Text Segmentation Research Articles

A Qualitative Exploration of ACPE Standard 4 Key Elements From the Perspective of Student Pharmacists

ObjectiveAttitudes and skills reflecting the Accreditation Council for Pharmacy Education Standard 4 (ACPE-S4) remain an important component of pharmacy education. The objectives of this study were (1) explore student perspectives on the importance of ACPE-S4 key elements, and (2) identify experiences that students described as supporting ACPE-S4 attitude and skill development. MethodsThis was a qualitative study using focus groups and dyadic interviews with student pharmacists to collect data on ACPE-S4 key element importance and skill acquisition. A general inductive qualitative approach was used to analyze data, with 2 study authors independently reviewing all transcripts and assigning codes to text segments. Authors then met to discuss coding, iteratively organize coded segments into categories and subcategories, and select representative quotes. ResultsOverall, 3 focus groups and 3 dyadic interviews were conducted with a total of 20 students. Students ranked professionalism and self-awareness as the 2 most important key elements. From qualitative analysis, 4 categories and 10 subcategories of educational activities, programs, and experiences were identified as follows: Mentorship (Identifying a Mentor, Formal Mentorship Programs), Curriculum (Practice Lab, Professional Development Course, Interprofessional Education, Experiential Education), Cocurriculum (Organizational Involvement, Volunteer Hours), and Opportunities Outside Academic Setting (Work, Conferences, and Innovative Pharmacy Practices). ConclusionParticipants articulated several opportunities for developing the skills and attitudes reflecting ACPE-S4 key elements. Incorporating student input into ACPE-S4 attitude and skill development may help pharmacy programs consider how to design, refine, and implement initiatives and experiences to help students develop skills across ACPE-S4 focus areas.

Heuristic-based text segmentation of bilingual handwritten documents for Gurumukhi-Latin scripts

Heuristic-based text segmentation of bilingual handwritten documents for Gurumukhi-Latin scripts

The conceptualisation and operationalisation of ‘marketing’ in public health research: a review of reviews focused on food marketing using principles from critical interpretive synthesis

BackgroundExtensive public health research reports the nature, scope and effects of various marketing activities used by food and drinks companies to support the sale of their products. Such literature informs the regulation of food marketing that encourages unhealthy eating behaviours and poor diet-related health outcomes. However, it is not clear whether this literature consistently conceptualises and applies marketing, which could in turn influence the approach and efficacy of policies to regulate food marketing. We aimed to understand the conceptualisation and operationalisation of marketing in public health research of food marketing, eventually focusing on the conceptualisation of integrated marketing.MethodsWe conducted a review of reviews that drew on scoping review methods and applied principles of critical interpretive synthesis. Five databases of peer-reviewed literature and websites of relevant organisations were searched in June – August 2020. Articles were screened against inclusion criteria to identify reviews examining food marketing in a health context. Informative text segments from included articles were coded using NVivo. Codes were grouped into synthetic constructs and a synthesising argument.ResultsAfter screening against inclusion criteria, 60 publications were eligible for inclusion. Informative text segments from 24 publications were coded, after which no new codes were identified. Our synthesising argument was that the understanding of integrated marketing appeared inconsistent across publications, such as by differences in use of underlying conceptual frameworks and in the application of terms such as marketing strategy and tactics.ConclusionsUsing our synthesising argument, we suggest ways to improve the future study of food marketing in public health research, for example by using in-depth case studies to understand the integrated operation and effect of multi-component marketing strategies. Improving conceptual clarity in the study of food marketing in public health research has the potential to inform policy that is more reflective of the true nature of marketing, and thus more effective in combating food marketing effects and protecting public health.Protocol registrationThe review protocol was made publicly available on Open Science Framework prior to the start of the study (DOI: https://doi.org/10.17605/OSF.IO/VSJCW).

