There are different goals for literature research, from understanding an unfamiliar topic to generate hypothesis for the next research project. The nature of literature research also varies according to user's familiarity level of the topic. For inexperienced researchers, identifying gaps in the existing literature and generating feasible hypothesis are crucial but challenging. While general ``deep research'' tools can be used, they are not designed for such use case, thus often not effective. In addition, the ``black box" nature and hallucination of Large Language Models (LLMs) often lead to distrust. In this paper, we introduce a human-agent collaborative visualization system AwesomeLit to address this need. It has several novel features: a transparent user-steerable agentic workflow; a dynamically generated query exploring tree, visualizing the exploration path and provenance; and a semantic similarity view, depicting the relationships between papers. It enables users to transition from general intentions to detailed research topics. Finally, a qualitative study involving several early researchers showed that AwesomeLit is effective in helping users explore unfamiliar topics, identify promising research directions, and improve confidence in research results.
AbstractThis article presents an analysis of the French author Tran Minh Huy's memoir Un enfant sans histoire (A Child Without Story, 2022) and two Norwegian novels: Lars Amund Vaage's Syngja (Sing, 2012) and Olaug Nilssen's Tung tids tale (A Tale of Terrible Times, 2017). Huy, Vaage, and Nilssen—all established authors—draw explicitly on their personal experiences as parents of nonverbal children diagnosed with autism. Their works challenge and reshape conventional genre norms and narrative expectations to address the narratological, ethical, and ontological difficulties of representing lives often perceived as lacking traditional “narratability”—that is, lives that appear to have limited dramaturgical progression or apparent agency. The analysis identifies three narrative strategies of triangulation through which the authors construct narrative development, arguing that these approaches contest prevailing assumptions linking narrative coherence with autonomy, agency, and human value. The article further reflects on the potential role of literature within the trilemmatic structure of political ontologies of difference (Boger, 2023). By foregrounding literary form as a site of ethical and ontological inquiry, the article demonstrates how narrative experimentation can interrogate and unsettle normative conceptions of what it means to “have a story.”
Simra Shahid, Marissa Radensky, Raymond Fok
et al.
Automated scientific idea generation systems have made remarkable progress, yet the automatic evaluation of idea novelty remains a critical and underexplored challenge. Manual evaluation of novelty through literature review is labor-intensive, prone to error due to subjectivity, and impractical at scale. To address these issues, we propose the Idea Novelty Checker, an LLM-based retrieval-augmented generation (RAG) framework that leverages a two-stage retrieve-then-rerank approach. The Idea Novelty Checker first collects a broad set of relevant papers using keyword and snippet-based retrieval, then refines this collection through embedding-based filtering followed by facet-based LLM re-ranking. It incorporates expert-labeled examples to guide the system in comparing papers for novelty evaluation and in generating literature-grounded reasoning. Our extensive experiments demonstrate that our novelty checker achieves approximately 13% higher agreement than existing approaches. Ablation studies further showcases the importance of the facet-based re-ranker in identifying the most relevant literature for novelty evaluation.
Aminul Islam, Mukta Bansal, Lena Felix Stephanie
et al.
Writing literature reviews is a common component of university curricula, yet it often poses challenges for students. Since generative artificial intelligence (GenAI) tools have been made publicly accessible, students have been employing them for their academic writing tasks. However, there is limited evidence of structured training on how to effectively use these GenAI tools to support students in writing literature reviews. In this study, we explore how university students use one of the most popular GenAI tools, ChatGPT, to write literature reviews and how prompting frameworks can enhance their output. To this aim, prompts and literature reviews written by a group of university students were collected before and after they had been introduced to three prompting frameworks, namely CO-STAR, POSE, and Sandwich. The results indicate that after being exposed to these prompting frameworks, the students demonstrated improved prompting behaviour, resulting in more effective prompts and higher quality literature reviews. However, it was also found that the students did not fully utilise all the elements in the prompting frameworks, and aspects such as originality, critical analysis, and depth in their reviews remain areas for improvement. The study, therefore, raises important questions about the significance of utilising prompting frameworks in their entirety to maximise the quality of outcomes, as well as the extent of prior writing experience students should have before leveraging GenAI in the process of writing literature reviews. These findings are of interest for educators considering the integration of GenAI into academic writing tasks such as literature reviews or evaluating whether to permit students to use these tools.
