Natural Language Processing Journal最新文献

{"title":"Unsupervised hypernymy directionality prediction using context terms","authors":"Thushara Manjari Naduvilakandy,&nbsp;Hyeju Jang,&nbsp;Mohammad Al Hasan","doi":"10.1016/j.nlp.2024.100118","DOIUrl":"10.1016/j.nlp.2024.100118","url":null,"abstract":"<div><div>Hypernymy directionality prediction is an important task in Natural Language Processing (NLP) due to its significant usages in natural language understanding and generation. Many supervised and unsupervised methods have been proposed for this task. Supervised methods require labeled examples, which are not readily available for many domains; besides, supervised models for this task that are trained on data from one domain performs poorly on data in a different domain. Therefore, unsupervised methods that are universally applicable for all domains are preferred. Existing unsupervised methods for hypernymy directionality prediction are outdated and suffer from poor performance. Specifically, they do not leverage distributional pre-trained vectors from neural language models, which have shown to be very effective in diverse NLP tasks. In this paper, we present DECIDE, a simple yet effective unsupervised method for hypernymy directionality prediction that exploits neural pre-trained vectors of words in context. By utilizing the distributional informativeness hypothesis over the context vectors, DECIDE predicts the hypernym directionality between a pair of words with a high accuracy. Extensive experiments on seven datasets demonstrate that DECIDE outperforms or achieves comparable performance to existing unsupervised and supervised methods.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100118"},"PeriodicalIF":0.0,"publicationDate":"2024-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142660100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning-based natural language processing in human–agent interaction: Applications, advancements and challenges","authors":"Nafiz Ahmed ,&nbsp;Anik Kumar Saha ,&nbsp;Md. Abdullah Al Noman ,&nbsp;Jamin Rahman Jim ,&nbsp;M.F. Mridha ,&nbsp;Md Mohsin Kabir","doi":"10.1016/j.nlp.2024.100112","DOIUrl":"10.1016/j.nlp.2024.100112","url":null,"abstract":"<div><div>Human–Agent Interaction is at the forefront of rapid development, with integrating deep learning techniques into natural language processing representing significant potential. This research addresses the complicated dynamics of Human–Agent Interaction and highlights the central role of Deep Learning in shaping the communication between humans and agents. In contrast to a narrow focus on sentiment analysis, this study encompasses various Human–Agent Interaction facets, including dialogue systems, language understanding and contextual communication. This study systematically examines applications, algorithms and models that define the current landscape of deep learning-based natural language processing in Human–Agent Interaction. It also presents common pre-processing techniques, datasets and customized evaluation metrics. Insights into the benefits and challenges of machine learning and Deep Learning algorithms in Human–Agent Interaction are provided, complemented by a comprehensive overview of the current state-of-the-art. The manuscript concludes with a comprehensive discussion of specific Human–Agent Interaction challenges and suggests thoughtful research directions. This study aims to provide a balanced understanding of models, applications, challenges and research directions in deep learning-based natural language processing in Human–Agent Interaction, focusing on recent contributions to the field.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100112"},"PeriodicalIF":0.0,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Persian readability classification using DeepWalk and tree-based ensemble methods","authors":"Mohammad Mahmoodi Varnamkhasti","doi":"10.1016/j.nlp.2024.100116","DOIUrl":"10.1016/j.nlp.2024.100116","url":null,"abstract":"<div><div>The Readability Classification (Difficulty classification) problem is the task of assessing the readability of text by categorizing it into different levels or classes based on its difficulty to understand. Applications ranging from language learning tools to website content optimization depend on readability classification. While numerous techniques have been proposed for readability classification in various languages, the topic has received little attention in the Persian (Farsi) language. Persian readability analysis poses unique challenges due to its complex morphology and flexible syntax, which necessitate a customized approach for accurate classification. In this research, we have proposed a method based on the nodes graph embedding and tree-based classification methods for sentence-level readability classification in the Persian language. The results indicate an F1-score of up to 0.961 in predicting the readability of Persian sentences.