Single Statement Research Articles

A Comparison of ChatGPT and Expert Consensus Statements on Surgical Site Infection Prevention in High-Risk Paediatric Spine Surgery.

Artificial intelligence (AI) represents and exciting shift for orthopaedic surgery, where its role is rapidly evolving. ChatGPT is an AI language model which is preeminent among those leading the mass consumer uptake of AI. Artamonov and colleagues compared ChatGPT with orthopaedic surgeons when considering the diagnosis and management of anterior shoulder instability; they found a limited correlation between them. This study aims to further explore how reliable ChatGPT is compared with orthopaedic surgeons. Twenty-three statements were extracted from the article "Building Consensus: Development of a Best Practice Guideline (BPG) for Surgical Site Infection (SSI) Prevention in High-risk Pediatric Spine Surgery" by Vitale and colleagues. These included 14 consensus statements and an additional 9 statements that did not reach consensus. ChatGPT was asked to state the extent to which it agreed with each statement. ChatGPT appeared to demonstrate a fair correlation with most expert responses to the 14 consensus statements. It appeared less emphatic than the experts, often stating that it "agreed" with a statement, where the most frequent response from experts was "strongly agree." It reached the opposite conclusion to the majority of experts on a single consensus statement regarding the use of ultraviolet light in the operating theatre; it may have been that ChatGPT was drawing from more up to date literature that was published subsequent to the consensus statement. This study demonstrated a reasonable correlation between ChatGPT and orthopaedic surgeons when providing simple responses. ChatGPT's function may be limited when asked to provide more complex answers. This study adds to a growing body of discussion and evidence exploring AI and whether its function is reliable enough to enter the high-accountability world of health care. This article is of high clinical relevance to orthopaedic surgery given the rapidly emerging applications of AI. This creates a need to understand the level to which AI can function in the clinical setting and the risks that would entail.

Does conversation lead to better searches? Investigating single-shot and multi-turn spoken searches with children

Interactive speech technologies such as Siri and Amazon Echo allow people to search for information via speech. These systems generally take a single spoken statement or query as input, which may not be enough to fully capture the user’s information need. Therefore, multi-turn conversations have been proposed to establish more detail. A conversational approach can particularly benefit children who experience more difficulty formulating precise queries. We carried out a study with 32 children comparing multi-turn conversations with single-shot (or Query-Response) interactions. We compared the descriptions of the information needs, as well as children’s search experience. Findings indicate that more elaborate descriptions of children’s information needs (more keywords) were gathered through conversational search, leading to more focused search results compared with the Query-Response search. Further analysis of children’s responses yielded insights into their preferences for conversing with speech-based interfaces. This paper also offers design and methodological recommendations for conversational agents that support children’s information search.

Разум, язык, культура: как сегодня прочитывать концепцию «арабского разума» ал-Джа̄бири

In his four-volume Critique of Arab Reason the late Moroccan philosopher Muḥammad ‘Ābid al-Jābirī (1935–2010) argues that human history witnessed “Greco-European” and “Arab” variants of reason, and not the single one. He challenges the confidence of philosophers in the universality of reason which disclosed itself in the course of European history. Reading al-Jābirī’s theory in Kantian optics will prove fruitful in view of significant number of observations accumulated in Arabic philosophical studies and directly related to the question of the type of rationality that grounds Arab-Muslim culture. The Arabic linguistic tradition bases itself not on the being of substance but on the flux of action. The indivisible segment of the Arabic speech is the flow of sound between the two “sides” (ḥ arf, pl. ḥ urūf). The copulative function ensures the gluing of the subject and predicate into a single statement due to the action of “leaning” (isnād) between them, which excludes the use of the “to be”-like copula and removes “being” from the list of the fundamental metaphysical categories. All Arabic words, with few exceptions, are “branches” growing from “roots” represented by names of actions. The “root – branches” (’aṣl – furū‘) model, developed by the representatives of the Arabic linguistic tradition and Islamic jurists, explains the diversity of Islamic culture which allows for direct contradictions, legitimizing them by tracing down to a single “root”, which contrasts with the law of contradiction fundamental to European rationality. Islamic ethics is concerned not with the moral perfection of a single substantial individual, but with maintaining a network of interactions between human being and God and between members of society, constituting society not as a hierarchy of strata and classes, but as multiple networks of interactions. The concept of the “Arab reason” by al-Jābirī, read in Kantian optics on the basis of the achievements of philosophical Arabic studies, allows to substantiate the hypothesis of the plurality of types of rationality developed through the history of humankind.