Compounding errors pose a significant challenge in automatic literature review generation, as inaccuracies can cascade across multi-stage retrieval and generation workflows. Existing self-correction strategies often lack mechanisms to effectively track and consolidate verified information throughout the process, making it difficult to prevent error accumulation and propagation. In this paper, we propose Structure-Guided Memory Consolidation (SGMC), a novel framework that incrementally consolidates and verifies information using structured representations at each stage of the literature review pipeline. SGMC consists of three key modules: Tree-Guided Memory for hierarchical literature retrieval and outline generation, Hub-Guided Memory for evidence extraction and iterative content refinement, and Self-Loop Memory for proactive error correction via historical feedback. Extensive experiments on public benchmarks and a newly constructed large-scale dataset demonstrate that SGMC achieves state-of-the-art performance in citation accuracy and content quality, significantly mitigating compounding errors in long-form literature review generation.
This study proposes an interpretable prediction framework with literature-informed fine-tuned (LIFT) LLMs for truck driving risk prediction. The framework integrates an LLM-driven Inference Core that predicts and explains truck driving risk, a Literature Processing Pipeline that filters and summarizes domain-specific literature into a literature knowledge base, and a Result Evaluator that evaluates the prediction performance as well as the interpretability of the LIFT LLM. After fine-tuning on a real-world truck driving risk dataset, the LIFT LLM achieved accurate risk prediction, outperforming benchmark models by 26.7% in recall and 10.1% in F1-score. Furthermore, guided by the literature knowledge base automatically constructed from 299 domain papers, the LIFT LLM produced variable importance ranking consistent with that derived from the benchmark model, while demonstrating robustness in interpretation results to various data sampling conditions. The LIFT LLM also identified potential risky scenarios by detecting key combination of variables in truck driving risk, which were verified by PERMANOVA tests. Finally, we demonstrated the contribution of the literature knowledge base and the fine-tuning process in the interpretability of the LIFT LLM, and discussed the potential of the LIFT LLM in data-driven knowledge discovery.
This technical report presents a natural language processing (NLP)-based approach for systematically classifying scientific literature on childhood speech disorders. We retrieved and filtered 4,804 relevant articles published after 2015 from the PubMed database using domain-specific keywords. After cleaning and pre-processing the abstracts, we applied two topic modeling techniques - Latent Dirichlet Allocation (LDA) and BERTopic - to identify latent thematic structures in the corpus. Our models uncovered 14 clinically meaningful clusters, such as infantile hyperactivity and abnormal epileptic behavior. To improve relevance and precision, we incorporated a custom stop word list tailored to speech pathology. Evaluation results showed that the LDA model achieved a coherence score of 0.42 and a perplexity of -7.5, indicating strong topic coherence and predictive performance. The BERTopic model exhibited a low proportion of outlier topics (less than 20%), demonstrating its capacity to classify heterogeneous literature effectively. These results provide a foundation for automating literature reviews in speech-language pathology.
Ingunn Saur Modahl, Hanne Lerche Raadal, Tor Haakon Bakken
Purpose: This article serves to evaluate and increase the understanding of water footprint results when using recommended impact categories available for Life Cycle Assessment (LCA) practitioners. Results are exemplified by analyses of reservoir hydroelectricity with regard to local net evaporation and water scarcity conditions. Methods: An existing Norwegian hydropower plant with connecting reservoirs has been ‘moved’ virtually to several locations in order to evaluate different conditions concerning evaporation rates and water scarcities. The study is based on LCA methodology, as standardised in ISO 14044 and other literature, as well as ISO 14046 for water footprinting. The functional unit is 1 kWh electricity produced from reservoir hydropower and thereafter distributed to the point of use. Net evaporative losses have been employed. The results are presented for the categories water footprint inventory and water scarcity footprint, using the ReCiPe 2016 Midpoint and AWARE methods, as implemented by SimaPro. Results and discussion: Regions have varying levels of evaporation and are more or less water scarce, but these two dimensions are not necessarily correlated. Hence, a reservoir-based hydropower plant’s most burdensome value chain activity might vary depending on the location and the chosen impact category. For water scarcity footprint, the site-specific net evaporation rates have in general more importance than the site-specific water scarcity factors. Conclusions: Both the LCA-based water footprint inventory and the water scarcity footprint are key impact categories in the context of water consumption. They address different facets and are both critical for grasping how local conditions shape the effects of water use. This insight is especially important when using off-the-shelf LCIA methods, without taking time-dependent scarcity factors into account. The study has identified inconsistencies in the literature regarding calculation methods for water footprint inventories and water scarcity footprint. Future studies should aim to include a broader range of locations and employ more spatially explicit and detailed characterisation factors.