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100116"},"PeriodicalIF":0.0,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554497","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hawk: An industrial-strength multi-label document classifier","authors":"Arshad Javeed","doi":"10.1016/j.nlp.2024.100115","DOIUrl":"10.1016/j.nlp.2024.100115","url":null,"abstract":"<div><div>There are a plethora of methods for solving the classical multi-label document classification problem. However, when it comes to deployment and usage in an industry setting, most if not all the contemporary approaches fail to address some of the vital aspects or requirements of an ideal solution: i) ability to operate on variable-length texts or rambling documents, ii) catastrophic forgetting problem, and iii) ability to visualize the model’s predictions. The paper describes the significance of these problems in detail and adopts the hydranet architecture to address these problems. The proposed architecture views documents as a sequence of sentences and leverages sentence-level embeddings for input representation, turning the problem into a sequence classification task. Furthermore, two specific architectures are explored as the architectures for the heads, Bi-LSTM and transformer heads. The proposed architecture is benchmarked on some of the popular benchmarking datasets such as Web of Science - 5763, Web of Science - 11967, BBC Sports, and BBC News datasets. The experimental results reveal that the proposed model performs at least as best as previous SOTA architectures and even outperforms prior SOTA in a few cases, along with the added advantages of the practicality issues discussed. The ablation study includes comparisons of the impact of the attention mechanism and the application of weighted loss functions to train the task-specific heads in the hydranet. The claims regarding catastrophic forgetfulness are further corroborated by empirical evaluations under incremental learning scenarios. The results reveal the robustness of the proposed architecture compared to other benchmarks.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100115"},"PeriodicalIF":0.0,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142577899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Validating pretrained language models for content quality classification with semantic-preserving metamorphic relations","authors":"Pak Yuen Patrick Chan,&nbsp;Jacky Keung","doi":"10.1016/j.nlp.2024.100114","DOIUrl":"10.1016/j.nlp.2024.100114","url":null,"abstract":"<div><h3>Context:</h3><div>Utilizing pretrained language models (PLMs) has become common practice in maintaining the content quality of question-answering (Q&amp;A) websites. However, evaluating the effectiveness of PLMs poses a challenge as they tend to provide local optima rather than global optima.</div></div><div><h3>Objective:</h3><div>In this study, we propose using semantic-preserving Metamorphic Relations (MRs) derived from Metamorphic Testing (MT) to address this challenge and validate PLMs.</div></div><div><h3>Methods:</h3><div>To validate four selected PLMs, we conducted an empirical experiment using a publicly available dataset comprising 60000 data points. We defined three groups of Metamorphic Relations (MRGs), consisting of thirteen semantic-preserving MRs, which were then employed to generate “Follow-up” testing datasets based on the original “Source” testing datasets. The PLMs were trained using a separate training dataset. A comparison was made between the predictions of the four trained PLMs for “Source” and “Follow-up” testing datasets in order to identify instances of violations, which corresponded to inconsistent predictions between the two datasets. If no violation was found, it indicated that the PLM was insensitive to the associate MR; thereby, the MR can be used for validation. In cases where no violation occurred across the entire MRG, non-violation regions were identified and supported simulation metamorphic testing.</div></div><div><h3>Results:</h3><div>The results of this study demonstrated that the proposed MRs could effectively serve as a validation tool for content quality classification on Stack Overflow Q&amp;A using PLMs. One PLM did not violate the “Uppercase conversion” MRG and the “Duplication” MRG. Furthermore, the absence of violations in the MRGs allowed for the identification of non-violation regions, confirming the ability of the proposed MRs to support simulation metamorphic testing.</div></div><div><h3>Conclusion:</h3><div>The experimental findings indicate that the proposed MRs can validate PLMs effectively and support simulation metamorphic testing for PLMs. However, further investigations are required to enhance the semantic comprehension and common sense knowledge of PLMs and explore highly informative statistical patterns of PLMs, in order to improve their overall performance.