The difficulty of accurately summarising Assamese text content is a significant barrier in natural language processing (NLP). Manually summarising lengthy Assamese texts is time-consuming and ...labor-intensive. As a result, automatic text summarization has developed as a critical NLP study topic. In this study, we integrate the Transformer and Self-Attention approaches to develop an abstract text summarization model. This Transformer-based technique uses self-attention approaches to successfully manage co-reference concerns in Assamese text, enhancing overall system understanding. This proposed approach improves the efficiency of text summarization greatly. We exhaustively evaluated the model using the Assamese dataset (AD-50), which contains human-produced summaries, to assess its performance. When compared to current state-of-the-art baseline models, our model outperformed them. On the AD-50 dataset, for example, our suggested model obtained a low training loss of 0.0022 during 20 training epochs, as well as an amazing model accuracy of 47.15 percentage. This research marks a substantial advancement in the field of Assamese abstractive text summarization, with intriguing implications for practical applications in NLP.
Psoriasis is an immune-mediated skin disease affecting approximately 3% of the global population. Proper management of this condition necessitates the assessment of the Body Surface Area (BSA) and ...the involvement of nails and joints. Recently, the integration of Natural Language Processing (NLP) with Electronic Medical Records (EMRs) has shown promise in advancing disease classification and research. This study evaluates the performance of ChatGPT-4, a commercial AI platform, in analyzing unstructured EMR data of psoriasis patients, particularly in identifying affected body areas.
The study analyzed EMR data from 94 patients treated at the Dermatology Department and Psoriasis Outpatient Clinic of Sheba Medical Center between 2008 and 2022. The data were processed using the ChatGPT-4 interface to identify and report the body areas affected by psoriasis. These identified areas were then categorized, and the accuracy of ChatGPT-4′s analysis was compared with that of a senior dermatologist.
The results revealed that the dermatologist identified 477 psoriasis-affected body areas. ChatGPT-4 accurately recognized 443 (92.8%) of these areas, missed 34, and incorrectly identified 30 areas as affected. From 94 cases, nail involvement was detected in 32 cases (34.0%), with ChatGPT-4 correctly identifying 29 cases. Joint involvement was noted in 25 cases (26.6%), with 24 correctly identified by ChatGPT-4. Complete accuracy was achieved in 54 cases (57.4%), while inaccuracies were observed in 40 cases (42.6%). We found that cases with more characters, words, or identified body areas were more prone to errors, suggesting that increased data complexity heightens the likelihood of inaccuracies in AI analysis.
ChatGPT-4 demonstrated a high performance in analyzing detailed and complex unstructured EMR data from psoriasis patients, effectively identifying involved body areas, including nails and joints. This highlights the potential of NLP algorithms in enhancing the analysis of unstructured EMR data for both clinical follow-up and research purposes.
Despite the increasing use of natural language processing (NLP) in the construction domain, no systematic comparison has been conducted between NLP applications in construction and the latest ...advancements in NLP within the computer science domain. Therefore, this study compares NLP studies in these two domains. Firstly, a bibliometric analysis was performed on 55 publications in state-of-the-art NLP studies, which identified four main research areas in NLP. Secondly, a systematic review of 202 NLP studies in construction was conducted, presenting representative application areas of NLP and their current technical status. The results reveal a decreasing technology gap between NLP in construction and the state-of-the-art. However, the comparison also highlighted gaps in application areas and methodologies, and eight future research opportunities were proposed. This review serves as a foundation for future studies that aim to apply state-of-the-art NLP technologies in the construction domain.
•Systematic comparison was conducted between NLP studies in construction and state-of-the-art.•State-of-the-art NLP methods and the major field of applications (i.e., NLP tasks) were reviewed.•PRISMA and bibliometic analysis were used for the review study.•Technology gaps between NLP in construction and state-of-the-art were presented.•Future research opportunities were suggested to fill the identified technology gaps.
