Textual Content Research Articles

Introducing 3D printed models of fractures in osteology learning improves clinical reasoning skills among first-year medical students: a pilot study

Abstract Background The human bone anatomy is commonly taught using normal adult bones. However, students often face difficulties comprehending the clinical correlations related to fractures, as they only rely on text content or diagrams without three-dimensional visual aids. Therefore, this study aims to evaluate the effectiveness of using 3D-printed models of limb bone fractures in routine osteology classes to enhance the clinical reasoning skills of first-year medical undergraduate students. Methodology In this experimental study, 105 first-year medical undergraduate students were divided into intervention and control groups based on their pre-assigned serial numbers. The control group was taught using dry adult human bones, with the teacher explaining clinical correlations verbally. Meanwhile, in two sessions, the intervention group was taught using 3D-printed models of fractures in addition to real bones. At the end of the second session, students were evaluated for their clinical reasoning ability using a case-based MCQ test (maximum score 5). The scores were compared between the two groups using an unpaired t-test. Students of the intervention group were asked to rate their learning experience using a 10-point Likert Scale questionnaire. Results The intervention group scored significantly higher (2.54 ± 1.15) than the control group (2.04 ± 0.94) (p = 0.015). The maximum score for both groups was five, and the minimum was one. Most students agreed that the 3D-printed models helped them understand the fractures’ clinical relevance and provided better orientation to the bones, joints, and structures involved in fractures (92%, n = 46). The students expressed a desire for more similar types of sessions. Conclusion Incorporating 3D-printed models of fractures was a novel approach to help students comprehend the clinical correlations. This strategy improved students’ clinical reasoning skills in the intervention group, as evidenced by their higher scores and feedback. Therefore, 3D-printed models are a valuable addition to the traditional teaching methods of learning osteology.

A Robust Recommender System Against Adversarial and Shilling Attacks Using Diffusion Networks and Self-Adaptive Learning

Shilling and adversarial attacks are two main types of attacks against recommender systems (RSs). In modern RSs, existing defense methods are hindered by the following two challenges: (1) the diversity of RSs’ information sources beyond the interaction matrix, such as user comments, textual data, and visual information; and (2) most defense methods are robust only against specific types of adversarial attacks. Ensuring the robustness of RSs against new adversarial attacks across different data sources remains an open problem. To address this problem, we propose a novel method that unifies adversarial attack detection, purification, and fake user detection in RSs by utilizing a guided diffusion adversarial purification network and a self-adaptive training technique. Our approach aims to simultaneously handle both known and unknown adversarial attacks on RSs’ inputs and outputs. We conducted extensive experiments on three large-scale datasets to evaluate the effectiveness of the proposed method. The results confirm that our method can effectively eliminate adversarial perturbations on images and textual content within RSs, surpassing state-of-the-art methods by a significant margin. Moreover, it achieved the best results in three out of five evaluated shilling attack types. Finally, for attacks with realistic magnitudes, it can maintain baseline performance levels even when multiple attacks are applied simultaneously.

“Some harm, some foul”: The editor’s duty of care in minimising the potential harm of creative nonfiction that contains problematic or distressing material

The popular Americanism, “no harm, no foul”, was first used in basketball when the play in question caused no physical harm. Today, the expression is used to suggest that, if no one has been harmed by a misdeed, the “foul” should be excused. Kate Clanchy’s controversial memoir, Some Kids I Taught and What They Taught Me (2019), caused considerable psychological harm to readers – and significant reputational damage to both author and publisher. The critical discussion that ensued divided readers, with critics drawing attention to the ableist, classist, racist, and sexist undertones of the work, while Clanchy’s most vocal supporters argued that her critics were “abusive trolls” (Rajesh, 2021, para. 4) who condemned the book without reading it. Thus, this paper scrutinises the public commentary that surrounded the controversy to demonstrate that there was, indeed, “some harm, some foul”. The paper then assesses the value of harm-preventative measures available to practising Australian editors who work with problematic content in creative nonfiction texts. The paper concludes by offering an evidence-based probability of harm model for editing problematic content, with the overarching goal of minimising potential harm, and promoting safe and responsible editing practice.

