Text Segmentation Research Articles

Improved bidirectional attention flow (BIDAF) model for Arabic machine reading comprehension

Abstract Machine reading comprehension (MRC) refers to the process of instructing machines to comprehend and respond to inquiries based on a provided text. There are two primary methodologies for achieving this: extracting answers directly from the text or predicting them. Extracting answers involves anticipating the specific segment of text containing the answer, pinpointed by its starting and ending indices within the paragraph. Despite the increasing interest in MRC, exploration within the framework of the Arabic language faces limitations due to various challenges. A significant impediment arises from the inadequacy of resources available for Arabic textual content, which impedes the development of effective models. Furthermore, the inherent intricacies of Arabic, manifesting in its diverse linguistic forms including classical, modern standard, and colloquial, present distinctive hurdles for tasks involving language comprehension. This paper proposes an enhanced version of the bidirectional attention flow (BIDAF) model for Arabic MRC, constructed upon the Arabic Span-Extraction-based Reading Comprehension Benchmark (ASER). ASER comprises 10,000 sets of questions, answers, and passages, partitioned into a training set constituting 90% of the data and a testing set making up the remaining 10%. By introducing a new input feature based on parts-of-speech (POS) word embeddings and replacing Bidirectional Long Short-Term Memory (bi-LSTM) with bidirectional gated recurrent unit, significant improvements were observed. Eight different POS word embeddings were generated using both Continuous Bag of Words (CBOW) and Skip-gram methods, with varying dimensionalities. Evaluation metrics, including exact match (EM) and F1-measure, were utilized to assess model performance, with emphasis on the latter for its accuracy. The proposed enhanced BIDAF model achieved a remarkable accuracy of 75.22% on the ASER dataset, demonstrating its efficacy in Arabic MRC tasks. Additionally, rigorous statistical evaluation using a two-tailed paired samples t-test further validated the findings, highlighting the significance of the proposed enhancements in advancing Arabic language processing capabilities.

Restoration and Segmentation of Old Jawi Manuscripts using Variational Image Inpainting and Active Contour Models

Old Jawi Manuscripts (OJM) are crucial to historical studies, offering insights into past societies. However, degradation from mishandling and environmental factors can impair their legibility. To preserve OJM, image inpainting and segmentation are essential for restoring corrupted areas and identifying text. Recently, the Gaussian Regularization Segmentation (GRS) model has shown effectiveness in intensity inhomogeneity grayscale image segmentation, though it was not designed for corrupted OJM images. Therefore, this study aimed to reformulate the GRS model to restore and segment text from real corrupted OJM images. The methodology begins with the incorporation of the Mumford-Shah and Bertalmio inpainting models into the GRS model as new fitting terms, resulting in the Modified Gaussian Regularization Segmentation Mumford-Shah (MGRSM) model and the Modified Gaussian Regularization Segmentation Bertalmio (MGRSB) model, respectively. MATLAB was used to implement these models, and their performance was assessed on 30 corrupted OJM samples from Malay Ethnomathematics Research Group, with expert evaluations and efficiency measured by average elapsed time. The MGRSM model achieved 38.4 percent and 12.4 percent higher overall total scores from experts in terms of segmentation accuracy compared to the GRS and MGRSB models, respectively. While the GRS model is the fastest, the MGRSM model provides superior accuracy, with an average processing time of 9.35 seconds, making it the most optimal for restoring and segmenting OJM images. This approach not only enhances the preservation of historical manuscripts but also provides a practical tool for researchers and historians in safeguarding our cultural heritage.

Application of large language models to intelligently analyze long construction contract texts

The emergence of large language models (LLMs) has provided an opportunity to assist humans in quickly reading, searching, and understanding the contents of construction contracts. However, the limited context length of LLMs restricts their ability to process lengthy contract documents, which hinders their application in the construction industry. This study proposes intelligent analysis methods for long construction contracts, which enables LLMs to handle contracts that exceed their context length through a two-stage text segmentation. We further utilized the segmented text chunks for content compression and intelligent question-answering applications. The FIDIC contract was used for testing. A condensed version of lengthy contracts summarizes the provisions in a shorter format. It maintains a high level of correctness and readability, offering practitioners additional options to read contracts of varying lengths based on their needs. The satisfaction rate of the question-answering outcomes reached 93.3%, allowing practitioners to quickly obtain specific clauses of interest and relevant contract knowledge through personalized queries. The performance of the 8K model using our methods is comparable to that of the 128K long-text models, while reducing computational power. This study expands the potential for applications of LLMs in contract management within the construction industry.