Mixed-Keying or Desirability-Matching in the Construction of Forced-Choice Measures? An Empirical Investigation and Practical Recommendations

Forced-choice (FC) measures are becoming increasingly popular as an alternative to single-statement (SS) measures. However, to ensure the practical usefulness of an FC measure, it is crucial to address the tension between psychometric properties and faking resistance by balancing mixed keying and social desirability matching. It is currently unknown from an empirical perspective whether the two design criteria can be reconciled, and how they impact respondent reactions. By conducting a two-wave experimental design, we constructed four FC measures with varying degrees of mixed-keying and social desirability matching from the same statement pool and investigated their differences in terms of psychometric properties, faking resistance, and respondent reactions. Results showed that all FC measures demonstrated comparable reliability and induced similar respondent reactions. Forced-choice measures with stricter social desirability matching were more faking resistant, while FC measures with more mixed keyed blocks had higher convergent validity with the SS measure and displayed similar discriminant and criterion-related validity profiles as the SS benchmark. More importantly, we found that it is possible to strike a balance between social desirability matching and mixed keying, such that FC measures can have adequate psychometric properties and faking resistance. A seven-step recommendation and a tutorial based on the autoFC R package were provided to help readers construct their own FC measures.

Nonlinear Jordan triple ∗-derivations on prime ∗-algebras

Let [Formula: see text] be a unital prime ∗-algebra containing a nontrivial projection [Formula: see text]. In this paper, it is shown that if a mapping [Formula: see text] satisfies [Formula: see text] or [Formula: see text] for all [Formula: see text], where [Formula: see text] and [Formula: see text], then [Formula: see text] preserves self-adjoint elements. Therefore, the two relations (0.1) and (0.2) will be equivalent for every [Formula: see text] and [Formula: see text]. This result makes possible to unify the conclusions in a single statement which implies the above two results. In fact, it is shown that if a map [Formula: see text] satisfies the conditions in (0.1) or (0.2) for every [Formula: see text] and [Formula: see text], then [Formula: see text] is an additive ∗-derivation.

Enhancing Engineering Education Through Pedagogical Change: “Application to Abstract

Students' understanding and capacity to apply and integrate principles from different units in each course are not being sufficed by the traditional teaching method, especially for the core courses. This research proposes a method, namely, “Application to Abstract” in which students learn the course by solving an application that draws on principles from various topics. The implementation of the same has been demonstrated with the help of a core course on Fluid Mechanics where a single problem statement with all the necessary information is defined and students work on the different aspect of the problem with the progress of the course. Compared to traditional method, more percentage of students showed interest in derivation of abstract principles (60%), the ability to apply the concepts (40%), ability to correlate with the basic principles and execute the project (50%). Their ability to comprehend and integrate the abstract concepts through the practical application has also enhanced their participation in laboratory classes. Keywords—Fluid Mechanics; Engineering applications; Project based learning; Rubrics design.

Facilitating the implementation of AI-based assistive technologies for persons with disabilities in vocational rehabilitation: A practical design thinking approach

