achieves a 4-fold speedup and eliminates 80% trainable parameters while achieving competitive performance on downstream tasks. Further analysis shows that the approach maintains the speed advantage under various settings, even when the sentence encoder is scaled up.

scholarly journals TRS: Transformers for Remote Sensing Scene Classification

2021 ◽  
Vol 13 (20) ◽  
pp. 4143
Author(s):  
Jianrong Zhang ◽  
Hongwei Zhao ◽  
Jiao Li
Keyword(s):  
Remote Sensing ◽  
Language Processing ◽  
State Of The Art ◽  
Representation Learning ◽  
Learning Performance ◽  
Great Success ◽  
Scene Classification ◽  
Linear Classifier ◽  
Classification Tasks ◽  
Multiple Patches

Remote sensing scene classification remains challenging due to the complexity and variety of scenes. With the development of attention-based methods, Convolutional Neural Networks (CNNs) have achieved competitive performance in remote sensing scene classification tasks. As an important method of the attention-based model, the Transformer has achieved great success in the field of natural language processing. Recently, the Transformer has been used for computer vision tasks. However, most existing methods divide the original image into multiple patches and encode the patches as the input of the Transformer, which limits the model’s ability to learn the overall features of the image. In this paper, we propose a new remote sensing scene classification method, Remote Sensing Transformer (TRS), a powerful “pure CNNs→Convolution + Transformer → pure Transformers” structure. First, we integrate self-attention into ResNet in a novel way, using our proposed Multi-Head Self-Attention layer instead of 3 × 3 spatial revolutions in the bottleneck. Then we connect multiple pure Transformer encoders to further improve the representation learning performance completely depending on attention. Finally, we use a linear classifier for classification. We train our model on four public remote sensing scene datasets: UC-Merced, AID, NWPU-RESISC45, and OPTIMAL-31. The experimental results show that TRS exceeds the state-of-the-art methods and achieves higher accuracy.

scholarly journals Distill BERT to Traditional Models in Chinese Machine Reading Comprehension (Student Abstract)

2020 ◽  
Vol 34 (10) ◽  
pp. 13901-13902
Author(s):  
Xingkai Ren ◽  
Ronghua Shi ◽  
Fangfang Li
Keyword(s):  
Reading Comprehension ◽  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Language Model ◽  
Representation Learning ◽  
Original Model ◽  
Language Models ◽  
Machine Reading

Recently, unsupervised representation learning has been extremely successful in the field of natural language processing. More and more pre-trained language models are proposed and achieved the most advanced results especially in machine reading comprehension. However, these proposed pre-trained language models are huge with hundreds of millions of parameters that have to be trained. It is quite time consuming to use them in actual industry. Thus we propose a method that employ a distillation traditional reading comprehension model to simplify the pre-trained language model so that the distillation model has faster reasoning speed and higher inference accuracy in the field of machine reading comprehension. We evaluate our proposed method on the Chinese machine reading comprehension dataset CMRC2018 and greatly improve the accuracy of the original model. To the best of our knowledge, we are the first to propose a method that employ the distillation pre-trained language model in Chinese machine reading comprehension.

scholarly journals BioVerbNet: a large semantic-syntactic classification of verbs in biomedicine

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Olga Majewska ◽  
Charlotte Collins ◽  
Simon Baker ◽  
Jari Björne ◽  
Susan Windisch Brown ◽  
...  
Keyword(s):  
Natural Language ◽  
Language Processing ◽  
Model Performance ◽  
Representation Learning ◽  
Verb Classes ◽  
Expert Annotation ◽  
Biomedical Texts ◽  
Time Required ◽  
Verb Meaning

Abstract Background Recent advances in representation learning have enabled large strides in natural language understanding; However, verbal reasoning remains a challenge for state-of-the-art systems. External sources of structured, expert-curated verb-related knowledge have been shown to boost model performance in different Natural Language Processing (NLP) tasks where accurate handling of verb meaning and behaviour is critical. The costliness and time required for manual lexicon construction has been a major obstacle to porting the benefits of such resources to NLP in specialised domains, such as biomedicine. To address this issue, we combine a neural classification method with expert annotation to create BioVerbNet. This new resource comprises 693 verbs assigned to 22 top-level and 117 fine-grained semantic-syntactic verb classes. We make this resource available complete with semantic roles and VerbNet-style syntactic frames. Results We demonstrate the utility of the new resource in boosting model performance in document- and sentence-level classification in biomedicine. We apply an established retrofitting method to harness the verb class membership knowledge from BioVerbNet and transform a pretrained word embedding space by pulling together verbs belonging to the same semantic-syntactic class. The BioVerbNet knowledge-aware embeddings surpass the non-specialised baseline by a significant margin on both tasks. Conclusion This work introduces the first large, annotated semantic-syntactic classification of biomedical verbs, providing a detailed account of the annotation process, the key differences in verb behaviour between the general and biomedical domain, and the design choices made to accurately capture the meaning and properties of verbs used in biomedical texts. The demonstrated benefits of leveraging BioVerbNet in text classification suggest the resource could help systems better tackle challenging NLP tasks in biomedicine.