“I see wonderful things!” Socio-cognitive experience of Cypriot graffiti in ancient Egypt

Abstract This article analyses Cypriot graffiti in Egypt engraved by tourists and mercenaries in the Archaic and Classical periods (sixth through fourth centuries BC). These graffiti were inscribed on walls of majestic monuments, e.g., the pyramids in Giza and the temples of Karnak and Abydos. They were written in different languages and scripts all employed on the island, such as Cypriot-syllabic Greek, Eteocypriot – an indigenous language still undeciphered – alphabetic Greek and Phoenician, spread throughout the city-kingdoms into which Cyprus was subdivided. The examination is conducted through a socio-cognitive approach, moving beyond previous scholarship which exclusively focused on the graffiti’s concise texts – often limited to the names and provenance of the writers. By comparing Cypriot Egyptian graffiti with contemporary graffiti and twenty-first-century graffiti written in Cypriot-syllabic script, the analysis will shed light on the role of socio-political environment and landscapes in triggering individuals’ – in this case Cypriots’ – cognitive-behavioral responses which lead them to write these texts. This pioneering approach will allow a better understanding of the underlying reasons why these graffiti were engraved and in which locations. The investigation will provide a better knowledge of the origins of the Cypriot writers, level of literacy, and of their role in society (e.g., insiders vs. outsiders), shedding light on communities that would otherwise be unknown. Furthermore, this study lays the foundations for developing an innovative methodology applicable to epigraphic studies, based on integrating landscape interactions and visual impacts to more traditional examination strategies.

Explainable Graph Spectral Clustering of text documents.

Spectral clustering methods are known for their ability to represent clusters of diverse shapes, densities etc. However, the results of such algorithms, when applied e.g. to text documents, are hard to explain to the user, especially due to embedding in the spectral space which has no obvious relation to document contents. Therefore, there is an urgent need to elaborate methods for explaining the outcome of the clustering. We have constructed in this paper a theoretical bridge linking the clusters resulting from Graph Spectral Clustering and the actual document content, given that similarities between documents are computed as cosine measures in tf or tfidf representation. This link enables to provide with explanation of cluster membership in clusters produced by GSA. We present a proposal of explanation of the results of combinatorial and normalized Laplacian based graph spectral clustering. For this purpose, we show (approximate) equivalence of combinatorial Laplacian embedding and of K-embedding (proposed in this paper) and term vector space embedding. We performed an experimental study showing that K-embedding approximates well Laplacian embedding under favourable block matrix conditions and show that approximation is good enough under other conditions. We show also perfect equivalence of normalized Laplacian embedding and the [Formula: see text]-embedding (proposed in this paper) and (weighted) term vector space embedding. Hence a bridge is constructed between the textual contents and the clustering results using both combinatorial and normalized Laplacian based Graph Spectral Clustering methods. We provide a theoretical background for our approach. An initial version of this paper is available at arXiv, (Starosta B 2023). The Reader may refer to that text to get acquainted with formal aspects of our method and find a detailed overview of motivation.

An ontology-oriented approach to traditional dance: on the example of Kazakh dance.