Teaching text structure in science education

Outcomes from a body of research on reading education stress the importance of text structure instruction and the integration of reading instruction in content learning in primary schools. We conducted an analysis of Dutch textbooks and workbooks of the seven most frequently used science teaching programs developed for grades 3-6 in order to find out to what extent these materials are appropriate for text structure instruction. The results show that the materials do offer opportunities for text structure instruction but are not suitable for a step-by-step implementation and practice of knowledge and skills related to text structure. The textbooks offer too few clear organized single-structured texts and not all text structures are sufficiently represented. Also, the programs differ strongly in text length, text segmentation and use of introductions. Illustrations and assignments hardly connect to the text structure. We finish the paper with recommendations for educational publishers.

TextStyleBrush: Transfer of Text Aesthetics From a Single Example.

We present a novel approach for disentangling the content of a text image from all aspects of its appearance. The appearance representation we derive can then be applied to new content, for one-shot transfer of the source style to new content. We learn this disentanglement in a self-supervised manner. Our method processes entire word boxes, without requiring segmentation of text from background, per-character processing, or making assumptions on string lengths. We show results in different text domains which were previously handled by specialized methods, e.g., scene text, handwritten text. To these ends, we make a number of technical contributions: (1) We disentangle the style and content of a textual image into a non-parametric, fixed-dimensional vector. (2) We propose a novel approach inspired by StyleGAN but conditioned over the example style at different resolution and content. (3) We present novel self-supervised training criteria which preserve both source style and target content using a pre-trained font classifier and text recognizer. Finally, (4) we also introduce Imgur5K, a new challenging dataset for handwritten word images. We offer numerous qualitative photo-realistic results of our method. We further show that our method surpasses previous work in quantitative tests on scene text and handwriting datasets, as well as in a user study.

A Performance Comparison of Deep Learning Models for Scene Text Segmentation

A Performance Comparison of Deep Learning Models for Scene Text Segmentation

FindZebra online search delving into rare disease case reports using natural language processing.

Early diagnosis is crucial for well-being and life quality of the rare disease patient. Access to the most complete knowledge about diseases through intelligent user interfaces can play an important role in supporting the physician reaching the correct diagnosis. Case reports may offer information about heterogeneous phenotypes which often further complicate rare disease diagnosis. The rare disease search engine FindZebra.com is extended to also access case report abstracts extracted from PubMed for several diseases. A search index for each disease is built in Apache Solr adding age, sex and clinical features extracted using text segmentation to enhance the specificity of search. Clinical experts performed retrospective validation of the search engine, utilising real-world Outcomes Survey data on Gaucher and Fabry patients. Medical experts evaluated the search results as being clinically relevant for the Fabry patients and less clinically relevant for the Gaucher patients. The shortcomings for Gaucher patients mainly reflect a mismatch between the current understanding and treatment of the disease and how it is reported in PubMed, notably in the older case reports. In response to this observation, a filter for the publication date was added in the final version of the tool available from deep.findzebra.com/<disease> with <disease> = gaucher, fabry, hae (Hereditary angioedema).

Profile of publications on tourism with elderly people in Brazil

Tourism and leisure are sectors that have been awakening the demand of elderly consumers. The aim of this article was to map the literature on Brazilian productions on tourism and human aging. For this purpose, a Scoping review was carried out using the PRISMA method. Data collection did not have publication time delimitation and CAPES and BDTD databases were used. The 46 selected documents were analyzed by IRaMuTeQ. The results were presented using the descending hierarchical classification method and the word cloud. The corpus was divided into 365 text segments and 59.06% of these data were evaluated. 12,723 different elements were analyzed. Descending hierarchical classification separated the corpus into five classes of textual elements. The importance of these publications for a national panorama and future analyzes of the development of this theme in our society are highlighted.