Simen Alexander Steindal, Anette Winger, Kirsti Riiser
et al.
BackgroundFamilies with children who need pediatric palliative care (PPC) often stay at home to preserve a sense of normalcy. However, families may experience challenges regarding communication and follow-up from health care professionals (HCPs). Health technology is suggested as a way to facilitate communication between families and HCPs, but no previous scoping review has mapped existing studies on health and communication technologies and infrastructures for supporting children who need PPC and their families.
ObjectiveThe objective of this scoping review was to systematically map the literature on health technologies and infrastructures to support communication in home-based PPC.
MethodsWe conducted a scoping review based on Arksey and O’Malley’s framework with a systematic search for relevant publications in the ASSIA, CINAHL, Embase, MEDLINE, PsycINFO, and Web of Science databases in November 2023, updated on August 28, 2025. Eligible publications comprised children (aged 0-18 years) with life-limiting or life-threating conditions requiring PPC and their families; HCPs, social care workers, or teachers caring for children in need of PPC and using any health technologies and infrastructures to support 2-way communication in home-based care; and literature published between January 1, 2018, and August 28, 2025, in Danish, English, Norwegian, or Swedish. Pairs of authors independently assessed eligibility and extracted data, which were summarized using a descriptive approach.
ResultsThis review included 41 publications: 20 empirical papers, 6 protocol papers, 7 abstracts, 3 brief publications, 2 review papers, and 3 case publications. In 29.3% (12/41) of the publications, the researchers applied user-centered phased-design approaches to develop health technology for PPC. Children with cancer were most often studied in the publications. The most frequent delivery of health technology for communication in home-based PPC combined asynchronous and synchronous modes (19/41, 46.3%). Furthermore, the most frequent health technology apps for communication in home-based care were symptom monitoring apps (15/41, 36.6%), video technology (8/41, 19.5%), and health monitoring and video technology (3/41, 7.3%). Smartphones (14/41, 36.6%), internet and Wi-Fi (12/41, 29.3%), computers or laptops (9/41, 22%), and tablets (9/41, 22%) were the most frequently reported infrastructures.
ConclusionsChildren with cancer and their families are the most frequently reported users of health technology for communicating with HCPs in home-based PPC. However, research on children with diagnoses other than cancer and their families is limited. Combining asynchronous and synchronous modes is the most frequent way to deliver health technology, and children and their families often communicate with HCPs using symptom monitoring apps. Reports of health technology infrastructure for home-based PPC were insufficiently accounted for. Future studies should strive to include the voices of children in the development of health technology to align more closely with their needs.
Trial RegistrationOpen Science Framework t9h4c; https://osf.io/t9h4c/
Computer applications to medicine. Medical informatics, Public aspects of medicine
This paper introduces LLAssist, an open-source tool designed to streamline literature reviews in academic research. In an era of exponential growth in scientific publications, researchers face mounting challenges in efficiently processing vast volumes of literature. LLAssist addresses this issue by leveraging Large Language Models (LLMs) and Natural Language Processing (NLP) techniques to automate key aspects of the review process. Specifically, it extracts important information from research articles and evaluates their relevance to user-defined research questions. The goal of LLAssist is to significantly reduce the time and effort required for comprehensive literature reviews, allowing researchers to focus more on analyzing and synthesizing information rather than on initial screening tasks. By automating parts of the literature review workflow, LLAssist aims to help researchers manage the growing volume of academic publications more efficiently.