This article is of interest because it approaches traditional Kazakh dance using an ontological perspective. The importance of this method in studying traditional dance is clarified through the example of the Kazakh national dance "Kamyzhai." Kazakh art, which includes aitys (oral poetry contests), epic poetry, instrumental art, singing, dance, handicrafts, and sports traditions, occupies a special place in national culture. National art reflects the specific features of the people’s lifestyle and traditions. In recent years, discussions in Kazakhstan have intensified around the need to consolidate society and strengthen national identity based on the country’s historical heritage and cultural potential, which fosters connections between the past and the present. In his recent address, President Kassym-Jomart Tokayev emphasized the importance of preserving national heritage, considering it a relic passed down from ancestors to the new generation. This article presents a comprehensive ontologically-oriented study of the dance "Kamyzhai," which serves as an initial method for exploring traditional dance, revealing six key components of analysis: dance styles, context and symbolism, musical foundation, poses and movements, costumes and attributes, and traditional rituals and customs. These six main aspects of "Kamyzhai" reveal the connection between textual and dance content, illustrating the upbringing of a girl and her preparation for future social life. The novelty of this study lies in the fact that it is the first research of "Kamyzhai" based on newly discovered factual data. Additional methods used in the article include observation, analytical analysis, and comparison. This article will contribute to enriching the historical and theoretical significance of Kazakh dance, both in practical application and in the educational process.

Urban climate action plans in the United States: A textual content analysis and evaluation

Urban climate action plans in the United States: A textual content analysis and evaluation

Impact of dual-carbon attention competition from local government on regional carbon emissions in China.

Extreme climate change induced by carbon emissions has received extensive attention from governments worldwide. Strong competition in local governments' dual-carbon attention (GCA) produces an effective influence on the reduction of regional carbon emissions, confirming crucial policy implications. In this study, textual content analysis is employed to measure the GCA level and GCA competition, and the mechanism by which GCA competition reduces regional carbon emissions in China is explored from the perspective of competition behaviors. The findings demonstrate that the increase of GCA competition positively influences the reduction of regional carbon emissions. Influence mechanism analyses verify that increasing GCA competition primarily stimulates the greater reduction of regional carbon emissions by intensifying competition in energy consumption reduction and the optimization of the energy consumption structure. The influence of GCA competition on reduced regional carbon emissions is significantly related to industrial structure upgrading, the energy consumption structure, and environmental governance investment, as well as inter-government competition in these areas. The detailed findings of this research can provide economic and environmental benefits for policymakers, and can provide corporations with more targeted policy recommendations related to dual-carbon attention and carbon emission reduction.

Examining correlation between online readiness and content-type preferences of adult learners in an online learning environment

ABSTRACT Current research has two main aims: (a) to identify adults’ content type preferences and (b) examine correlation between adults’ content type preferences and online readiness in online learning environments. For this purpose, the research was conducted as a survey research. The research examined five types of content namely; video, interactive video, textual content, presentation, and video conference, as well as four psycho-educational constructs, which are self-efficacy, self-directed learning, learner control, and motivation as online readiness. A questionnaire was administered to adult learners to identify their preferences, and Configural Frequency Analysis was used to analyze the data. The findings revealed that adults with low levels of psycho-educational constructs tended to prefer only one type of interaction, such as video, presentation, or video conference, while those with high levels of these constructs preferred video conferences to a lesser extent. The findings obtained within the scope of the research will contribute to providing adaptive content for adults in digital learning environments and explaining the relationship between content types and online readiness.

Exploring Students’ Perception on The Implementation of a Web-Based Materials in Intensive Reading Class

Intensive Reading is one of reading subjects learned by English Department students. This course is a reading activity to find information and understand the contents of various reading texts. Good reading skills will enable students find and understand information from various texts proficiently. Besides knowing and applying reading strategies, students can improve their reading skills by doing reading practice continuously. Therefore, students need to have an easy access of various texts that are used as intensive reading materials. Website is one of the media that can be used to provide intensive reading materials. This is a descriptive qualitative research that is aimed to explore students’ perception on the implementation of a web-based materials in intensive reading class. The participants were the second semester of English Department students learned intensive reading. The data were obtained from observation, interview, and documentation. The findings showed that students have positive perception on the implementation of a web-based materials in intensive reading class. It was found that the implementation of websites can motivate students in learning to read, help students improve their reading comprehension, and form students’ reading habits.