Functional Analysis of Thematic Approach to the Division of Surahs in Two English Contemporary Translations of the Qur’an

Thematic division of Qur’anic chapters (henceforth, surahs) into paragraphs is a recent trend in translating the Qur’an. This thematic approach refers to the translation of Qur’anic surah based on their underlying themes rather than verse-by-verse translation in a bilingual text. Disregarding source text division, some translators divide each Qur’anic surah into paragraphs with subheadings while translating the surahs into English. This paper aimed to reveal how the division and chosen themes for each paragraph of the surahs often differ among translators who choose this thematic approach to the Qur’an translation. The data is collected from two modern thematic English translations of the Qur’an, namely, The Clear Quran by Khattab (2015) and The Majestic Quran by Hussain (2018). Adopting a descriptive design and qualitative research method, the paper aimed at applying Nord’s model of functionalist text analysis to compare the units of analysis selected from three Qur’anic surahs: Q. 86, Q. 87 and Q. 93. These surahs are particularly chosen because they are divided differently in the two translations. The paper queries the impact of thematic translation and text segmentation on the Qur’an translation process, especially in relation to producing loyalty and functional adequacy in the two translations. Nord’s functional model was used to explore the effect of thematic translation in conveying the functions for the intended readership without betraying their trust in that they are reading God’s word in their own language. The findings emphasise that thematic translation tends to violate rather than maintain the original intended functions (skopos) and loyalty to the source text.

Journalism Versus Churnalism: How News Factors in Press Releases Affect Journalistic Processing of Ocean Plastic Research in Newspapers Globally

ABSTRACT This study evaluates the impact of news factors in press releases sent from scientific organizations on the representation of ocean plastic research in newspapers. We tracked 84 press releases discussing peer-reviewed ocean plastic studies, published between 2017 and 2021 on EurekAlert! and identified their presence in 495 English-language newspapers worldwide. Text comparison showed that press releases contribute highly to the construction of newspaper articles, with copied text segments in over half of the articles. News factors highlighting the prestige of the scientific study, like research published in high-impact journals or conducted by elite researchers were indicators of high newspaper coverage, as were the news factors bad news, magnitude, and social/economic relevance. Scientific press releases usually provide essential details about research, such as sample characteristics, research sites, and methods. However, funding information and research limitations were only marginally present. This study showed that providing research details does not reduce newsworthiness, instead sometimes even increasing it. Providing research details in press releases can influence journalists’ understanding and evaluation of the scientific study and thereby their decision to report. Hence, to assess the newsworthiness of scientific research, we propose that news factor analysis should include research details.

Enhancing Aortic Aneurysm Surveillance: Transformer NLP for Flagging and Measuring in Radiology Reports