The development of AI-assisted chemical synthesis tools requires comprehensive datasets covering diverse reaction types, yet current high-throughput experimental (HTE) approaches are expensive and limited in scope. Chemical literature represents a vast, underexplored data source containing thousands of reactions published annually. However, extracting reaction information from literature faces significant challenges including varied writing styles, complex coreference relationships, and multimodal information presentation. This paper proposes ChemMiner, a novel end-to-end framework leveraging multiple agents powered by large language models (LLMs) to extract high-fidelity chemical data from literature. ChemMiner incorporates three specialized agents: a text analysis agent for coreference mapping, a multimodal agent for non-textual information extraction, and a synthesis analysis agent for data generation. Furthermore, we developed a comprehensive benchmark with expert-annotated chemical literature to evaluate both extraction efficiency and precision. Experimental results demonstrate reaction identification rates comparable to human chemists while significantly reducing processing time, with high accuracy, recall, and F1 scores. Our open-sourced benchmark facilitates future research in chemical literature data mining.
With the development of data-centric AI, the focus has shifted from model-driven approaches to improving data quality. Academic literature, as one of the crucial types, is predominantly stored in PDF formats and needs to be parsed into texts before further processing. However, parsing diverse structured texts in academic literature remains challenging due to the lack of datasets that cover various text structures. In this paper, we introduce AceParse, the first comprehensive dataset designed to support the parsing of a wide range of structured texts, including formulas, tables, lists, algorithms, and sentences with embedded mathematical expressions. Based on AceParse, we fine-tuned a multimodal model, named AceParser, which accurately parses various structured texts within academic literature. This model outperforms the previous state-of-the-art by 4.1% in terms of F1 score and by 5% in Jaccard Similarity, demonstrating the potential of multimodal models in academic literature parsing. Our dataset is available at https://github.com/JHW5981/AceParse.
Harleen Kaur Bagga, Jasmine Bernard, Sahil Shaheen
et al.
Sarcasm is hard to interpret as human beings. Being able to interpret sarcasm is often termed as a sign of intelligence, given the complex nature of sarcasm. Hence, this is a field of Natural Language Processing which is still complex for computers to decipher. This Literature Survey delves into different aspects of sarcasm detection, to create an understanding of the underlying problems faced during detection, approaches used to solve this problem, and different forms of available datasets for sarcasm detection.
Recent breakthroughs in Large Language Models (LLMs) have revolutionized scientific literature analysis. However, existing benchmarks fail to adequately evaluate the proficiency of LLMs in this domain, particularly in scenarios requiring higher-level abilities beyond mere memorization and the handling of multimodal data. In response to this gap, we introduce SciAssess, a benchmark specifically designed for the comprehensive evaluation of LLMs in scientific literature analysis. It aims to thoroughly assess the efficacy of LLMs by evaluating their capabilities in Memorization (L1), Comprehension (L2), and Analysis \& Reasoning (L3). It encompasses a variety of tasks drawn from diverse scientific fields, including biology, chemistry, material, and medicine. To ensure the reliability of SciAssess, rigorous quality control measures have been implemented, ensuring accuracy, anonymization, and compliance with copyright standards. SciAssess evaluates 11 LLMs, highlighting their strengths and areas for improvement. We hope this evaluation supports the ongoing development of LLM applications in scientific literature analysis. SciAssess and its resources are available at \url{https://github.com/sci-assess/SciAssess}.
Previous research on real-time sentence processing in German has shown that listeners use the morphological marking of accusative case on a sentence-initial noun phrase to not only interpret the current argument as the object and patient, but also to predict a plausible agent. So far, less is known about the use of case marking to predict the semantic role of upcoming arguments after the subject/agent has been encountered. In the present study, we examined the use of case marking for argument interpretation in transitive as well as ditransitive structures. We aimed to control for multiple factors that could have influenced processing in previous studies, including the animacy of arguments, world knowledge, and the perceptibility of the case cue. Our results from eye- and mouse-tracking indicate that the exploitation of the first case cue that enables the interpretation of the unfolding sentence is influenced by (i) the strength of argument order expectation and (ii) the perceptual salience of the case cue.PsycINFO code: 2720 Linguistics & Language & Speech.
William B. Reinar, Ole K. Tørresen, Alexander J. Nederbragt
et al.
AbstractRepetitive DNA make up a considerable fraction of most eukaryotic genomes. In fish, transposable element (TE) activity has coincided with rapid species diversification. Here, we annotated the repetitive content in 100 genome assemblies, covering the major branches of the diverse lineage of teleost fish. We investigated if TE content correlates with family level net diversification rates and found support for a weak negative correlation. Further, we demonstrated that TE proportion correlates with genome size, but not to the proportion of short tandem repeats (STRs), which implies independent evolutionary paths. Marine and freshwater fish had large differences in STR content, with the most extreme propagation detected in the genomes of codfish species and Atlantic herring. Such a high density of STRs is likely to increase the mutational load, which we propose could be counterbalanced by high fecundity as seen in codfishes and herring.