A Multimodal Dataset of Fact-Checked News from Chile’s Constitutional Processes: Collection, Processing, and Analysis

This paper presents a multimodal dataset capturing fact-checked news coverage of Chile’s constitutional processes from 2019–2023. The collection comprises 300 articles from three sources: Fast Check, Fact Checking UC, and BioBioChile, containing 242,687 words of text and visual content in 168 entries. The dataset implements advanced natural language processing through RoBERTa and computer vision techniques via EfficientNet, with unified multimodal analysis using the CLIP model. Technical validation through clustering analysis and expert review demonstrates the dataset’s effectiveness in identifying narrative patterns within constitutional process coverage. The structured format includes verification metadata, precomputed embeddings, and documented relationships between textual and visual elements. This enables research into how misinformation propagates through multiple channels during significant political events. This paper details the dataset’s composition, collection methodology, and validation while acknowledging specific limitations. This contribution addresses a gap in current research resources by providing verified multimodal content spanning two constitutional processes, supporting investigations in computational social science and misinformation studies.

(Semi-)automatic Extraction of Urban Planning Rules in French for Better Management of Land Artificialization

Land artificialization is a significant modern concern, as it is irreversible, diminishes agriculturally suitable land and causes environmental problems. Our project, Hérelles, aims to address this challenge by developing a framework for land artificialization management. In this framework, we associate urban planning rules in text form with clusters extracted from time series of satellite images. To achieve this, it is crucial to understand the planning rules with two key objectives: (1) to verify if the constraints derived from the rules are verifiable on satellite images and (2) to use these constraints to guide the labelling (or semantization) of clusters. The first step in this process involves the automatic extraction of rules from urban planning documents written in the French language. To solve this problem, we propose a method based on the multilabel classification of textual segments and their subsequent summarization. This method includes a special format for representing segments, in which each segment has a title and a subtitle. We then propose a cascade approach to address the hierarchy of class labels. Additionally, we develop several text augmentation techniques for French texts that can improve prediction results. Finally, we reformulate classified segments into concise text portions containing necessary elements for expert rule construction. We adapt an approach based on Meaning Representation (AMR) graphs to generate these portions in the French language and conduct a comparative analysis with ChatGPT. We experimentally demonstrate that the resulting framework correctly classifies each type of segment with more than 90% accuracy. Furthermore, our results indicate that ChatGPT outperforms the AMR-based approach, leading to a discussion of the advantages and limitations of both methods.

Automatic summarization of cooking videos using transfer learning and transformer-based models

The proliferation of cooking videos on the internet these days necessitates the conversion of these lengthy video contents into concise text recipes. Many online platforms now have a large number of cooking videos, in which, there is a challenge for viewers to extract comprehensive recipes from lengthy visual content. Effective summary is necessary in order to translate the abundance of culinary knowledge found in videos into text recipes that are easy to read and follow. This will make the cooking process easier for individuals who are searching for precise step by step cooking instructions. Such a system satisfies the needs of a broad spectrum of learners while also improving accessibility and user simplicity. As there is a growing need for easy-to-follow recipes made from cooking videos, researchers are looking on the process of automated summarization using advanced techniques. One such approach is presented in our work, which combines simple image-based models, audio processing, and GPT-based models to create a system that makes it easier to turn long culinary videos into in-depth recipe texts. A systematic workflow is adopted in order to achieve the objective. Initially, Focus is given for frame summary generation which employs a combination of two convolutional neural networks and a GPT-based model. A pre-trained CNN model called Inception-V3 is fine-tuned with food image dataset for dish recognition and another custom-made CNN is built with ingredient images for ingredient recognition. Then a GPT based model is used to combine the results produced by the two CNN models which will give us the frame summary in the desired format. Subsequently, Audio summary generation is tackled by performing Speech-to-text functionality in python. A GPT-based model is then used to generate a summary of the resulting textual representation of audio in our desired format. Finally, to refine the summaries obtained from visual and auditory content, Another GPT-based model is used which combines the output of the frame summary and audio summary modules and give the final enhanced summary. By minimizing the complications involved with traditional and sophisticated methodologies, this research helps with the development of a straightforward but efficient cooking video summarization system. The results achieved in the work are on par with the existing work in the respective field which demonstrates comparable performance and efficacy in converting cooking videos into detailed recipe texts.