Incidental findings of aortic aneurysms (AAs) often go unreported, and established patients are frequently lost to follow-up. Natural language processing (NLP) offers a promising solution to address these issues. While rule-based NLP methods have shown some success, recent advancements in transformer-based large language models (LLMs) remain underutilized. This study has three aims: (1) to evaluate the effectiveness of our innovative transformer-based NLP pipeline regarding AA detection; (2) to detail the clinical impact by quantifying the number of patients who could benefit from such technology; and (3) to use this information to help coordinate appointments with patients, ensuring proper monitoring and management. 3229 radiology reports were divided into three batches with varying class balance. Each entry was processed through our innovative NLP pipeline, where it was fragmented using regular expression (regex) functions to isolate relevant textual segments. These segments were subsequently processed through our "question and find" (Q&F) function, powered by Google's BERT, a well-established transformer LLM. This Q&F function extracted aortic diameter measurements, flagging measurements that exceeded a predefined threshold. Following detection, we conducted comprehensive chart reviews and contacted primary care providers (PCPs) and patients to categorize aneurysms as "known" or "incidental." We also assessed whether patients with known aneurysms were adhering to regular yearly screenings and coordinated follow-up appointments. Evaluation of the three batches showed high F1 scores: 99.4% (95% CI [98.5-100]), 96.7% (95% CI [95.0-98.2]), and 98.9% (95% CI [98.0-99.6]). Overall measurement accuracy was 98.9% (95% CI [97.6-100]), 99.6% (95% CI [99.3-99.9]), and 98.1% (95% CI [96.8-99.4]). Compared to manual chart reviews, the NLP system demonstrated superior accuracy and fewer errors: 12 vs. 22 (p=0.084), 47 vs. 98 (p=0.000021), and 31 vs. 53 (p=0.015). Of the 412 patients investigated, 58 (14.1%) involved incidental findings, 54 patients (15.3%) were lost to follow-up, 39 patients (55.7%) were successfully contacted, and 37 follow-up appointments (12.1%) were successfully coordinated. The high-performance metrics from our study demonstrate that transformer-based NLP can enhance aortic aneurysm surveillance. Our subsequent comprehensive patient profiling highlighted the need for such a system as a safety net within the electronic medical record (EMR), systematically reviewing radiology reports to detect incidental findings and patients lost to follow-up. This ensures appropriate referrals and monitoring, improving patient outcomes and healthcare efficiency through timely clinical interventions.

Deep handwritten diagram segmentation

AbstractHandwriting is a natural way to communicate and exchange ideas, but converting handwritten diagrams to application‐specific digital formats requires skill and time. Automatic handwritten document conversion can save time, but diagrams and text require different recognition engines. Since accurate segmentation of handwritten diagrams can improve the accuracy of later diagram recognition steps, the authors propose to solve the problem of segmentation of text and non‐text elements of handwritten diagrams using deep semantic segmentation. The model, DeepDP is a flexible U‐net style architecture that can be tuned in complexity to a level appropriate for a particular dataset and diagram type. Experiments on a public hand‐drawn flowchart dataset and a business process diagram dataset show excellent performance, with a per pixel accuracy of 98.6% on the public flowchart datasets and improvement over the 99.3% text stroke accuracy and 96.6% non‐text stroke accuracy obtained by state of the art methods that use online stroke information. On the smaller offline business process diagram dataset, the method obtains a per‐pixel accuracy of 96.9%.

Quantifying decision support level of explainable automatic classification of diagnoses in Spanish medical records

Background and Objective:In the realm of automatic Electronic Health Records (EHR) classification according to the International Classification of Diseases (ICD) there is a notable gap of non-black box approaches and more in Spanish, which is also frequently ignored in clinical language classification. An additional gap in explainability pertains to the lack of standardized metrics for evaluating the degree of explainability offered by distinct techniques. Methods:We address the classification of Spanish electronic health records, using methods to explain the predictions and improve the decision support level. We also propose Leberage a novel metric to quantify the decision support level of the explainable predictions.We aim to assess the explanatory ability derived from three model-independent methods based on different theoretical frameworks: SHapley Additive exPlanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), and Integrated Gradients (IG). We develop a system based on longformers that can process long documents and then use the explainability methods to extract the relevant segments of text in the EHR that motivated each ICD. We then measure the outcome of the different explainability methods by implementing a novel metric. Results:Our results beat those that carry out the same task by 7%. In terms of explainability degree LIME appears as a stronger technique compared to IG and SHAP. Discussion:Our research reveals that the explored techniques are useful for explaining the output of black box models as the longformer. In addition, the proposed metric emerges as a good choice to quantify the contribution of explainability techniques.

User Stress Detection Using Social Media Text: A Novel Machine Learning Approach

This paper introduces a novel Attention-based Bidirectional Long Short-Term Memory (Bi- LSTM) model for detecting stress in social media text, aiming to enhance mental health monitoring in the digital age. Utilizing the unique communicative nature of social media, this study employs user-generated content to analyze emotional and stress levels. The proposed model incorporates an attention mechanism with the Bi-LSTM architecture to improve the identification of temporal features and context relationships in text data, which is crucial for detecting stress indicators. This model stands out by dynamically focusing on text segments that significantly denote stress, thereby boosting the detection sensitivity and accuracy. Through rigorous testing against baseline models such as Text-CNN, LSTM, GRU, and standard Bi-LSTM, our method demonstrates superior performance, achieving the highest F1-score of 81.21%. These results underscore its potential for practical applications in mental health monitoring where accurate and timely detection of stress is essential.

