Evolution and Prospects of Foundation Models: From Large Language Models to Large Multimodal Models

Since the 1950s,when the Turing Test was introduced,there has been notable progress in machine language intelligence.Language modeling,crucial for AI development,has evolved from statistical to neural models over the last two decades.Recently,transformer-based Pre-trained Language Models(PLM)have excelled in Natural Language Processing(NLP)tasks by leveraging large-scale training corpora.Increasing the scale of these models enhances performance significantly,introducing abilities like context learning that smaller models lack.The advancement in Large Language Models,exemplified by the development of ChatGPT,has made significant impacts both academically and industrially,capturing widespread societal interest.This survey provides an overview of the development and prospects from Large Language Models(LLM)to Large Multimodal Models(LMM).It first discusses the contributions and technological advancements of LLMs in the field of natural language processing,especially in text generation and language understanding.Then,it turns to the discussion of LMMs,which integrates various data modalities such as text,images,and sound,demonstrating advanced capabilities in understanding and generating cross-modal content,paving new pathways for the adaptability and flexibility of AI systems.Finally,the survey highlights the prospects of LMMs in terms of technological development and application potential,while also pointing out challenges in data integration,cross-modal understanding accuracy,providing a comprehensive perspective on the latest developments in this field.

Enhancing Communication Accessibility:UrSL-CNN Approach to Urdu Sign Language Translation for Hearing-Impaired Individuals

Deaf people or people facing hearing issues can communicate using sign language(SL),a visual language.Many works based on rich source language have been proposed;however,the work using poor resource language is still lacking.Unlike other SLs,the visuals of the Urdu Language are different.This study presents a novel approach to translating Urdu sign language(UrSL)using the UrSL-CNN model,a convolutional neural network(CNN)architecture specifically designed for this purpose.Unlike existingworks that primarily focus on languageswith rich resources,this study addresses the challenge of translating a sign language with limited resources.We conducted experiments using two datasets containing 1500 and 78,000 images,employing a methodology comprising four modules:data collection,pre-processing,categorization,and prediction.To enhance prediction accuracy,each sign image was transformed into a greyscale image and underwent noise filtering.Comparative analysis with machine learning baseline methods(support vectormachine,GaussianNaive Bayes,randomforest,and k-nearest neighbors’algorithm)on the UrSL alphabets dataset demonstrated the superiority of UrSL-CNN,achieving an accuracy of 0.95.Additionally,our model exhibited superior performance in Precision,Recall,and F1-score evaluations.This work not only contributes to advancing sign language translation but also holds promise for improving communication accessibility for individuals with hearing impairments.

Systematizing Teacher Development:A Review of Foreign Language Teacher Learning

Foreign language teaching practice is developing rapidly,but research on foreign language teacher learning is currently relatively fragmented and unstructured.The book Foreign Language Teacher Learning,written by Professor Kang Yan from Capital Normal University,published in September 2022,makes a systematic introduction to foreign language teacher learning,which to some extent makes up for this shortcoming.Her book presents the lineage of foreign language teacher learning research at home and abroad,analyzes both theoretical and practical aspects,reviews the cuttingedge research results,and foresees the future development trend,painting a complete research picture for researchers in the field of foreign language teaching and teacher education as well as front-line teachers interested in foreign language teacher learning.This is an important inspiration for conducting foreign language teacher learning research in the future.And this paper makes a review of the book from aspects such as its content,major characteristics,contributions and limitations.