The ever-evolving global landscape of communication, driven by Information Technology advancements, underscores the importance of emotion detection in natural language processing. However, challenges ...persist in interpreting emotions within linguistically diverse contexts, notably in low-resource languages like Bengali, compounded by the emergence of Banglish. To address this gap, we present “Bengali & Banglish,” an extensive dataset comprising 80,098 labelled samples across six emotion classes. Our dataset fills a void in fine-grained emotion classification for Bengali and pioneers in emotion detection in Banglish. We achieve significant performance metrics through meticulous annotation and rigorous evaluation, including a weighted F1 score of 71.30% for Bengali and 64.59% for Banglish using BanglaBERT. Also, our dataset facilitates Bengali-to-Banglish Machine Translation, contributing to the advancement of language processing models. Furthermore, our dataset demonstrates a high Cohen's Kappa score of 93.5%, affirming the reliability and consistency of our annotations. This research underscores the importance of linguistic diversity in NLP and provides a valuable resource for enhancing Emotion Detection capabilities in Bengali and Banglish across digital platforms.
Knowing that language is often used as a classifier of human intelligence and the development of systems that understand human language remains a challenge all the time (Kryeziu & Shehu, 2022). ...Natural Language Processing is a very active field of study, where transformers have a key role. Transformers function based on neural networks and they are increasingly showing promising results. One of the first major contributions to transfer learning in Natural Language Processing was the use of pre-trained word embeddings in 2010 (Joseph, Lev, & Yoshua, 2010). Pre-trained models like ELMo (Matthew, et al., 2018) and BERT (Delvin, et al., 2019) are trained on large corpora of unlabeled text and as a result learning from text representations has achieved good performance on many of the underlying tasks on datasets from different domains. Pre-training in the language model has proven that there has been an improvement in some aspects of natural language processing, based on the paper (Dai & Le, 2015). In present paper, we will pre-train BERT on the task of Masked Language Modeling (MLM) with the Albanian language dataset (alb_dataset) that we have created for this purpose (Kryeziu et al., 2022). We will compare two approaches: training of BERT using the available OSCAR dataset and using our alb_dataset that we have collected. The paper shows some discrepancies during training, especially while evaluating the performance of the model.
Research has shown that accounting for moral sentiment in natural language can yield insight into a variety of on- and off-line phenomena such as message diffusion, protest dynamics, and social ...distancing. However, measuring moral sentiment in natural language is challenging, and the difficulty of this task is exacerbated by the limited availability of annotated data. To address this issue, we introduce the Moral Foundations Twitter Corpus, a collection of 35,108 tweets that have been curated from seven distinct domains of discourse and hand annotated by at least three trained annotators for 10 categories of moral sentiment. To facilitate investigations of annotator response dynamics, we also provide psychological and demographic metadata for each annotator. Finally, we report moral sentiment classification baselines for this corpus using a range of popular methodologies.
Reinforcement learning (RL) has proven its worth in a series of artificial domains, and is beginning to show some successes in real-world scenarios. However, much of the research advances in RL are ...hard to leverage in real-world systems due to a series of assumptions that are rarely satisfied in practice. In this work, we identify and formalize a series of independent challenges that embody the difficulties that must be addressed for RL to be commonly deployed in real-world systems. For each challenge, we define it formally in the context of a Markov Decision Process, analyze the effects of the challenge on state-of-the-art learning algorithms, and present some existing attempts at tackling it. We believe that an approach that addresses our set of proposed challenges would be readily deployable in a large number of real world problems. Our proposed challenges are implemented in a suite of continuous control environments called realworldrl-suite which we propose an as an open-source benchmark.
We study the problem of training deep fully connected neural networks with Rectified Linear Unit (ReLU) activation function and cross entropy loss function for binary classification using gradient ...descent. We show that with proper random weight initialization, gradient descent can find the global minima of the training loss for an over-parameterized deep ReLU network, under certain assumption on the training data. The key idea of our proof is that Gaussian random initialization followed by gradient descent produces a sequence of iterates that stay inside a small perturbation region centered at the initial weights, in which the training loss function of the deep ReLU networks enjoys nice local curvature properties that ensure the global convergence of gradient descent. At the core of our proof technique is (1) a milder assumption on the training data; (2) a sharp analysis of the trajectory length for gradient descent; and (3) a finer characterization of the size of the perturbation region. Compared with the concurrent work (Allen-Zhu et al. in A convergence theory for deep learning via over-parameterization,
2018a
; Du et al. in Gradient descent finds global minima of deep neural networks,
2018a
) along this line, our result relies on milder over-parameterization condition on the neural network width, and enjoys faster global convergence rate of gradient descent for training deep neural networks.