What is vocational learning? It refers to education and training that focuses on acquiring practical skills and knowledge directly related to specific jobs or careers. All over the world, it is unfortunate that a person with a disability is less likely to be employed than one without. AI has served as an important tool in providing assistive technologies for persons with disabilities in vocational learning. Luckily, person which is an extension of innovation that allows you to design solutions for end users with a single problem statement in mind was developed. (Abou-Zahra, Brewer and Copper: 2018). In this research we sought out ways in which AI has helped those with disabilities. Are the AIs really effective? Do they come with any challenges? Can these problems be solved? We discovered some challenges encountered through the use of these AI and found a way to solve them. By leveraging AI algorithms, researchers and developers can create cost-effective solutions that cater to a wide range of disabilities. This includes AI-powered prosthetics, smart glasses for visual assistance, AI voice technologies, such as Siri, Alexa, accessibility tools such as screen readers, captioning services and other assistive devices that enhance mobility, communication, and overall independence and autonomy. AI can track and analyse an individual’s progress in vocational training then adaptive learning platforms can adjust the difficulty and pace of training materials based on the learner’s performance and progress. Incorporating AI into vocational training can significantly enhance accessibility, individualised support, and the overall learning experience for people with disabilities. It empowers them to acquire the skills and knowledge they need to pursue meaningful careers and contributes to greater inclusivity in the workforce.

Log statements generation via deep learning: Widening the support provided to developers

Logging assists in monitoring events that transpire during the execution of software. Previous research has highlighted the challenges confronted by developers when it comes to logging, including dilemmas such as where to log, what data to record, and which log level to employ (e.g., info, fatal). In this context, we introduced LANCE, an approach rooted in deep learning (DL) that has demonstrated the ability to correctly inject a log statement into Java methods in ∼15% of cases. Nevertheless, LANCE grapples with two primary constraints: (i) it presumes that a method necessitates the inclusion of logging statements and; (ii) it allows the injection of only a single (new) log statement, even in situations where the injection of multiple log statements might be essential. To address these limitations, we present LEONID, a DL-based technique that can distinguish between methods that do and do not require the inclusion of log statements. Furthermore, LEONID supports the injection of multiple log statements within a given method when necessary, and it also enhances LANCE’s proficiency in generating meaningful log messages through the combination of DL and Information Retrieval (IR).

EXPLORING THE POTENTIAL OF CHATGPT FOR ASSESSING DELIRIUM SYMPTOMS IN OLDER ADULTS WITH COGNITIVE IMPAIRMENT

Abstract Delirium is a common and serious geriatric syndrome, especially in older adults with cognitive impairment. Accurate caregiver assessment and reporting of delirium in community-dwelling older adults can help healthcare providers diagnose and treat delirium more effectively. Yet, caregivers often lack confidence in their ability to accurately assess and report delirium. An automated method to detect the presence and symptoms of delirium could potentially boost caregiver confidence and assist with more accurate delirium assessment and reporting. This study explored the use of ChatGPT, a generative AI based on the Large Language Model (LLM), to detect the presence and symptoms of delirium. We trained ChatGPT using a series of prompts specifically designed to enhance its understanding of the domain knowledge and the scoring mechanism of a caregiver-centered, seven-item delirium assessment, the Sour Seven. The trained ChatGPT model was asked to identify and score various delirium symptoms in five previously validated geriatric case vignettes using the Sour Seven scale. Each case was repeated three times to test for reliability. The ChatGPT assessment results were compared against those of human experts. The preliminary analysis showed that items 3, 6 and 7 were scored correctly in all cases, but mixed results were found on items 1, 2, 4, and 5, mostly because the AI applied a single symptom statement to multiple items. Overall, the results indicate that generative AI has the potential to accurately capture and represent complex delirium symptoms in natural language, which could assist caregivers in accurately identifying and reporting delirium symptoms.

Facilitating the implementation of AI-based assistive technologies for persons with disabilities in vocational rehabilitation: A practical design thinking approach

What is vocational learning? It refers to education and training that focuses on acquiring practical skills and knowledge directly related to specific jobs or careers. All over the world, it is unfortunate that a person with a disability is less likely to be employed than one without. AI has served as an important tool in providing assistive technologies for persons with disabilities in vocational learning. Luckily, design thinking which is an extension of innovation that allows you to design solutions for end users with a single problem statement in mind was developed.
 In this research we sought out ways in which AI has helped those with disabilities. Are the AIs really effective? Do they come with any challenges? Can these problems be solved? We discovered some challenges encountered through the use of these AI and found a way to solve them. By leveraging AI algorithms, researchers and developers can create cost-effective solutions that cater to a wide range of disabilities.
 This includes AI-powered prosthetics, smart glasses for visual assistance, AI voice technologies, such as Siri, Alexa, accessibility tools such as screen readers, captioning services and other assistive devices that enhance mobility, communication, and overall independence and autonomy. AI can track and analyse an individual’s progress in vocational training then adaptive learning platforms can adjust the difficulty and pace of training materials based on the learner’s performance and progress.
 Incorporating AI into vocational training can significantly enhance accessibility, individualised support, and the overall learning experience for people with disabilities. It empowers them to acquire the skills and knowledge they need to pursue meaningful careers and contributes to greater inclusivity in the workforce.