Plain language in the healthcare of Japan:a systematic review of“plain Japanese”

Objective:Despite the decrease in the number of foreign visitors and residents in Japan due to the coronavirus disease 2019,a resurgence is remarkable from 2022.However,Japan's medical support system for foreign patients,especially residents,is inadequate,with language barriers potentially causing health disparities.Comprehensive interpretation and translation services are challenging,but“plain Japanese”may be a viable alternative for foreign patients with basic Japanese language skills.This study explores the application and obstacles of plain Japanese in the medical sector.Methods:A literature review was performed across these databases:Web of Science,PubMed,Google Scholar,Scopus,CINAHL Plus,Springer Link and Ichushi-Web(Japanese medical literature).The search covered themes related to healthcare,care for foreign patients,and scholarly articles,and was conducted in July 2023.Results:The study incorporated five papers.Each paper emphasized the language barriers foreign residents in Japan face when accessing healthcare,highlighting the critical role and necessity of plain Japanese in medical environments.Most of the reports focused on the challenges of delivering medical care to foreign patients and the training of healthcare professionals in using plain Japanese for communication.Conclusion:The knowledge and application of plain Japanese among healthcare professionals are inadequate,and literature also remains scarce.With the increasing number of foreign residents in Japan,the establishment of a healthcare system that effectively uses plain Japanese is essential.However,plain Japanese may not be the optimal linguistic assistance in certain situations,thus it is imperative to encourage more research and reports on healthcare services using plain Japanese.

Impact of BRICS Trade Facilitation on China's Exports Using Cross-Border E-Commerce

With the rapid growth of the global digital economy, cross-border e-commerce, as an emerging form of trade, has gradually become a powerful engine to promote the development of global trade. BRICS is an important force in the global economy, and the progress of the BRICS countries' trade facilitation level has an important impact on the global trade environment. This paper conducts an in-depth study of the dynamic changes in BRICS trade facilitation from 2013 to 2022, and uses an extended gravity model to analyze the specific impact of this change on China's exports using cross-border e-commerce. The results show that although the BRICS countries have made some progress in trade facilitation, the overall level still needs to be improved, and there are obvious differences among member countries. However, the improvement of trade facilitation among BRICS countries has undoubtedly brought significant positive effects to China's exports using cross-border e-commerce.

Literature classification and its applications in condensed matter physics and materials science by natural language processing

The exponential growth of literature is constraining researchers’access to comprehensive information in related fields.While natural language processing(NLP)may offer an effective solution to literature classification,it remains hindered by the lack of labelled dataset.In this article,we introduce a novel method for generating literature classification models through semi-supervised learning,which can generate labelled dataset iteratively with limited human input.We apply this method to train NLP models for classifying literatures related to several research directions,i.e.,battery,superconductor,topological material,and artificial intelligence(AI)in materials science.The trained NLP‘battery’model applied on a larger dataset different from the training and testing dataset can achieve F1 score of 0.738,which indicates the accuracy and reliability of this scheme.Furthermore,our approach demonstrates that even with insufficient data,the not-well-trained model in the first few cycles can identify the relationships among different research fields and facilitate the discovery and understanding of interdisciplinary directions.

Automatic Generation of Attribute-Based Access Control Policies from Natural Language Documents

In response to the challenges of generating Attribute-Based Access Control(ABAC)policies,this paper proposes a deep learning-based method to automatically generate ABAC policies from natural language documents.This method is aimed at organizations such as companies and schools that are transitioning from traditional access control models to the ABAC model.The manual retrieval and analysis involved in this transition are inefficient,prone to errors,and costly.Most organizations have high-level specifications defined for security policies that include a set of access control policies,which often exist in the form of natural language documents.Utilizing this rich source of information,our method effectively identifies and extracts the necessary attributes and rules for access control from natural language documents,thereby constructing and optimizing access control policies.This work transforms the problem of policy automation generation into two tasks:extraction of access control statements andmining of access control attributes.First,the Chat General Language Model(ChatGLM)isemployed to extract access control-related statements from a wide range of natural language documents by constructing unique prompts and leveraging the model’s In-Context Learning to contextualize the statements.Then,the Iterated Dilated-Convolutions-Conditional Random Field(ID-CNN-CRF)model is used to annotate access control attributes within these extracted statements,including subject attributes,object attributes,and action attributes,thus reassembling new access control policies.Experimental results show that our method,compared to baseline methods,achieved the highest F1 score of 0.961,confirming the model’s effectiveness and accuracy.