Impact of Social Media on Political Campaigns: Management, Strategies and Results

Introduction: The relevance of this topic is that social media has become a crucial tool for communication and political campaigns. As a significant portion of society now resides in the online space, it is crucial to study the strategies and outcomes of using social media in politics to understand the impact of these platforms on public opinion and electoral processes. Analysing this issue enables the identification of significant trends and the development of approaches for the optimal use of social media in a political context. This research aims to comprehensively analyse and systematise the strategies used in political campaigns on social media. Additionally, it seeks to evaluate the influence of these strategies on public opinion and electoral processes. Objectives: The study aims to uncover the critical aspects of social media usage in politics, identify vital strategies, assess their impact on voters, and examine the formation of political leaders' images. The objective also includes studying the positive and negative aspects of using social media in political campaigns and determining opportunities for optimising this processin line with legalandso cialre quirements. Methods: There search methodology in volvesan alysing literary our cesto identify critical theories and concepts related to social media interaction with political campaigns. Case studies are used to explore actual examples of social media usage in political campaigns. Expert surveys are conducted to gather specialist opinions. The content analysis method is applied to thoroughly study textual and visual content on social media during political campaigns. This approach aims to comprehensively analyse the impact of social media on political campaign sand determine the factors that contribute toits effectiveness. Results: The article reveals strategies for influencing social media on political campaigns based on obtained data. It highlights key aspects of their impact, such as interactivity, microtargeting, and the use of viral content. The research also considers legal and sociological aspects of this phenomenon, including the regulation of social campaigns and criteria for the acceptability or unacceptability of certain strategies. Conclusions: Thearticleanalyses social media's influence on political campaigns, exploring different strategies for utilising these platforms and their outcomes. Various methods, such as literature analysis, case studies, expert surveys, and content analysis, have enabled the systematisation of knowledge about the interaction between social media and political campaigns. Various aspects that influence the effectiveness of social media usage in politics have been identified, including microtargeting, viral content, and interactivity. The text discusses different types of political campaigns, such as electoral, lobbying, and socio-activist, providing the reader with a deep understanding and contextualisation of the issue.

Large Language Model Approach for Zero-Shot Information Extraction and Clustering of Japanese Radiology Reports: Algorithm Development and Validation.

The application of natural language processing in medicine has increased significantly, including tasks such as information extraction and classification. Natural language processing plays a crucial role in structuring free-form radiology reports, facilitating the interpretation of textual content, and enhancing data utility through clustering techniques. Clustering allows for the identification of similar lesions and disease patterns across a broad dataset, making it useful for aggregating information and discovering new insights in medical imaging. However, most publicly available medical datasets are in English, with limited resources in other languages. This scarcity poses a challenge for development of models geared toward non-English downstream tasks. This study aimed to develop and evaluate an algorithm that uses large language models (LLMs) to extract information from Japanese lung cancer radiology reports and perform clustering analysis. The effectiveness of this approach was assessed and compared with previous supervised methods. This study employed the MedTxt-RR dataset, comprising 135 Japanese radiology reports from 9 radiologists who interpreted the computed tomography images of 15 lung cancer patients obtained from Radiopaedia. Previously used in the NTCIR-16 (NII Testbeds and Community for Information Access Research) shared task for clustering performance competition, this dataset was ideal for comparing the clustering ability of our algorithm with those of previous methods. The dataset was split into 8 cases for development and 7 for testing, respectively. The study's approach involved using the LLM to extract information pertinent to lung cancer findings and transforming it into numeric features for clustering, using the K-means method. Performance was evaluated using 135 reports for information extraction accuracy and 63 test reports for clustering performance. This study focused on the accuracy of automated systems for extracting tumor size, location, and laterality from clinical reports. The clustering performance was evaluated using normalized mutual information, adjusted mutual information , and the Fowlkes-Mallows index for both the development and test data. The tumor size was accurately identified in 99 out of 135 reports (73.3%), with errors in 36 reports (26.7%), primarily due to missing or incorrect size information. Tumor location and laterality were identified with greater accuracy in 112 out of 135 reports (83%); however, 23 reports (17%) contained errors mainly due to empty values or incorrect data. Clustering performance of the test data yielded an normalized mutual information of 0.6414, adjusted mutual information of 0.5598, and Fowlkes-Mallows index of 0.5354. The proposed method demonstrated superior performance across all evaluation metrics compared to previous methods. The unsupervised LLM approach surpassed the existing supervised methods in clustering Japanese radiology reports. These findings suggest that LLMs hold promise for extracting information from radiology reports and integrating it into disease-specific knowledge structures.