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100114"},"PeriodicalIF":0.0,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unveiling personality traits through Bangla speech using Morlet wavelet transformation and BiG","authors":"Md. Sajeebul Islam Sk.,&nbsp;Md. Golam Rabiul Alam","doi":"10.1016/j.nlp.2024.100113","DOIUrl":"10.1016/j.nlp.2024.100113","url":null,"abstract":"<div><div>Speech serves as a potent medium for expressing a wide array of psychologically significant attributes. While earlier research on deducing personality traits from user-generated speech predominantly focused on other languages, there is a noticeable absence of prior studies and datasets for automatically assessing user personalities from Bangla speech. In this paper, our objective is to bridge the research gap by generating speech samples, each imbued with distinct personality profiles. These personality impressions are subsequently linked to OCEAN (Openness, Conscientiousness, Extroversion, Agreeableness, and Neuroticism) personality traits. To gauge accuracy, human evaluators, unaware of the speaker’s identity, assess these five personality factors. The dataset is predominantly composed of around 90% content sourced from online Bangla newspapers, with the remaining 10% originating from renowned Bangla novels. We perform feature level fusion by combining MFCCs with LPC features to set MELP and MEWLP features. We introduce MoMF feature extraction method by transforming Morlet wavelet and fusing MFCCs feature. We develop two soft voting ensemble models, DistilRo (based on DistilBERT and RoBERTa) and BiG (based on Bi-LSTM and GRU), for personality classification in speech-to-text and speech modalities, respectively. The DistilRo model has gained F-1 score 89% in speech-to-text and the BiG model has gained F-1 score 90% in speech modality.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100113"},"PeriodicalIF":0.0,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"V-LTCS: Backbone exploration for Multimodal Misogynous Meme detection","authors":"Sneha Chinivar ,&nbsp;Roopa M.S. ,&nbsp;Arunalatha J.S. ,&nbsp;Venugopal K.R.","doi":"10.1016/j.nlp.2024.100109","DOIUrl":"10.1016/j.nlp.2024.100109","url":null,"abstract":"<div><div>Memes have become a fundamental part of online communication and humour, reflecting and shaping the culture of today’s digital age. The amplified Meme culture is inadvertently endorsing and propagating casual Misogyny. This study proposes V-LTCS (Vision- Language Transformer Combination Search), a framework that encompasses all possible combinations of the most fitting Text (<em>i.e.</em> BERT, ALBERT, and XLM-R) and Vision (<em>i.e.</em> Swin, ConvNeXt, and ViT) Transformer Models to determine the backbone architecture for identifying Memes that contains misogynistic contents. All feasible Vision-Language Transformer Model combinations obtained from the recognized optimal Text and Vision Transformer Models are evaluated on two (smaller and larger) datasets using varied standard metrics (<em>viz.</em> Accuracy, Precision, Recall, and F1-Score). The BERT-ViT combinational Transformer Model demonstrated its efficiency on both datasets, validating its ability to serve as a backbone architecture for all subsequent efforts to recognize Multimodal Misogynous Memes.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100109"},"PeriodicalIF":0.0,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Recent advancements in automatic disordered speech recognition: A survey paper","authors":"Nada Gohider,&nbsp;Otman A. Basir","doi":"10.1016/j.nlp.2024.100110","DOIUrl":"10.1016/j.nlp.2024.100110","url":null,"abstract":"<div><div>Automatic Speech Recognition (ASR) technology has recently witnessed a paradigm shift with respect to performance accuracy. Nevertheless, impaired speech remains a significant challenge, evidenced by the inadequate accuracy of existing ASR solutions. This lacking is reported in various research reports. While this lacking has motivated new directions in <em>Automatic Disordered Speech Recognition</em> (ADSR), the gap between ASR performance accuracy and that of ADSR remains significant. In this paper, we report a consolidated account of research work conducted to date to address this gap, highlighting the root causes of such performance discrepancy and discussing prominent research directions in this area. The paper raises some fundamental issues and challenges that ADSR research faces today. Firstly, we discuss the adequacy of impaired speech representation in existing datasets, in terms of the diversity of speech impairments, speech continuity, speech style, vocabulary, age group, and the environments of the data collection process. We argue that disordered speech is poorly represented in the existing datasets; thus, it is expected that several fundamental components needed for training ADSR models are absent. Most of the open-access databases of impaired speech focus on adult dysarthric speakers, ignoring a wide spectrum of speech disorders and age groups. Furthermore, the paper reviews prominent research directions adopted by the ADSR research community in its effort to advance speech recognition technology for impaired speakers. We categorize this research effort into directions such as personalized models, model adaptation, data augmentation, and multi-modal learning. Although these research directions have advanced the performance of ADSR models, we believe there is still potential for further advancement since current efforts, in essence, make the false assumption that there is a limited distribution shift between the source and target data. Finally, we stress the need to investigate performance measures other than Word Error Rate (WER)- measures that can reliably encode the contribution of erroneous output tokens in the final uttered message.