Comparing Fine-Tuning, Zero and Few-Shot Strategies with Large Language Models in Hate Speech Detection in English

Large Language Models(LLMs)are increasingly demonstrating their ability to understand natural language and solve complex tasks,especially through text generation.One of the relevant capabilities is contextual learning,which involves the ability to receive instructions in natural language or task demonstrations to generate expected outputs for test instances without the need for additional training or gradient updates.In recent years,the popularity of social networking has provided a medium through which some users can engage in offensive and harmful online behavior.In this study,we investigate the ability of different LLMs,ranging from zero-shot and few-shot learning to fine-tuning.Our experiments show that LLMs can identify sexist and hateful online texts using zero-shot and few-shot approaches through information retrieval.Furthermore,it is found that the encoder-decoder model called Zephyr achieves the best results with the fine-tuning approach,scoring 86.811%on the Explainable Detection of Online Sexism(EDOS)test-set and 57.453%on the Multilingual Detection of Hate Speech Against Immigrants and Women in Twitter(HatEval)test-set.Finally,it is confirmed that the evaluated models perform well in hate text detection,as they beat the best result in the HatEval task leaderboard.The error analysis shows that contextual learning had difficulty distinguishing between types of hate speech and figurative language.However,the fine-tuned approach tends to produce many false positives.

DeBERTa-GRU: Sentiment Analysis for Large Language Model

Modern technological advancements have made social media an essential component of daily life.Social media allow individuals to share thoughts,emotions,and ideas.Sentiment analysis plays the function of evaluating whether the sentiment of the text is positive,negative,neutral,or any other personal emotion to understand the sentiment context of the text.Sentiment analysis is essential in business and society because it impacts strategic decision-making.Sentiment analysis involves challenges due to lexical variation,an unlabeled dataset,and text distance correlations.The execution time increases due to the sequential processing of the sequence models.However,the calculation times for the Transformer models are reduced because of the parallel processing.This study uses a hybrid deep learning strategy to combine the strengths of the Transformer and Sequence models while ignoring their limitations.In particular,the proposed model integrates the Decoding-enhanced with Bidirectional Encoder Representations from Transformers(BERT)attention(DeBERTa)and the Gated Recurrent Unit(GRU)for sentiment analysis.Using the Decoding-enhanced BERT technique,the words are mapped into a compact,semantic word embedding space,and the Gated Recurrent Unit model can capture the distance contextual semantics correctly.The proposed hybrid model achieves F1-scores of 97%on the Twitter Large Language Model(LLM)dataset,which is much higher than the performance of new techniques.

Recent Advances on Deep Learning for Sign Language Recognition

Sign language,a visual-gestural language used by the deaf and hard-of-hearing community,plays a crucial role in facilitating communication and promoting inclusivity.Sign language recognition(SLR),the process of automatically recognizing and interpreting sign language gestures,has gained significant attention in recent years due to its potential to bridge the communication gap between the hearing impaired and the hearing world.The emergence and continuous development of deep learning techniques have provided inspiration and momentum for advancing SLR.This paper presents a comprehensive and up-to-date analysis of the advancements,challenges,and opportunities in deep learning-based sign language recognition,focusing on the past five years of research.We explore various aspects of SLR,including sign data acquisition technologies,sign language datasets,evaluation methods,and different types of neural networks.Convolutional Neural Networks(CNN)and Recurrent Neural Networks(RNN)have shown promising results in fingerspelling and isolated sign recognition.However,the continuous nature of sign language poses challenges,leading to the exploration of advanced neural network models such as the Transformer model for continuous sign language recognition(CSLR).Despite significant advancements,several challenges remain in the field of SLR.These challenges include expanding sign language datasets,achieving user independence in recognition systems,exploring different input modalities,effectively fusing features,modeling co-articulation,and improving semantic and syntactic understanding.Additionally,developing lightweight network architectures for mobile applications is crucial for practical implementation.By addressing these challenges,we can further advance the field of deep learning for sign language recognition and improve communication for the hearing-impaired community.