DQ-DETR: Dual Query Detection Transformer for Phrase Extraction and Grounding

In this paper, we study the problem of visual grounding by considering both phrase extraction and grounding (PEG). In contrast to the previous phrase-known-at-test setting, PEG requires a model to extract phrases from text and locate objects from image simultaneously, which is a more practical setting in real applications. As phrase extraction can be regarded as a 1D text segmentation problem, we formulate PEG as a dual detection problem and propose a novel DQ-DETR model, which introduces dual queries to probe different features from image and text for object prediction and phrase mask prediction. Each pair of dual queries are designed to have shared positional parts but different content parts. Such a design effectively alleviates the difficulty of modality alignment between image and text (in contrast to a single query design) and empowers Transformer decoder to leverage phrase mask-guided attention to improve the performance. To evaluate the performance of PEG, we also propose a new metric CMAP (cross-modal average precision), analogous to the AP metric in object detection. The new metric overcomes the ambiguity of Recall@1 in many-box-to-one-phrase cases in phrase grounding. As a result, our PEG pre-trained DQ-DETR establishes new state-of-the-art results on all visual grounding benchmarks with a ResNet-101 backbone. For example, it achieves 91.04% and 83.51% in terms of recall rate on RefCOCO testA and testB with a ResNet-101 backbone.

SegFormer: A Topic Segmentation Model with Controllable Range of Attention

Topic segmentation aims to reveal the latent structure of a document and divide it into multiple parts. However, current neural solutions are limited in the context modeling of sentences and feature representation of candidate boundaries. This causes the model to suffer from inefficient sentence context encoding and noise information interference. In this paper, we design a new text segmentation model SegFormer with unidirectional attention blocks to better model sentence representations. To alleviate the problem of noise information interference, SegFormer uses a novel additional context aggregator and a topic classification loss to guide the model to aggregate the information within the appropriate range. In addition, SegFormer applies an iterative prediction algorithm to search for optimal boundaries progressively. We evaluate SegFormer's generalization ability, multilingual ability, and application ability on multiple challenging real-world datasets. Experiments show that our model significantly improves the performance by 7.5% on the benchmark WIKI-SECTION compared to several strong baselines. The application of SegFormer to a real-world dataset to separate normal and advertisement segments in product marketing essays also achieves superior performance in the evaluation with other cutting-edge models.

End-to-End Entity Linking with Hierarchical Reinforcement Learning

Entity linking (EL) is the task of linking the text segments to the referring entities in the knowledge graph, typically decomposed into mention detection, and entity disambiguation. Compared to traditional methods treating the two tasks separately, recent end-to-end entity linking methods exploit the mutual dependency between mentions and entities to achieve better performance. However, existing end-to-end EL methods have problems utilizing the dependency of mentions and entities in the task. To this end, we propose to model the EL task as a hierarchical decision-making process and design a hierarchical reinforcement learning algorithm to solve the problem. We conduct extensive experiments to show that the proposed method achieves state-of-the-art performance in several EL benchmark datasets. Our code is publicly available at https://github.com/lhlclhl/he2eel.

Nursing work during the COVID-19 pandemic: (dis)satisfaction and (de)motivation

Objective: to show the factors that lead to (dis)satisfaction and (de)motivation in the work of nursing professionals in the COVID-19 pandemic. Methods: qualitative study, developed virtually, with professionals from nursing teams. Data was processed using the software Interface de R pour Analyses Multidimensionnelles de Textes et de Questionnaires and analyzed according to Thematic Analysis. Results: 72 workers participated. 45 (62.5%) stated to be satisfied and 46 (63.9%) to be motivated for work. The text segments were separated into four classes, each showing a factor that interfered in work (dis)satisfaction and (de)motivation. Conclusion: factors that intervene in satisfaction and motivation were mostly associated with patient recovery and helping one another. Among the factors that impacted on dissatisfaction and demotivation, the lack of needed supplies and work overload stood out. Contributions to practice: we identified needs and strategies relevant to improve the work environment.

Align-then-abstract representation learning for low-resource summarization

Generative transformer-based models have achieved state-of-the-art performance in text summarization. Nevertheless, they still struggle in real-world scenarios with long documents when trained in low-resource settings of a few dozen labeled training instances, namely in low-resource summarization (LRS). This paper bridges the gap by addressing two key research challenges when summarizing long documents, i.e., long-input processing and document representation, in one coherent model trained for LRS. Specifically, our novel align-then-abstract representation learning model (Athena) jointly trains a segmenter and a summarizer by maximizing the alignment between the chunk-target pairs in output from the text segmentation. Extensive experiments reveal that Athena outperforms the current state-of-the-art approaches in LRS on multiple long document summarization datasets from different domains.