A novel masking model for Buddhist literature understanding by using Generative Adversarial Networks

This paper is focused on ancient Chinese Buddhist literature understanding. Buddhist literature incorporates a plethora of dialects and slang, which makes it challenging to extract semantic meaning seamlessly. To address this issue, a Generative Adversarial Network Masking Model (GAN-MM) is proposed to pre-train BERT models. This method focuses on optimizing the Masked Language Model (MLM) and takes advantage of the rich semantics of Buddhist terminologies as well as the absence of functional words in Buddhist literature. Furthermore, a semi-supervised learning algorithm was developed to train the GAN-MM. The model was evaluated by assessing its performance on two private tasks related to Buddhist literature understanding: sentiment classification and text segmentation, as well as two public tasks pertaining to the ancient Chinese understanding. The experimental results demonstrated a significant enhancement in the pre-training of BERT models when utilizing GAN-MM, as compared to conventional MLM methods. A large-scale Buddhist dataset, including 20,075 utterance documents and 146 million tokens, is public released at https://data.mendeley.com/datasets/5hzs8w46jh/1.

A Comprehensive Natural Language Processing Pipeline for the Chronic Lupus Disease.

Electronic Health Records (EHRs) contain a wealth of unstructured patient data, making it challenging for physicians to do informed decisions. In this paper, we introduce a Natural Language Processing (NLP) approach for the extraction of therapies, diagnosis, and symptoms from ambulatory EHRs of patients with chronic Lupus disease. We aim to demonstrate the effort of a comprehensive pipeline where a rule-based system is combined with text segmentation, transformer-based topic analysis and clinical ontology, in order to enhance text preprocessing and automate rules' identification. Our approach is applied on a sub-cohort of 56 patients, with a total of 750 EHRs written in Italian language, achieving an Accuracy and an F-score over 97% and 90% respectively, in the three extracted domains. This work has the potential to be integrated with EHR systems to automate information extraction, minimizing the human intervention, and providing personalized digital solutions in the chronic Lupus disease domain.

Enhancing privacy policy comprehension through Privacify: A user-centric approach using advanced language models

As the digital age advances, the collection, usage, and dissemination of personal data have become critical concerns for users, regulators, and the cybersecurity community. Questions surrounding the extent of identifiable data collection, its usage, sharing, selling, and the mechanisms of consent are increasingly central to discussions on user data privacy. These issues highlight the need for effective management and comprehension of privacy policies. To this end, this paper introduces Privacify— a production-ready web application designed to enhance the accessibility and understandability of privacy policies, thus empowering users to make more informed decisions about their data. At its backend, Privacify leverages a combination of text segmentation, summarization using Large Language Model (LLM), and map-reduce technologies to facilitate BASE analysis for single-document insights and WRT and REV for comprehensive cross-document analysis. Designed with a user-centric approach, Privacify features an intuitive interface that presents all relevant user privacy information in easy-to-understand language, complete with a detailed explainability component. This design not only simplifies privacy policies but also aids users in effortlessly navigating complex privacy terms, significantly boosting their ability to protect and manage their personal information. Our evaluation employs robust methodologies, including reliability and accuracy assessments, alongside rigorous functionality verification through ROUGE metrics and human analysis, validating the system’s efficacy and performance. Privacify’s architecture promotes scalability, replicability, and seamless deployment, advancing the domain of user data protection through improved privacy comprehension.