LKPNR: Large Language Models and Knowledge Graph for Personalized News Recommendation Framework

Accurately recommending candidate news to users is a basic challenge of personalized news recommendation systems.Traditional methods are usually difficult to learn and acquire complex semantic information in news texts,resulting in unsatisfactory recommendation results.Besides,these traditional methods are more friendly to active users with rich historical behaviors.However,they can not effectively solve the long tail problem of inactive users.To address these issues,this research presents a novel general framework that combines Large Language Models(LLM)and Knowledge Graphs(KG)into traditional methods.To learn the contextual information of news text,we use LLMs’powerful text understanding ability to generate news representations with rich semantic information,and then,the generated news representations are used to enhance the news encoding in traditional methods.In addition,multi-hops relationship of news entities is mined and the structural information of news is encoded using KG,thus alleviating the challenge of long-tail distribution.Experimental results demonstrate that compared with various traditional models,on evaluation indicators such as AUC,MRR,nDCG@5 and nDCG@10,the framework significantly improves the recommendation performance.The successful integration of LLM and KG in our framework has established a feasible way for achieving more accurate personalized news recommendation.Our code is available at https://github.com/Xuan-ZW/LKPNR.

A large language model-powered literature review for high-angle annular dark field imaging

High-angle annular dark field(HAADF)imaging in scanning transmission electron microscopy(STEM)has become an indispensable tool in materials science due to its ability to offer sub-°A resolution and provide chemical information through Z-contrast.This study leverages large language models(LLMs)to conduct a comprehensive bibliometric analysis of a large amount of HAADF-related literature(more than 41000 papers).By using LLMs,specifically ChatGPT,we were able to extract detailed information on applications,sample preparation methods,instruments used,and study conclusions.The findings highlight the capability of LLMs to provide a new perspective into HAADF imaging,underscoring its increasingly important role in materials science.Moreover,the rich information extracted from these publications can be harnessed to develop AI models that enhance the automation and intelligence of electron microscopes.

Large language models in laparoscopic surgery: A transformative opportunity

This opinion paper explores the transformative potential of large language models(LLMs)in laparoscopic surgery and argues for their integration to enhance surgical education,decision support,reporting,and patient care.LLMs can revolutionize surgical education by providing personalized learning experiences and accelerating skill acquisition.Intelligent decision support systems powered by LLMs can assist surgeons in making complex decisions,optimizing surgical workflows,and improving patient outcomes.Moreover,LLMs can automate surgical reporting and generate personalized patient education materials,streamlining documentation and improving patient engagement.However,challenges such as data scarcity,surgical semantic capture,real-time inference,and integration with existing systems need to be addressed for successful LLM integration.The future of laparoscopic surgery lies in the seamless integration of LLMs,enabling autonomous robotic surgery,predictive surgical planning,intraoperative decision support,virtual surgical assistants,and continuous learning.By harnessing the power of LLMs,laparoscopic surgery can be transformed,empowering surgeons and ultimately benefiting patients.

Evaluating Privacy Leakage and Memorization Attacks on Large Language Models (LLMs) in Generative AI Applications

The recent interest in the deployment of Generative AI applications that use large language models (LLMs) has brought to the forefront significant privacy concerns, notably the leakage of Personally Identifiable Information (PII) and other confidential or protected information that may have been memorized during training, specifically during a fine-tuning or customization process. We describe different black-box attacks from potential adversaries and study their impact on the amount and type of information that may be recovered from commonly used and deployed LLMs. Our research investigates the relationship between PII leakage, memorization, and factors such as model size, architecture, and the nature of attacks employed. The study utilizes two broad categories of attacks: PII leakage-focused attacks (auto-completion and extraction attacks) and memorization-focused attacks (various membership inference attacks). The findings from these investigations are quantified using an array of evaluative metrics, providing a detailed understanding of LLM vulnerabilities and the effectiveness of different attacks.