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100110"},"PeriodicalIF":0.0,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142419914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Recurrent neural network based multiclass cyber bullying classification","authors":"Silvia Sifath ,&nbsp;Tania Islam ,&nbsp;Md Erfan ,&nbsp;Samrat Kumar Dey ,&nbsp;MD. Minhaj Ul Islam ,&nbsp;Md Samsuddoha ,&nbsp;Tazizur Rahman","doi":"10.1016/j.nlp.2024.100111","DOIUrl":"10.1016/j.nlp.2024.100111","url":null,"abstract":"<div><div>Cyberbullying is one of the crimes that arise rapidly through the daily use of technology by different types of people and, most notably, by sharing one’s opinions or feelings on social media in a harmful manner. It has several negative effects on society such as depression, anxiety, suicide, and so on. At the same time, it reduces productivity, causes psychological damage that can last a lifetime and increases violence among people. To prevent cyberbullying or take necessary steps against the harasser, the first step is to detect cyberbullying. Several works exist to detect and classify cyberbullying but a few works have been carried out to classify cyberbullying in the Bengali Language. As the number of people is increased day by day who communicate on social media using the Bengali language, it is crucial to address this situation and improve both accuracy and robustness to detect and classify cyberbullying. For this purpose, we propose an NLP-based model using machine learning and deep learning algorithms to detect and classify Bengali comments on social media. This research specifies cyberbullying comments using a multiclass classification strategy. Kaggle and Melany are used to collect the dataset to train and evaluate our model. The dataset contains 56308 Bengali comments, consisting of four distinct categories. The categories are not bully, trolls, sexual, and threats. We use different machine learning algorithms such as Support Vector Machine, Logistic Regression, Random Forest, XGBOOST, Multinomial Naïve Bayes, Deep learning algorithm, Recurrent Neural Network (RNN), and two fusion models. Along with that effective preprocessing steps are implemented to get a suitable dataset. In this study, the Recurrent Neural Network gives the best accuracy, which is 86%. The accuracy of our model is good enough to help social media users and encourage them to practice morality.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100111"},"PeriodicalIF":0.0,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of deep neural network architectures for authorship obfuscation of Portuguese texts","authors":"Antônio Marcos Rodrigues Franco,&nbsp;Ítalo Cunha,&nbsp;Leonardo B. Oliveira","doi":"10.1016/j.nlp.2024.100107","DOIUrl":"10.1016/j.nlp.2024.100107","url":null,"abstract":"<div><div>Preserving authorship anonymity is paramount to protect activists, freedom of expression, and critical journalism. Although there are several mechanisms to provide anonymity on the Internet, one can still identify anonymous authors through their writing style. With the advances in neural network and natural language processing research, the success of a classifier when identifying the author of a text is growing. On the other hand, new approaches that use recurrent neural networks for automatic generation of obfuscated texts have also arisen to fight anonymity adversaries. In this work, we evaluate two approaches that use neural networks to generate obfuscated texts. The first approach uses Generative Adversarial Networks to train an encoder–decoder to transform sentences from an input style into a target style. The second one trains an auto encoder with Gradient Reversal Layer to learn invariant representations. In our experiments, we compared the efficiency of both techniques when removing the stylistic attributes of a text and preserving its original semantics. Our evaluation on real texts clarifies each technique’s trade-offs for Portuguese texts and provides guidance on practical deployment.</div></div>","PeriodicalId":100944,"journal":{"name":"Natural Language Processing Journal","volume":"9 ","pages":"Article 100107"},"PeriodicalIF":0.0,"publicationDate":"2024-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142323946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}