CapGAN: Text-to-Image Synthesis Using Capsule GANs

Text-to-image synthesis is one of the most critical and challenging problems of generative modeling. It is of substantial importance in the area of automatic learning, especially for image creation, modification, analysis and optimization. A number of works have been proposed in the past to achieve this goal; however, current methods still lack scene understanding, especially when it comes to synthesizing coherent structures in complex scenes. In this work, we propose a model called CapGAN, to synthesize images from a given single text statement to resolve the problem of global coherent structures in complex scenes. For this purpose, skip-thought vectors are used to encode the given text into vector representation. This encoded vector is used as an input for image synthesis using an adversarial process, in which two models are trained simultaneously, namely: generator (G) and discriminator (D). The model G generates fake images, while the model D tries to predict what the sample is from training data rather than generated by G. The conceptual novelty of this work lies in the integrating capsules at the discriminator level to make the model understand the orientational and relative spatial relationship between different entities of an object in an image. The inception score (IS) along with the Fréchet inception distance (FID) are used as quantitative evaluation metrics for CapGAN. IS recorded for images generated using CapGAN is 4.05 ± 0.050, which is around 34% higher than images synthesized using traditional GANs, whereas the FID score calculated for synthesized images using CapGAN is 44.38, which is ab almost 9% improvement from the previous state-of-the-art models. The experimental results clearly demonstrate the effectiveness of the proposed CapGAN model, which is exceptionally proficient in generating images with complex scenes.

Was Stephen Timoshenko right about the Jewish scientists in Germany?

AbstractStephen Timoshenko was the author of numerous textbooks on various aspects of applied mechanics, and through them contributed to its rapid development in the world. As always, the community is also interested in the personal side of the scientists. This paper is devoted to a single statement made by Stephen Timoshenko in his autobiographical book. Specifically, it concerns Timoshenko's at least partial “explanation” of the Holocaust that took place during WWII. Whereas Timoshenko's statement constituted, to the present writer, open antisemitism, it does not seem to disturb many researchers who continue to honor him in various ways. “Nonsense is nonsense, but the history of nonsense is scholarship,” according to Saul Lieberman. Hence the present article presents an investigation of Timoshenko's nonsensical and overly antisemitic statement. This author corresponded with a number of historians of science all repudiating Timoshenko's assertions. Most importantly, the correspondents shed additional vivid light on the Holocaust as experienced by Jewish scientists.

O.4.5-4 What do young people want in a sport program: formative evaluation using Concept Mapping

Abstract Purpose Engaging young people in sport and physical activity (PA) requires awareness of their preferences and priorities. This study provided less active young people (14–17 years of age) with an opportunity to articulate their perspective on an ideal sport program. Methods We used online concept mapping, involving qualitative data collection and quantitative data analysis. Forty-eight less active young people in regional Victoria, Australia brainstormed 65 relevant, unique and single-focus statements in response to the prompt: The features of a sport program that I would really like to participate in are…. Twenty-four participants then sorted these statements into groups based on similarity of meaning, and 37 participants rated each statement on 5 point Likert scales for its importance (1 = completely unimportant; 5 = very important) in a program and how common (1 = never; 5 = always) it is in the programs they participate in. We analysed sorting data using multi-dimensional scaling and hierarchical cluster analysis, and rating data using descriptive statistics. Results We identified a map with the following nine clusters (in order of relative mean importance ratings)—access and convenience (mean importance rating = 4.11 out of 5; mean common rating 3.83 out of 5); social connection (3.98; 3.78); inclusion and belonging (3.98; 3.53); confidence (3.86; 3.46); design and structure (3.85; 3.65); skill development and goals (3.84; 3.60); cost and time (3.80; 3.30); age and gender (3.61; 3.41); and specific features (3.53; 3.25)—as the most appropriate representation of the participants’ sorting data. The two single statements—both in the inclusion and belonging cluster—rated most important were a sport program where you are accepted (4.57) and in which you do not feel judged (4.51). Conclusions Young people want affordable, accessible, convenient, safe, welcoming, socially engaging and inclusive sport and PA programs. They want programs to be well organised, delivered by experienced coaches with flexible content that builds their confidence, fitness and skills through challenging and fun activities. Youth sport and PA program designers can use this data to inform program development and delivery planning. Support/Funding Source Victorian Health Promotion Foundation (VicHealth).