LGBTQI+ representation in pre-licensure nursing textbooks: A qualitative descriptive analysis

BackgroundTo improve health equity, nursing curricula should include content specific to the needs of marginalized and underserved communities, such as lesbian, gay, bisexual, transgender, queer, and intersex groups (LGBTQI+). Harmful and absent academic discussions of LGBTQI+ patients lead to provider discomfort and inadequacy in treating this patient population. Nursing schools are well-positioned to increase comfort with LGBTQI+ content as part of pre-licensure curricula. This article presents a systematic evaluation of LGBTQI+ content in nursing pre-licensure textbooks and the nature and quality of the representations. MethodsA qualitative descriptive analysis of LGBTQI+ content from 14 nursing-specific textbooks required by a pre-licensure degree program at the Johns Hopkins School of Nursing was conducted by a student-led team with faculty oversight. A priori and iterative search terms were used to identify and extract text segments that referenced LGBTQI+ content in each textbook. An iterative codebook was developed, codes were applied, and analysis of the information and context in which the terms were presented was performed. ResultsThe research team observed gaps and notable patterns in distribution of LGBTQI+ terms and health content areas across the textbooks reviewed. The majority of LGBTQI+ search terms were identified in the following health content areas: social determinants of health, sexual/reproductive health, pediatric sexual & gender diversity, intersectionality, and infectious disease. Based on qualitative descriptive analyses, the data were organized into the following categories: a) Language; b) Medicalization; c) Vague, Incomplete, or Lacking Specificity; and d) Comprehensive Approach. ConclusionFindings highlight the need for increased academic exposure for pre-licensure nursing students regarding the care of LGBTQI+ patients. Thoughtful inclusion of LGBTQI+ content may better foster the delivery of evidence-based care for this patient population. These findings underscore the need for improved nursing curricula to support nurses in delivering affirming care for LGBTQI+ populations.

CDText: Scene text detector based on context-aware deformable transformer

Scene text detection task aims to precisely locate text regions in natural scenes. However, the existing methods still face challenges in detecting arbitrary-shaped text, due to their limited feature representation capability. To alleviate this problem, we propose a scene text detector, i.e., CDText, based on structure of context-aware deformable transformer. Specifically, CDText firstly adopts different convolution kernel designs for feature extraction, which designs receptive fields with different size for multi-scale feature perception and fusion. Meanwhile, multi-head self-attention mechanism is used to strengthen the reasoning ability of CDText in a global sense, thus enhancing feature maps with abundant context information by extracting implicit relationship between multi-scale text features. Moreover, CDText designs a segmentation head to segment text instances of arbitrary shapes from rectangular detection boxes. Experiments show that CDText is superior to comparative methods in detection accuracy, achieving F-scores of 92.7, 81.9, and 82.9 on ICDAR2013, Total Text, and CTW-1500 datasets, respectively.

Multi-labeling strategy based on a heuristic text segmentation for educational content

Multi-labeling strategy based on a heuristic text segmentation for educational content

An interpretable diagnostic approach for lung cancer: Combining maximal clique and improved BERT

AbstractThe lung cancer incidence and mortality in China have always been high. Moreover, due to the limited level of professional technology, misdiagnosis and missed diagnosis of lung cancer often occur. To improve the accuracy of diagnosis, this paper proposes an interpretable diagnostic method for lung cancer based on Chinese electronic medical records (EMRs). First, to overcome the difficulty in word segmentation of clinical texts in Chinese EMRs, a dictionary construction method is proposed based on the idea of maximal clique, and 730 medical professional terms related to lung diseases are identified. Then, the ProbSparse self‐attention mechanism and self‐attention distilling operation in Informer are used to improve the Bidirectional Encoder Representations from Transformer (BERT) to realize the representation of long clinical texts with lower time complexity and memory consumption. Finally, the convolutional neural network with an attention mechanism is employed to process the representation results to realize the interpretable prediction of lung cancer. This method is applied to the lung cancer diagnosis of inpatients in a tertiary hospital in Hunan Province, obtaining excellent results of about 0.9 for area under the receiver operating characteristic curve (AUROC) and area under the precision‐recall curve (AUPRC). In addition, the results of the comparative analysis with existing dictionaries, word embedding methods and diagnostic methods further confirm the superiority of the proposed method. Specifically, the proposed method improves the precision by at least 6%, the recall by at least 2.6%, the F1 score by at least 5.2%, AUROC by at least 7.3% and AUPRC by at least 7.7% compared with all these state‐of‐the‐art methods.