Over the past two decades, support vector machines (SVMs) have become a popular supervised machine learning model, and plenty of distinct algorithms are designed separately based on different KKT ...conditions of the SVM model for classification/regression with different losses, including convex and or nonconvex loss. In this paper, we propose an algorithm that can train different SVM models in a
unified
scheme. First, we introduce a definition of the least squares type of difference of convex loss (LS-DC) and show that the most commonly used losses in the SVM community are LS-DC loss or can be approximated by LS-DC loss. Based on the difference of convex algorithm (DCA), we then propose a unified algorithm called
UniSVM
which can solve the SVM model with any convex or nonconvex LS-DC loss, wherein only a vector is computed by the specifically chosen loss. UniSVM has a dominant advantage over all existing algorithms for training robust SVM models with nonconvex losses because it has a closed-form solution per iteration, while the existing algorithms always need to solve an L1SVM/L2SVM per iteration. Furthermore, by the low-rank approximation of the kernel matrix, UniSVM can solve large-scale nonlinear problems efficiently. To verify the efficacy and feasibility of the proposed algorithm, we perform many experiments on small artificial problems and large benchmark tasks both with and without outliers for classification and regression for comparison with state-of-the-art algorithms. The experimental results demonstrate that UniSVM can achieve comparable performance in less training time. The foremost advantage of UniSVM is that its core code in Matlab is less than 10 lines; hence, it can be easily grasped by users or researchers.
Chat Generative Pre-trained Transformer (ChatGPT) is a 175-billion-parameter natural language processing model that can generate conversation-style responses to user input.
This study aimed to ...evaluate the performance of ChatGPT on questions within the scope of the United States Medical Licensing Examination (USMLE) Step 1 and Step 2 exams, as well as to analyze responses for user interpretability.
We used 2 sets of multiple-choice questions to evaluate ChatGPT's performance, each with questions pertaining to Step 1 and Step 2. The first set was derived from AMBOSS, a commonly used question bank for medical students, which also provides statistics on question difficulty and the performance on an exam relative to the user base. The second set was the National Board of Medical Examiners (NBME) free 120 questions. ChatGPT's performance was compared to 2 other large language models, GPT-3 and InstructGPT. The text output of each ChatGPT response was evaluated across 3 qualitative metrics: logical justification of the answer selected, presence of information internal to the question, and presence of information external to the question.
Of the 4 data sets, AMBOSS-Step1, AMBOSS-Step2, NBME-Free-Step1, and NBME-Free-Step2, ChatGPT achieved accuracies of 44% (44/100), 42% (42/100), 64.4% (56/87), and 57.8% (59/102), respectively. ChatGPT outperformed InstructGPT by 8.15% on average across all data sets, and GPT-3 performed similarly to random chance. The model demonstrated a significant decrease in performance as question difficulty increased (P=.01) within the AMBOSS-Step1 data set. We found that logical justification for ChatGPT's answer selection was present in 100% of outputs of the NBME data sets. Internal information to the question was present in 96.8% (183/189) of all questions. The presence of information external to the question was 44.5% and 27% lower for incorrect answers relative to correct answers on the NBME-Free-Step1 (P<.001) and NBME-Free-Step2 (P=.001) data sets, respectively.
ChatGPT marks a significant improvement in natural language processing models on the tasks of medical question answering. By performing at a greater than 60% threshold on the NBME-Free-Step-1 data set, we show that the model achieves the equivalent of a passing score for a third-year medical student. Additionally, we highlight ChatGPT's capacity to provide logic and informational context across the majority of answers. These facts taken together make a compelling case for the potential applications of ChatGPT as an interactive medical education tool to support learning.