Reporting comprehensive income and firms' earnings management

This study investigates the impact of reporting comprehensive income and other income through performance or equity statements on firms' earnings management through the selective sale of available-for-sale (AFS) securities. The objective is to explore how Canadian firms can improve transparency by altering managers' accounting behavior by switching from equity statements to income statements. The study used a difference-in-differences method in a quasi-experimental framework. The results showed that negative net income and post-period net income in performance statements can reduce realized gains and losses. Available-for-sale securities also affect these gains and losses in performance and equity statements. Single statement adopters only smooth earnings when net income is positive or when available-for-sale gains are large enough to offset negative earnings. Income statement (IS) adopters reduced earnings smoothing more than other treatment firms. Managers with less job security were found to be more likely to engage in earnings smoothing, and earnings smoothing decreased more during the pre-to-post period. Equity incentives reduced the tendency to smooth earnings, particularly in treatment firms with CEOs whose pay is more sensitive to stock price changes. Treatment firms outperform control firms in predicting future earnings using realized gains and losses on AFS securities per share. The findings suggest that firms manage earnings less effectively by selectively selling AFS securities when comprehensive income (CI) or other comprehensive income (OCI) is presented in performance statements. Firms win when earnings management is reduced because it reduces the informativeness of realized gains and losses on AFS securities. Managers will use more earnings management strategies as long as they have incentives.

Nanopublications for Biodiversity Go Live

A nanopublication is the smallest unit of publishable information: a scientifically meaningful assertion about anything that can be uniquely identified and attributed to its author(s) and serve to communicate a single statement, its original source (provenance) and citation record (publication info) (Mons and Velterop 2009, Kuhn et al. 2021). Nanopublications are fully expressed in both human-readable and machine-interpretable formats and can be cited thorugh their unique URI identifiers (IDs). The development of the concept, formats and use cases of biodiversity-specific nanopublications started back in 2016 (Penev et al. 2016). Recently, Pensoft and Knowledge Pixels created a workflow to publish nanopublications related to a published article (Fig. 1). At its core, the generic biodiversity nanopublication consists of a triple structure with a Subject element, a Relation element, and an Object element. The Subject element refers to a specific organism identified by an Organism ID, or groups of organisms (Taxa, identified by Taxon Concept IDs). Universal Unique Identifiers (UUIDs) are minted for both in all cases, although if the given organism already has an identifier, e.g., from digitised collections or observations, then this identifier can be specified too. The Organism is linked via class membership (rdf:type) to a Taxon Concept, which is identified with another UUID that is minted upon nanopublication creation. The need to mint IDs is due to the fact that Organism or Taxon Concept IDs (which refer to taxon names but also details about their interpretation and use) are not currently available in any comprehensive vocabulary or database. The minted Taxon Concept IDs are further defined by linking them to a Taxon Name ID from Catalogue of Life / ChecklistBank (via a relation called "has Taxon Name") and by linking to Taxon Name's concept (interpretation of the name) by a publication's Digital Object Identifier (DOI) or TreatmentBank entry (via a relation called "used as defined in"). Apart from assigning a Taxon Concept, the subject organism can optionally be linked to a life cycle stage (based on the Uber-anatomy (UBERON) Ontology) and to occurrence data via Occurrence ID. The Relation element expresses the association between Subject and Object, according to a particular ontology (e.g., Relation Ontology). The Object element may have different formats. In the template “Association between organisms" the Object has the same structure as the Subject; while in the template “Association between an organism and environment”, the Object is a type of habitat available from the Environment Ontolog (ENVO). In addition to an assertion, a nanopublication has a provenance record (e.g., DOI of the article where the assertion has been published) and the type of evidence entered in the BasisOfRecord field. The publication info always contains the ORCID (Open Researcher and Contributor ID) of the nanopublication's creator, and the timestamp of creation. Another type of nanopublication template annotates an entire article or selected text from it by using the ORCID of the creator as the Subject, the article's DOI as the Object, and the Citation Typing Ontology (CiTO) to express evaluation of the article (e.g., “supports”, “agrees with”, “discusses”). These templates are complemented by free text comments and are available under the Nanopub tab on the article webpages at several Pensoft journals (see example and blog). Nanopublications can also prompt other types of publications, e.g., hypotheses, to become more machine-interpretable, as discussed for invasion biology (Heger et al. 2023).