Optical Character Recognition of Balochi Script

Optical Character Recognition is considered one of the fastest methods of data entry. OCR converts the text image representation of x and y coordinates representing pixel information to be converted into text data in a particular language. OCR as a field of pattern recognition and document image understanding, OCR requires a challenging job once a different language text is available on the image. Difference in language script will pose different challenges for OCR which requires entirely different approaches and algorithms. Latin scripts require a different approach whereas the Balochi adopted language scripts require a different approach. In this regard, various solutions have been proposed for different languages. Segmentation is considered one of the important tasks in the process of OCR. A good segmentation will definitely increase the accuracy of an OCR. Segmentation includes the segmentation of text lines from text images which are further divided into words. These segmented words are further divided into characters which are to be recognized. A single segmentation algorithm to segment various scripts of the languages is proposed in this study which checks the script and then segments the text image for the further processing in OCR. The proposed generalized algorithm will check the style, direction and other properties of the script and then adopts the segmentation process to segment text lines, words and characters of the language. The proposed algorithm segments more than ten languages of three scripts and segments for their OCRs. These images can be further processed for feature extraction and classification further. The process of OCR for selected languages will be made easier to recognize. Multiple scripts, languages and images were experimented, and the proposed algorithm successfully segmented 42,833 images of text line, words and character image. The algorithm provides 97% accuracy while segmenting these images and can be extended to further languages as well as scripts .

Immersive e-learning application in intelligent teaching of English composition based on neural network algorithm

Technical solutions for evaluating traditional English teaching composition and other problems, such as excessive subjectivity, waste of time, slow feedback, short scoring time, heavy tasks and so on. This paper first analyzes the theoretical basis of neural network algorithm, and completes the preprocessing of the composition text submitted by students from three aspects: text segmentation, text representation and text fragment retrieval. Then, an intelligent evaluation system of English composition is established, and the scoring standard of English composition is determined. On this basis, the design of English scoring module is improved, and an anti cheating module for common plagiarism problems in English composition is added. Finally, the English composition intelligent evaluation system is tested. It can effectively complete the target task, evaluate and correct the composition, and has a higher recognition success rate and accuracy rate. By studying the algorithm and applying it to the construction of English composition intelligent evaluation system, this paper successfully designs a kind of English composition intelligent evaluation system, which promotes the development of teaching work.

ANALYSIS OF FUNCTIONAL SPEECH ADAPTATION STRATEGIES IN MILITARY TRANSLATION

The research is focused on the aspect of functional speech adaptation of language interpreters who are gaining practical experience at the initial stage in the sphere of foreign-language military-professional communication and paying special attention to the search for original speech strategies to achieve the targets of the translation process in the most efficient way. The critical objective of the article is to theorize the functional speech adaptation and underscore the need and importance to develop interpreters’ unique creative strategies in the process of their adaptive professional maturation for the due carrying out of normal and special translation tasks under the conditions of operational military activities abroad. A detailed review of the scientific sources is carried out concerning research into definitions of linguistic terms such as: “adaptation,” “speech adaptation,” “speech strategies.” The author’s definition of the core of the term “Functional speech adaptation” is given and its main differences are outlined versus the definitions of the linguistic categories “adaptation” and “pragmatic adaptation” proposed by Russian linguistic theorists earlier. The most effective strategies for interpreting are reviewed, viz.: implementation of an interpreter’s anticipation, a strategy for target text forming from the ground of precision information, and a strategy for source text segmentation into minimal translation cells. The study applies the methods of definitional analysis, comparative analysis of theoretical principles with their generalization, theoretical synthesis of the concepts’ definitions, linguistic description method and linguistic modeling. The results of the conducted research make a significant contribution to studying functional speech adaptation of military interpreters in the implementation of interpreting and translation tasks on the foreign military sub-languages material. Moreover, the results of the study can be used for future research purposes of translation studies, theoretical and applied linguistics.

What do Contemporary Publications Report about the Generation of Urban Solid Waste (MSW) and/or Consumption from the Perspective of Chemistry Teaching?

The present study aimed to highlight the research on the topics of urban solid waste generation and consumerism from the perspective of Chemistry teaching between the years 2018 and 2023. To this end, a State-of-the-Art research, according to Romanowski and Ens (2006), in Periódicos Capes database on the aforementioned themes was realized. The methodology used was qualitative. The forms of analysis applied were both mediated by the IRAMuTeQ software and processed by the content analysis of Bardin (2002). In relation to the survey in Periódicos Capes, a textual corpus was obtained with 15 article summaries. Still as results, for textual statistics, the emergence of 82 text segments (ST) was observed, with the use of 65 ST (79.3%) and six classes in the Descending Hierarchical Classification (DHC). This is a satisfactory use and allows us to classify this analysis as consistent and consequently move on to content analysis based on the DHC. The mottos Waste and Generation were the highlights of the similarity analysis, preceded by the mottos Technology, Environmental, Student and Consumerism. In relation to content analysis, two very important categories emerged that communicate well with this object of study. As a consideration, it appears that in the contemporaneity of the last five years, there are strong signs of gaps regarding the use of these themes in the teaching of Chemistry, making it important for researchers to invest more time in the development of the area.