P0499 A multicenter study utilizes multi-scale clinical data from Crohn’s disease patients to develop large-scale language-human interaction models for predicting the progression of intestinal diseases

Abstract Background The hospital admission records of patient with Crohn’s disease (CD) contain extensive textual content that provides a valuable resource for gaining a thorough understanding of the disease development and progression. The medical application of large language model (LLM) has been extensively researched. We aim to employ LLMs to analyze multi-scale clinical data from patients with CD and predict their gastrointestinal outcomes. Methods By analyzing the clinical data and relevant literature, we established a comprehensive database incorporating intestinal disease related factors while selecting three highly representative LLMs (ChatGPT-4omini, ChatGPT-4o, and Gemini-Pro) through prompt engineering of original LLMs with chain-of-thought, knowledge-guided prompting, and self question. Nine combined models were then developed and their performances to predict the intestinal disease were compared. The comprehensiveness, accuracy, conciseness, and readability of logicality reports generated by different LLMs were evaluated. Results The clinical data of 649 patients from six clinical centers were retrospectively collected. Following two rounds of evaluation by ten experts in CD diagnosis and treatment, a total of 117 disease related factors out of 199 were selected to construct the LLM prompt. The ChatGPT-4o’s level-3 model demonstrated superior predictive performance in predicting intestinal outcomes for patients within center-1 as training set, achieving an accuracy of 0.844, sensitivity of 1.000, and specificity with 0.783 (ChatGPT4o-mini: 0.688, 1.000, 0.665; Gemini-pro: 0.758, 1.000, 0.708). For patients from other five medical centers as testing set, the accuracy reached a value at 0.835 (sensitivity: 1.000, specificity: 0.768). However, the readability tests indicate that Gemini-pro offers a better comprehensiveness than others. Conclusion The ChatGPT-4o human-computer interaction model based on multi-scale clinical data of CD patients can accurately predict the progression of intestinal diseases, providing accurate and effective basis for precise management of Crohn’s disease in clinical practice. The Gemini-pro offers a more comprehensible method for non-clinical-experts to under their condition of intestinal disease.

Multi-Modal Twitter Data Analysis for Identifying Offensive Posts Using a Deep Cross Attention based Transformer Framework

In today’s society dissemination of information among the individuals occur very rapidly due to the widespread usage of social media platforms like Twitter (now-a-days acclaimed as X). However, information may pose challenges to maintaining a healthy online environment because often it contains harmful content. This paper presents a novel approach to identify different categories of offensive posts such as hate speech, profanity, targeted insult, and derogatory commentary by analyzing multi-modal image and text data, collected from Twitter. We propose a comprehensive deep learning framework, “Value Mixed Cross Attention Transformer” (VMCA-Trans) that leverage a combination of computer vision and natural language processing methodologies to effectively classify the posts into four classes with binary labels. We have created an in-house dataset (OffenTweet) comprising of Twitter posts having textual content, accompanying with images to build the proposed model. The dataset is carefully annotated by several experts with offensive labels such as hate speech, profanity, targeted insult, and derogatory commentary. VMCA-Trans utilizes fine-tuned state-of-the-art transformer based backbones such as ViT, BERT, RoBERTA etc. The combined representation of image and text embeddings obtained by these fine-tuned transformer encoders, is fed into a classifier to categorize the posts into offensive and non-offensive classes. To assess its effectiveness, we extensively evaluate the VMCA-Trans model using various performance metrics. The results indicate that the proposed multi-modal approach, achieves superior performance compared to traditional unimodal methods.