Cecilie Haraldseid-Driftland, Hilda Bø Lyng, Veslemøy Guise
et al.
Abstract Background Theories of learning are of clear importance to resilience in healthcare since the ability to successfully adapt and improve patient care is closely linked to the ability to understand what happens and why. Learning from both positive and negative events is crucial. While several tools and approaches for learning from adverse events have been developed, tools for learning from successful events are scarce. Theoretical anchoring, understanding of learning mechanisms, and establishing foundational principles for learning in resilience are pivotal strategies when designing interventions to develop or strengthen resilient performance. The resilient healthcare literature has called for resilience interventions, and new tools to translate resilience into practice have emerged but without necessarily stipulating foundational learning principles. Unless learning principles are anchored in the literature and based on research evidence, successful innovation in the field is unlikely to occur. The aim of this paper is to explore: What are key learning principles for developing learning tools to help translate resilience into practice? Methods This paper reports on a two-phased mixed methods study which took place over a 3-year period. A range of data collection and development activities were conducted including a participatory approach which involved iterative workshops with multiple stakeholders in the Norwegian healthcare system. Results In total, eight learning principles were generated which can be used to help develop learning tools to translate resilience into practice. The principles are grounded in stakeholder needs and experiences and in the literature. The principles are divided into three groups: collaborative, practical, and content elements. Conclusions The establishment of eight learning principles that aim to help develop tools to translate resilience into practice. In turn, this may support the adoption of collaborative learning approaches and the establishment of reflexive spaces which acknowledge system complexity across contexts. They demonstrate easy usability and relevance to practice.
Paul Bjerke, Birgitte Kjos Fonn, Lars Julius Halvorsen
Abstract Norway’s literary policy as a central part of the cultural policy was etablished in the 1960s with the aim to preserve and protect Norwegian language and culture, improve the writers’ economy and thus also secure an aftergrowth of new voices, a stable production infrastructure and—according to library studies—public accessibility to Norwegian literature. The system is among other things based on a purchasing scheme for new Norwegian quality books for public libraries, which provides equal service throughout the country. A VAT-exemption for books contributes to stability and predictability. A library compensation fund compensates authors for income they lose from sales when their books are available in libraries, and a copy compensation fund compensates writers collectively for the use of copyrighted material. The system was originally established for fiction writers. Over the years it has—through extensive political processes—been expanded with a limited number of titles for other genres, most notably non-fiction. All evaluations conclude that it has been highly successful. The main features of the policy have been remarkably stable despite varying governments, due to a largely political consensus on objectives like democratization, free speech and diversity. This does not mean that the system has not been contested. Scholars from the fields of sociology of literature and cultural economics have shown that internationalization and liberalization have gone hand in hand with an enhanced pro-competition stance and that digital distribution contributes to putting the policy under pressure. In later years, the public-management logic has also strengthened its position at the expense of the independence of the field. There is however still robust demand for Norwegian fiction and non-fiction written by award-winning authors as well as new and interesting voices.
In scientific research, the method is an indispensable means to solve scientific problems and a critical research object. With the advancement of sciences, many scientific methods are being proposed, modified, and used in academic literature. The authors describe details of the method in the abstract and body text, and key entities in academic literature reflecting names of the method are called method entities. Exploring diverse method entities in a tremendous amount of academic literature helps scholars understand existing methods, select the appropriate method for research tasks, and propose new methods. Furthermore, the evolution of method entities can reveal the development of a discipline and facilitate knowledge discovery. Therefore, this article offers a systematic review of methodological and empirical works focusing on extracting method entities from full-text academic literature and efforts to build knowledge services using these extracted method entities. Definitions of key concepts involved in this review were first proposed. Based on these definitions, we systematically reviewed the approaches and indicators to extract and evaluate method entities, with a strong focus on the pros and cons of each approach. We also surveyed how extracted method entities are used to build new applications. Finally, limitations in existing works as well as potential next steps were discussed.