Students’ Digital Literacy Skill at Universitas PGRI Palembang

Developing digital literacy skills is necessary for students to survive in industrial revolution 4.0. The objective of this study aimed to find out the digital literacy skills of students in the English Education Study Program at Universitas PGRI Palembang. The sample of this study were 25 students enrolled in the second semester of the English education study program at Universitas PGRI Palembang. The survey used a Likert scale to score responses, with Strongly Disagree (SDA) equaling 1, Disagree (DA) equaling 2, Agree (A) equaling 3, and Strongly Agree (SA) equaling 4. Every single statement in the questionnaire was related to the data collected by percentage analysis. Based on The results of data analysis indicated that students of English Education Study Program at University of PGRI Palembang have digital literacy skill. They can use and utilize technology in their daily lives.

Program transformation landscapes for automated program modification using Gin

Automated program modification underlies two successful research areas — genetic improvement and program repair. Under the generate-and-validate strategy, automated program modification transforms a program, then validates the result against a test suite. Much work has focused on the search space of application of single fine-grained operators — copy, delete, replace, and swap at both line and statement granularity. This work explores the limits of this strategy. We scale up existing findings an order of magnitude from small corpora to 10 real-world Java programs comprising up to 500k LoC. We decisively show that the grammar-specificity of statement granular edits pays off: its pass rate triples that of line edits and uses 10% less computational resources. We confirm previous findings that delete is the most effective operator for creating test-suite equivalent program variants. We go farther than prior work by exploring the limits of delete ’s effectiveness by exhaustively applying it. We show this strategy is too costly in practice to be used to search for improved software variants. We further find that pass rates drop from 12–34% for single statement edits to 2–6% for 5-edit sequences, which implies that further progress will need human-inspired operators that target specific faults or improvements. A program is amenable to automated modification to the extent to which automatically editing it is likely to produce test-suite passing variants. We are the first to systematically search for a code measure that correlates with a program’s amenability to automated modification. We found no strong correlations, leaving the question open.

Kontroverse um die Alzheimer-Frühdiagnostik – eine literaturbasierte Übersicht über die Vor- und Nachteile