Роль мелодики речи в формировании семантики стихотворения Б. Л. Пастернака «Как у них»

The purpose of the study is to determine the role of intonation characteristics of B. L. Pasternak's lyrical works in constructing the meaning of the text. The article describes the functions of the melodics of speech in B. L. Pasternak's poem "Like Them", examines the main role of the melodics of a lyrical work, influencing the content of the entire text, helping to create the image of a lyrical hero and revealing the author's intention. The scientific novelty of the research consists in the analysis of the lyrical composition of B. L. Pasternak's poem "Like Them" and in identifying the influence of the melody of the lyrical text on its semantic content. The article describes the techniques of constructing a poetic text, with the help of which the melodic structure of a poem is created, which is directly related to the meaning, semantic and syntactic aspects of the content of a lyrical work. The melody of a particular text is considered as an expressive sign that reveals the state of the hero, the author's position, and the intent of the entire lyrical work. The results of the study of the features of the melodic-forming system of lyrics, phonological and rhythmic-rhymological intersections in the melodic pattern of a poetic text are to identify ways to construct the semantic space of the text and determine the influence of the melodic poetic speech of B. L. Pasternak on the meaning of the work.

Patient education on Instagram? Structured content analysis of the hashtag "obstructivesleepapnea".

Instagram ranks among the most used social media platforms worldwide. An increasing number of posts are dedicated to specific medical topics, such as sleep medicine. The educational content of these posts is largely unknown. Therefore, a structured content analysis of posts linked to the hashtag "obstructivesleepapnea" was conducted, as obstructive sleep apnea (OSA) represents the most common sleep-related breathing disorder. The hashtag "obstructivesleepapnea" was entered into Instagram's search field. The first linked post was selected and then subdivided into visual content and text content for systematic analysis with a focus on educational information on OSA. Demographic factors of the post such as likes, hashtags and the posting account were also included in the analysis. The data collection was completed for N = 150 consecutive posts. 37.3% of the visual content and 32.7% of the text content addressed educational information on OSA. In both subgroups, the most frequently discussed aspects were OSA symptoms, comorbidities, and therapy (visual content: 50.0%, 39.3, and 41.1%, respectively; text content: 42.9%, 44.9%, and 24.9%, respectively). The most common (professional) background of the account, as self-stated by the holder, was dentists (29.5%). Additional sleep medicine content was posted by 34.3% of all accounts. Instagram offers informative content about OSA and is therefore a potential source for patient education. However, the content available is often poorly organized and in most cases incomplete. Patients may have difficulty categorizing the information provided to benefit from it.

Hate Speech Automatic Detection

This study offers a system studying-primarily based absolutely device for automated hate speech detection in textual content the usage of natural Language Processing (NLP) techniques. fashions like Logistic Regression and deep mastering algorithms (LSTM, BERT) are educated on classified datasets with features extracted via TF-IDF. The system evaluates overall overall performance using metrics together with accuracy, precision, keep in mind, and F1-score. A multilingual translation issue powered with the aid of Google Translator enables detection across languages. A Tkinter-based totally GUI allows clients to categorise remarks in actual-time, making sure accessibility and simplicity of use. elegance imbalances are addressed with weighted Logistic Regression, improving detection accuracy. The gadget demonstrates strong overall performance, promoting greater communities with the resource of offering an inexperienced device for computerized hate speech moderation.