scholarly journals Automated Source Code Generation and Auto-Completion Using Deep Learning: Comparing and Discussing Current Language Model-Related Approaches

AI ◽  
2021 ◽  
Vol 2 (1) ◽  
pp. 1-16
Author(s):  
Juan Cruz-Benito ◽  
Sanjay Vishwakarma ◽  
Francisco Martin-Fernandez ◽  
Ismael Faro
Keyword(s):  
Deep Learning ◽  
Learning Community ◽  
Programming Languages ◽  
Language Processing ◽  
Code Generation ◽  
Language Model ◽  
Language Models ◽  
Stochastic Gradient Descent ◽  
Network Architectures ◽  
Learning Architectures

In recent years, the use of deep learning in language models has gained much attention. Some research projects claim that they can generate text that can be interpreted as human writing, enabling new possibilities in many application areas. Among the different areas related to language processing, one of the most notable in applying this type of modeling is programming languages. For years, the machine learning community has been researching this software engineering area, pursuing goals like applying different approaches to auto-complete, generate, fix, or evaluate code programmed by humans. Considering the increasing popularity of the deep learning-enabled language models approach, we found a lack of empirical papers that compare different deep learning architectures to create and use language models based on programming code. This paper compares different neural network architectures like Average Stochastic Gradient Descent (ASGD) Weight-Dropped LSTMs (AWD-LSTMs), AWD-Quasi-Recurrent Neural Networks (QRNNs), and Transformer while using transfer learning and different forms of tokenization to see how they behave in building language models using a Python dataset for code generation and filling mask tasks. Considering the results, we discuss each approach’s different strengths and weaknesses and what gaps we found to evaluate the language models or to apply them in a real programming context.

Children With Cochlear Implants Use Semantic Prediction to Facilitate Spoken Word Recognition

2021 ◽  
pp. 1-14
Author(s):  
Christina Blomquist ◽  
Rochelle S. Newman ◽  
Yi Ting Huang ◽  
Jan Edwards
Keyword(s):  
Word Recognition ◽  
Target Word ◽  
Cochlear Implants ◽  
Language Processing ◽  
Spoken Word Recognition ◽  
Spoken Language ◽  
Spoken Word ◽  
Semantic Cues ◽  
Matched Group ◽  
Spoken Language Processing

Purpose Children with cochlear implants (CIs) are more likely to struggle with spoken language than their age-matched peers with normal hearing (NH), and new language processing literature suggests that these challenges may be linked to delays in spoken word recognition. The purpose of this study was to investigate whether children with CIs use language knowledge via semantic prediction to facilitate recognition of upcoming words and help compensate for uncertainties in the acoustic signal. Method Five- to 10-year-old children with CIs heard sentences with an informative verb ( draws ) or a neutral verb ( gets ) preceding a target word ( picture ). The target referent was presented on a screen, along with a phonologically similar competitor ( pickle ). Children's eye gaze was recorded to quantify efficiency of access of the target word and suppression of phonological competition. Performance was compared to both an age-matched group and vocabulary-matched group of children with NH. Results Children with CIs, like their peers with NH, demonstrated use of informative verbs to look more quickly to the target word and look less to the phonological competitor. However, children with CIs demonstrated less efficient use of semantic cues relative to their peers with NH, even when matched for vocabulary ability. Conclusions Children with CIs use semantic prediction to facilitate spoken word recognition but do so to a lesser extent than children with NH. Children with CIs experience challenges in predictive spoken language processing above and beyond limitations from delayed vocabulary development. Children with CIs with better vocabulary ability demonstrate more efficient use of lexical-semantic cues. Clinical interventions focusing on building knowledge of words and their associations may support efficiency of spoken language processing for children with CIs. Supplemental Material https://doi.org/10.23641/asha.14417627