BackgroundThe continuum of Alzheimer's disease (AD) comprises three stages: the pre-clinical stage (with few to no subjective symptoms), the prodromal stage (measurable mild cognitive impairment) and the final stage of clinically manifest (AD) dementia. Neuropathological correlates in accordance with these stages have been found with varying frequency. The aim of early AD diagnosis is to determine such correlates in these preclinical and prodromal stages and to determine the probability of the manifestation of a later AD dementia. In this regard, the prognostic validity is of vital importance. From the perspective of those affected, the issue of early diagnosis cannot be reduced to AD, although Alzheimer's dementia is by far the most common form of dementia. The aim of this review is to provide a literature-based overview of the advantages and disadvantages of early AD diagnosis in contrast to other types of dementia. Based on this, recommendations will be formulated for the prioritisation of early diagnostic advantages and disadvantages in the patient counselling situation in clinical practice. MethodsThree databases were searched for current reviews addressing the advantages, disadvantages and ethical aspects of early AD diagnosis. The search was limited to current German or English reviews published between January 1, 2018 and November 30, 2022. The systematic search strategy was based on the PICO model and included both a Boolean and a focus-expanding keyword search using previously defined search terms. In addition, snowballing was used as a search strategy. Qualitative synoptic content analysis in accordance with Mayring was used to analyse both advantages and disadvantages. Similar arguments were combined into single statements. The resulting statements were categorized into three main groups: social, individual or clinical aspects. To emphasize the aforementioned aims, arguments were additionally distinguished into “only valid for early AD diagnosis” and ”valid for early diagnosis of all types of dementia“. ResultsOverall, seven reviews with suitable content were included. Social aspects of early AD diagnosis comprise the protection of the affected person and the society as well as potential stigmatisation of the affected and their family members. Individual arguments range from the right to a “rational suicide”, including the avoidance of financial and caregiving burdens on family members, to the right not to know the diagnosis. Clinical arguments include, for example, the scientific research context (facilitating the identification of suitable clinical trial participants) and various negative effects of a false positive or false negative early diagnosis. ConclusionEarly diagnosis of AD as well as of other forms of dementia is characterised by numerous advantages and disadvantages of a social, individual and clinical nature. The decision for or against early diagnosis should always be made by weighing the current advantages against the disadvantages in a specific case. In particular, the desires of the person affected, the validity of the diagnostic procedures available and, above all, the availability of effective preventive or therapeutic measures must be taken into account.

Understanding Chinese Moral Stories with Further Pre-Training

The goal of moral understanding is to grasp the theoretical concepts embedded in a narrative by delving beyond the concrete occurrences and dynamic personas. Specifically, the narrative is compacted into a single statement without involving any characters within the original text, necessitating a more astute language model that can comprehend connotative morality and exhibit commonsense reasoning. The “pretraining + fine-tuning” paradigm is widely embraced in neural language models. In this paper, we propose an intermediary phase to establish an improved paradigm of “pre-training + further pre-training + finetuning”. Further pre-training generally refers to continual learning on task-specific or domain-relevant corpora before being applied to target tasks, which aims at bridging the gap in data distribution between the phases of pre-training and fine-tuning. Our work is based on a Chinese dataset named STORAL-ZH that composes of 4k human-written story-moral pairs. Furthermore, we design a two-step process of domain-adaptive pre-training in the intermediary phase. The first step depends on a newly-collected Chinese dataset of Confucian moral culture. And the second step bases on the Chinese version of a frequently-used commonsense knowledge graph (i.e. ATOMIC) to enrich the backbone model with inferential knowledge besides morality. By comparison with several advanced models including BERTbase, RoBERTa-base and T5-base, experimental results on two understanding tasks demonstrate the effectiveness of our proposed three-phase paradigm.

Food safety culture in food companies: evaluation of the perception of food safety culture in three Tuscan food companies.

The Commission Regulation (EU) No. 2021/382 (European Commission, 2021), amending the Regulation (EC) No. 852/2004 (European Commission, 2004), introduced the obligation for companies to establish and maintain a food safety culture (FSC). The methodology to evaluate, implement, and enhance the level of FSC is up to the individual companies. This study aimed to investigate the perception of FSC among the employees of 3 Tuscan medium-sized enterprises in the food sector, producing cured meat (A), dairy products (B), and frozen fish products (C). The survey was conducted through the development and administration of a questionnaire based on a 5 points Likert scale, referring to different aspects of FSC, organized in 6 sections with 5-6 statements each and subjected to a percentage of employees between 76 and 85%, classified also by the length of service (≤3 and >3 years). For all the companies, the minimum median and mode value for scores obtained by the different sections was 4, and the minimum median and mode value for the single statement was 3 (A, B; except for a bimodal value 2-4) and 4 (C). The section awareness and perception of risk showed the highest mean scores in all companies. As for the length of service, senior employees gave lower scores than junior ones in all sections in B and 3 sections in C. Overall, the results of the questionnaires showed a good perception of FSC, even though it was possible to identify some partial weaknesses.