Preface to the Second Issue

I consider it my privilege to introduce the second issue of the Artificial Intelligence and Applications (AIA) Journal. I extend my warm welcome to the second issue and thank all the authors of the papers included. The highlight of the second issue is the exploration of the latest AI techniques addressing the new challenges like text segmentation from the historical documents and images, developing smart sensing technologies, studying industrial revolution readiness among industry players, precipitation prediction, human-centric functional computing, Devanagari digit recognition, and the various multi-party agent issues by the papers in this issue. These are new and emerging applications with immense scope for the understanding of the niche areas and future directions. It would also motivate the readers to find new directions to solve future challenges. I am sure the pertinent issue will draw the attention of the numerous budding researchers from various disciplines including academics and industry. Moreover, I think our journal has seized the right moment to release the second issue in resonance with the current flourishing of AI-based techniques and models which can solve more complex challenges and problems with great accuracy but with minimum expenses. For example, ChatGPT can solve problems related to natural language processing, and sometimes, it can even answer questions that do not have a solution at all. In the same way, 3D printers can be used to construct houses and buildings, etc. Therefore, one can expect an intelligent system could be able to overtake the human mind and thinking. I believe that the second issue and future issues of AIA will become a bellwether to human society. Finally, I appreciate the team led by EiC of the AIA. They have put their efforts into making AIA a successful journal and reaching society through publishing innovative technologies to meet the current demands. Thus, I wish for the whole team and the authors to be part of the great journey of AIA.

Reception Reader: Exploring Text Reuse in Early Modern British Publications

The Reception Reader is a web tool for studying text reuse in the Early English Books Online (EEBO-TCP) and Eighteenth Century Collections Online (ECCO) data. Users can: 1) explore a visual overview of the reception of a work, or its incoming connections, across time based on shared text segments, 2) interactively survey the details of connected documents, and 3) examine the context of reused text for "close reading". We show examples of how the tool streamlines research and exploration tasks, and discuss the utility and limitations of the user interface along with its current data sources.

A Convolutional Recurrent Neural-Network-Based Machine Learning for Scene Text Recognition Application

Optical character recognition (OCR) is the process of acquiring text and layout information through analysis and recognition of text data image files. It is also a process to identify the geometric location and orientation of the texts and their symmetrical behavior. It usually consists of two steps: text detection and text recognition. Scene text recognition is a subfield of OCR that focuses on processing text in natural scenes, such as streets, billboards, license plates, etc. Unlike traditional document category photographs, it is a challenging task to use computer technology to locate and read text information in natural scenes. Imaging sequence recognition is a longstanding subject of research in the field of computer vision. Great progress has been made in this field; however, most models struggled to recognize text in images of complex scenes with high accuracy. This paper proposes a new pattern of text recognition based on the convolutional recurrent neural network (CRNN) as a solution to address this issue. It combines real-time scene text detection with differentiable binarization (DBNet) for text detection and segmentation, text direction classifier, and the Retinex algorithm for image enhancement. To evaluate the effectiveness of the proposed method, we performed experimental analysis of the proposed algorithm, and carried out simulation on complex scene image data based on existing literature data and also on several real datasets designed for a variety of nonstationary environments. Experimental results demonstrated that our proposed model performed better than the baseline methods on three benchmark datasets and achieved on-par performance with other approaches on existing datasets. This model can solve the problem that CRNN cannot identify text in complex and multi-oriented text scenes. Furthermore, it outperforms the original CRNN model with higher accuracy across a wider variety of application scenarios.