Astrid

2020 ◽  
Vol 14 (4) ◽  
pp. 471-484
Author(s):  
Suraj Shetiya ◽  
Saravanan Thirumuruganathan ◽  
Nick Koudas ◽  
Gautam Das
Keyword(s):  
Deep Learning ◽  
Objective Function ◽  
Pattern Matching ◽  
Language Processing ◽  
Language Model ◽  
Language Models ◽  
Selectivity Estimation ◽  
Statistical Correlations ◽  
Benchmark Datasets ◽  
Traditional Approaches

Accurate selectivity estimation for string predicates is a long-standing research challenge in databases. Supporting pattern matching on strings (such as prefix, substring, and suffix) makes this problem much more challenging, thereby necessitating a dedicated study. Traditional approaches often build pruned summary data structures such as tries followed by selectivity estimation using statistical correlations. However, this produces insufficiently accurate cardinality estimates resulting in the selection of sub-optimal plans by the query optimizer. Recently proposed deep learning based approaches leverage techniques from natural language processing such as embeddings to encode the strings and use it to train a model. While this is an improvement over traditional approaches, there is a large scope for improvement. We propose Astrid, a framework for string selectivity estimation that synthesizes ideas from traditional and deep learning based approaches. We make two complementary contributions. First, we propose an embedding algorithm that is query-type (prefix, substring, and suffix) and selectivity aware. Consider three strings 'ab', 'abc' and 'abd' whose prefix frequencies are 1000, 800 and 100 respectively. Our approach would ensure that the embedding for 'ab' is closer to 'abc' than 'abd'. Second, we describe how neural language models could be used for selectivity estimation. While they work well for prefix queries, their performance for substring queries is sub-optimal. We modify the objective function of the neural language model so that it could be used for estimating selectivities of pattern matching queries. We also propose a novel and efficient algorithm for optimizing the new objective function. We conduct extensive experiments over benchmark datasets and show that our proposed approaches achieve state-of-the-art results.

scholarly journals TripletProt: Deep Representation Learning of Proteins based on Siamese Networks

2020 ◽  
Author(s):  
Esmaeil Nourani ◽  
Ehsaneddin Asgari ◽  
Alice C. McHardy ◽  
Mohammad R.K. Mofrad
Keyword(s):  
Functional Annotation ◽  
Cellular Localization ◽  
State Of The Art ◽  
Language Model ◽  
Representation Learning ◽  
Learning Problems ◽  
Ppi Network ◽  
New Approach ◽  
Protein Protein Interaction ◽  
Siamese Networks

AbstractWe introduce TripletProt, a new approach for protein representation learning based on the Siamese neural networks. We evaluate TripletProt comprehensively in protein functional annotation tasks including sub-cellular localization (14 categories) and gene ontology prediction (more than 2000 classes), which are both challenging multi-class multi-label classification machine learning problems. We compare the performance of TripletProt with the state-of-the-art approaches including recurrent language model-based approach (i.e., UniRep), as well as protein-protein interaction (PPI) network and sequence-based method (i.e., DeepGO). Our TripletProt showed an overall improvement of F1 score in the above mentioned comprehensive functional annotation tasks, solely relying on the PPI network. TripletProt and in general Siamese Network offer great potentials for the protein informatics tasks and can be widely applied to similar tasks.

scholarly journals EMOSIS Sentiment Analysis on Tweets with Emotion and Intensity Level Recognition Considering Ending Punctuation Marks

2019 ◽  
Vol 8 (4) ◽  
pp. 10289-10293
Keyword(s):  
Natural Language Processing ◽  
Natural Language ◽  
Emotion Recognition ◽  
Sentiment Analysis ◽  
Language Processing ◽  
Significant Role ◽  
Language Model ◽  
Intensity Level ◽  
Processing Stage ◽  
Overall Performance

Sentiment Analysis is a tool used for determining the Polarity or Emotion of a Sentence. It is a field of Natural Language Processing which focuses on the study of opinions. In this study, the researchers solved one key challenge in Sentiment Analysis, which is to consider the Ending Punctuation Marks present in a sentence. Ending punctuation marks plays a significant role in Emotion Recognition and Intensity Level Recognition. The research made used of tweets expressing opinions about Philippine President Rodrigo Duterte. These downloaded tweets served as the inputs. It was initially subjected to pre-processing stage to be able to prepare the sentences for processing. A Language Model was created to serve as the classifier for determining the scores of the tweets. The scores give the polarity of the sentence. Accuracy is very important in sentiment analysis. To increase the chance of correctly identifying the polarity of the tweets, the input undergone Intensity Level Recognition which determines the intensifiers and negations within the sentences. The system was evaluated with overall performance of 80.27%.