AraFast: Developing and Evaluating a Comprehensive Modern Standard Arabic Corpus for Enhanced Natural Language Processing

The research presented in the following paper focuses on the effectiveness of a modern standard Arabic corpus, AraFast, in training transformer models for natural language processing tasks, particularly in Arabic. In the study described herein, four experiments were conducted to evaluate the use of AraFast across different configurations: segmented, unsegmented, and mini versions. The main outcomes of the present study are as follows: Transformer models trained with larger and cleaner versions of AraFast, especially in question-answering, indicate the impact of corpus quality and size on model efficacy. Secondly, a dramatic reduction in training loss was observed with the mini version of AraFast, underscoring the importance of optimizing corpus size for effective training. Moreover, the segmented text format led to a decrease in training loss, highlighting segmentation as a beneficial strategy in Arabic NLP. In addition, using the study findings, challenges in managing noisy data derived from web sources are identified, which were found to significantly hinder model performance. These findings collectively demonstrate the critical role of well-prepared, segmented, and clean corpora in advancing Arabic NLP capabilities. The insights from AraFast’s application can guide the development of more efficient NLP models and suggest directions for future research in enhancing Arabic language processing tools.

Street office and care practices: emerging conceptions and reflexivities

Objective: To understand the perceptions of users of the "street office" about care practices. Methods: Mixed approach study carried out through semi-structured individual interviews with users of the "street office", with their reports analyzed through the Iramuteq software. Results: The general corpus of the reports consisted of 188 text segments (TS), with 178 TSs (94.68%), with three classes emerging: Class 1 – "Care understood by homeless people", with 85 TS (47.75%); Class 2 – "Satisfaction with access to health actions and services", with 40 TS (22.47%) and Class 3 – "Causes and coping with life on the street", with 53 TS (29.78%). The most recurrent word in the word cloud and in the analysis of similarity was “street”, showing reflections on social inequality and health disparities. Conclusion: The narratives revealed the needs of users regarding their conceptions of care practices, constituting reflectivities emerging from this service instrument.

Clinical research text summarization method based on fusion of domain knowledge

ObjectiveThe objective of this study is to integrate PICO knowledge into the clinical research text summarization process, aiming to enhance the model’s comprehension of biomedical texts while capturing crucial content from the perspective of summary readers, ultimately improving the quality of summaries. MethodsWe propose a clinical research text summarization method called DKGE-PEGASUS (Domain-Knowledge and Graph Convolutional Enhanced PEGASUS), which is based on integrating domain knowledge. The model mainly consists of three components: a PICO label prediction module, a text information re-mining unit based on Graph Convolutional Neural Networks (GCN), and a pre-trained summarization model. First, the PICO label prediction module is used to identify PICO elements in clinical research texts while obtaining word embeddings enriched with PICO knowledge. Then, we use GCN to reinforce the encoder of the pre-trained summarization model to achieve deeper text information mining while explicitly injecting PICO knowledge. Finally, the outputs of the PICO label prediction module, the GCN text information re-mining unit, and the encoder of the pre-trained model are fused to produce the final coding results, which are then decoded by the decoder to generate summaries. ResultsExperiments conducted on two datasets, PubMed and CDSR, demonstrated the effectiveness of our method. The Rouge-1 scores achieved were 42.64 and 38.57, respectively. Furthermore, the quality of our summarization results was found to significantly outperform the baseline model in comparisons of summarization results for a segment of biomedical text. ConclusionThe method proposed in this paper is better equipped to identify critical elements in clinical research texts and produce a higher-quality summary.

Application of elementary probability models for text homogeneity and segmentation: A case study of Bible

Application of elementary probability models for text homogeneity and segmentation: A case study of Bible