Identification of Software Bugs by Analyzing Natural Language-Based Requirements Using Optimized Deep Learning Features

Software project outcomes heavily depend on natural language requirements,often causing diverse interpretations and issues like ambiguities and incomplete or faulty requirements.Researchers are exploring machine learning to predict software bugs,but a more precise and general approach is needed.Accurate bug prediction is crucial for software evolution and user training,prompting an investigation into deep and ensemble learning methods.However,these studies are not generalized and efficient when extended to other datasets.Therefore,this paper proposed a hybrid approach combining multiple techniques to explore their effectiveness on bug identification problems.The methods involved feature selection,which is used to reduce the dimensionality and redundancy of features and select only the relevant ones;transfer learning is used to train and test the model on different datasets to analyze how much of the learning is passed to other datasets,and ensemble method is utilized to explore the increase in performance upon combining multiple classifiers in a model.Four National Aeronautics and Space Administration(NASA)and four Promise datasets are used in the study,showing an increase in the model’s performance by providing better Area Under the Receiver Operating Characteristic Curve(AUC-ROC)values when different classifiers were combined.It reveals that using an amalgam of techniques such as those used in this study,feature selection,transfer learning,and ensemble methods prove helpful in optimizing the software bug prediction models and providing high-performing,useful end mode.

Enhancing Relational Triple Extraction in Specific Domains:Semantic Enhancement and Synergy of Large Language Models and Small Pre-Trained Language Models

In the process of constructing domain-specific knowledge graphs,the task of relational triple extraction plays a critical role in transforming unstructured text into structured information.Existing relational triple extraction models facemultiple challenges when processing domain-specific data,including insufficient utilization of semantic interaction information between entities and relations,difficulties in handling challenging samples,and the scarcity of domain-specific datasets.To address these issues,our study introduces three innovative components:Relation semantic enhancement,data augmentation,and a voting strategy,all designed to significantly improve the model’s performance in tackling domain-specific relational triple extraction tasks.We first propose an innovative attention interaction module.This method significantly enhances the semantic interaction capabilities between entities and relations by integrating semantic information fromrelation labels.Second,we propose a voting strategy that effectively combines the strengths of large languagemodels(LLMs)and fine-tuned small pre-trained language models(SLMs)to reevaluate challenging samples,thereby improving the model’s adaptability in specific domains.Additionally,we explore the use of LLMs for data augmentation,aiming to generate domain-specific datasets to alleviate the scarcity of domain data.Experiments conducted on three domain-specific datasets demonstrate that our model outperforms existing comparative models in several aspects,with F1 scores exceeding the State of the Art models by 2%,1.6%,and 0.6%,respectively,validating the effectiveness and generalizability of our approach.

Classification of Conversational Sentences Using an Ensemble Pre-Trained Language Model with the Fine-Tuned Parameter

Sentence classification is the process of categorizing a sentence based on the context of the sentence.Sentence categorization requires more semantic highlights than other tasks,such as dependence parsing,which requires more syntactic elements.Most existing strategies focus on the general semantics of a conversation without involving the context of the sentence,recognizing the progress and comparing impacts.An ensemble pre-trained language model was taken up here to classify the conversation sentences from the conversation corpus.The conversational sentences are classified into four categories:information,question,directive,and commission.These classification label sequences are for analyzing the conversation progress and predicting the pecking order of the conversation.Ensemble of Bidirectional Encoder for Representation of Transformer(BERT),Robustly Optimized BERT pretraining Approach(RoBERTa),Generative Pre-Trained Transformer(GPT),DistilBERT and Generalized Autoregressive Pretraining for Language Understanding(XLNet)models are trained on conversation corpus with hyperparameters.Hyperparameter tuning approach is carried out for better performance on sentence classification.This Ensemble of Pre-trained Language Models with a Hyperparameter Tuning(EPLM-HT)system is trained on an annotated conversation dataset.The proposed approach outperformed compared to the base BERT,GPT,DistilBERT and XLNet transformer models.The proposed ensemble model with the fine-tuned parameters achieved an F1_score of 0.88.