A Natural Language Processing Pipeline to Identify COVID-19 Outbreaks from Contact Tracing Interview Forms (Preprint)

2022 ◽  
Author(s):  
John Caskey ◽  
Iain L McConnell ◽  
Madeline Oguss ◽  
Dmitriy Dligach ◽  
Rachel Kulikoff ◽  
...  
Keyword(s):  
Language Processing ◽  
Disease Surveillance ◽  
Language Model ◽  
Entity Recognition ◽  
Contact Tracing ◽  
Free Text ◽  
Local Health ◽  
Targeted Interventions ◽  
Mapping Tool ◽  
Location Mapping

BACKGROUND In Wisconsin, COVID-19 case interview forms contain free text fields that need to be mined to identify potential outbreaks for targeted policy making. We developed an automated pipeline to ingest the free text into a pre-trained neural language model to identify businesses and facilities as outbreaks. OBJECTIVE We aim to examine the performance of our pipeline. METHODS Data on cases of COVID-19 were extracted from the Wisconsin Electronic Disease Surveillance System (WEDSS) for Dane County between July 1, 2020, and June 30, 2021. Features from the case interview forms were fed into a Bidirectional Encoder Representations from Transformers (BERT) model that was fine-tuned for named entity recognition (NER). We also developed a novel location mapping tool to provide addresses for relevant NERs. The pipeline was validated against known outbreaks that were already investigated and confirmed. RESULTS There were 46,898 cases of COVID-19 with 4,183,273 total BERT tokens and 15,051 unique tokens. The recall and precision of the NER tool were 0.67 (95 % CI 0.66-0.68) and 0.55 (95 % CI: 0.54-0.57), respectively. For the location mapping tool, the recall and precision were 0.93 (95% CI: 0.92-0.95) and 0.93 (95% CI: 0.92-0.95), respectively. Across monthly intervals, the NER tool identified more potential clusters than were confirmed in the WEDSS system. CONCLUSIONS We developed a novel pipeline of tools that identified existing outbreaks and novel clusters with associated addresses. Our pipeline ingests data from a statewide database and may be deployed to assist local health departments for targeted interventions. CLINICALTRIAL Not applicable

scholarly journals BERTMeSH: deep contextual representation learning for large-scale high-performance MeSH indexing with full text

2020 ◽  
Author(s):  
Ronghui You ◽  
Yuxuan Liu ◽  
Hiroshi Mamitsuka ◽  
Shanfeng Zhu
Keyword(s):  
Full Text ◽  
High Performance ◽  
Large Scale ◽  
Learning Strategy ◽  
Learning To Rank ◽  
Representation Learning ◽  
Supplementary Information ◽  
Medical Subject Headings ◽  
The Difference ◽  
Contextual Representation

Abstract Motivation With the rapid increase of biomedical articles, large-scale automatic Medical Subject Headings (MeSH) indexing has become increasingly important. FullMeSH, the only method for large-scale MeSH indexing with full text, suffers from three major drawbacks: FullMeSH (i) uses Learning To Rank, which is time-consuming, (ii) can capture some pre-defined sections only in full text and (iii) ignores the whole MEDLINE database. Results We propose a computationally lighter, full text and deep-learning-based MeSH indexing method, BERTMeSH, which is flexible for section organization in full text. BERTMeSH has two technologies: (i) the state-of-the-art pre-trained deep contextual representation, Bidirectional Encoder Representations from Transformers (BERT), which makes BERTMeSH capture deep semantics of full text. (ii) A transfer learning strategy for using both full text in PubMed Central (PMC) and title and abstract (only and no full text) in MEDLINE, to take advantages of both. In our experiments, BERTMeSH was pre-trained with 3 million MEDLINE citations and trained on ∼1.5 million full texts in PMC. BERTMeSH outperformed various cutting-edge baselines. For example, for 20 K test articles of PMC, BERTMeSH achieved a Micro F-measure of 69.2%, which was 6.3% higher than FullMeSH with the difference being statistically significant. Also prediction of 20 K test articles needed 5 min by BERTMeSH, while it took more than 10 h by FullMeSH, proving the computational efficiency of BERTMeSH. Supplementary information Supplementary data are available at Bioinformatics online

Sign in / Sign up

Export Citation Format

Share Document