Impact of transcranial electrical stimulation on serum neurotrophic factors and language function in patients with speech disorders

BACKGROUND Speech disorders have a substantial impact on communication abilities and quality of life.Traditional treatments such as speech and psychological therapies frequently demonstrate limited effectiveness and patient compliance.Transcranial electrical stimulation(TES)has emerged as a promising non-invasive treatment to improve neurological functions.However,its effectiveness in enhancing language functions and serum neurofactor levels in individuals with speech disorders requires further investigation.AIM To investigate the impact of TES in conjunction with standard therapies on serum neurotrophic factor levels and language function in patients with speech disorders.METHODS In a controlled study spanning from March 2019 to November 2021,81 patients with speech disorders were divided into a control group(n=40)receiving standard speech stimulation and psychological intervention,and an observation group(n=41)receiving additional TES.The study assessed serum levels of ciliary neurotrophic factor(CNTF),glial cell-derived neurotrophic factor(GDNF),brainderived neurotrophic factor(BDNF),and nerve growth factor(NGF),as well as evaluations of motor function,language function,and development quotient scores.RESULTS After 3 wk of intervention,the observation group exhibited significantly higher serum levels of CNTF,GDNF,BDNF,and NGF compared to the control group.Moreover,improvements were noted in motor function,cognitive function,language skills,physical abilities,and overall development quotient scores.It is worth mentioning that the observation group also displayed superior perfor CONCLUSION This retrospective study concluded that TES combined with traditional speech and psychotherapy can effectively increase the levels of neurokines in the blood and enhance language function in patients with speech disorders.These results provide a promising avenue for integrating TES into standard treatment methods for speech disorders.

A Survey on Chinese Sign Language Recognition:From Traditional Methods to Artificial Intelligence

Research on Chinese Sign Language(CSL)provides convenience and support for individuals with hearing impairments to communicate and integrate into society.This article reviews the relevant literature on Chinese Sign Language Recognition(CSLR)in the past 20 years.Hidden Markov Models(HMM),Support Vector Machines(SVM),and Dynamic Time Warping(DTW)were found to be the most commonly employed technologies among traditional identificationmethods.Benefiting from the rapid development of computer vision and artificial intelligence technology,Convolutional Neural Networks(CNN),3D-CNN,YOLO,Capsule Network(CapsNet)and various deep neural networks have sprung up.Deep Neural Networks(DNNs)and their derived models are integral tomodern artificial intelligence recognitionmethods.In addition,technologies thatwerewidely used in the early days have also been integrated and applied to specific hybrid models and customized identification methods.Sign language data collection includes acquiring data from data gloves,data sensors(such as Kinect,LeapMotion,etc.),and high-definition photography.Meanwhile,facial expression recognition,complex background processing,and 3D sign language recognition have also attracted research interests among scholars.Due to the uniqueness and complexity of Chinese sign language,accuracy,robustness,real-time performance,and user independence are significant challenges for future sign language recognition research.Additionally,suitable datasets and evaluation criteria are also worth pursuing.

Japanese Sign Language Recognition by Combining Joint Skeleton-Based Handcrafted and Pixel-Based Deep Learning Features with Machine Learning Classification

Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japanese Sign Language(JSL)for communication.However,existing JSL recognition systems have faced significant performance limitations due to inherent complexities.In response to these challenges,we present a novel JSL recognition system that employs a strategic fusion approach,combining joint skeleton-based handcrafted features and pixel-based deep learning features.Our system incorporates two distinct streams:the first stream extracts crucial handcrafted features,emphasizing the capture of hand and body movements within JSL gestures.Simultaneously,a deep learning-based transfer learning stream captures hierarchical representations of JSL gestures in the second stream.Then,we concatenated the critical information of the first stream and the hierarchy of the second stream features to produce the multiple levels of the fusion features,aiming to create a comprehensive representation of the JSL gestures.After reducing the dimensionality of the feature,a feature selection approach and a kernel-based support vector machine(SVM)were used for the classification.To assess the effectiveness of our approach,we conducted extensive experiments on our Lab JSL dataset and a publicly available Arabic sign language(ArSL)dataset.Our results unequivocally demonstrate that our fusion approach significantly enhances JSL recognition accuracy and robustness compared to individual feature sets or traditional recognition methods.