Search

Background: Dementia is a cognitive decline that leads to the progressive deterioration of an individual's ability to perform daily activities independently. As a result, a considerable amount of time and resources are spent on caretaking. Early detection of dementia can significantly reduce the effort and resources needed for caretaking. Aims: This research proposes an approach for assessing cognitive decline by analysing speech data, specifically focusing on speech relevance as a crucial indicator for memory recall. Methods & Procedures: This is a cross‐sectional, online, self‐administered. The proposed method used deep learning architecture based on transformers, with BERT (Bidirectional Encoder Representations from Transformers) and Sentence‐Transformer to derive encoded representations of speech transcripts. These representations provide contextually descriptive information that is used to analyse the relevance of sentences in their respective contexts. The encoded information is then compared using cosine similarity metrics to measure the relevance of uttered sequences of sentences. The study uses the Pitt Corpus Dementia dataset for experimentation, which consists of speech data from individuals with and without dementia. The accuracy of the proposed multi‐QA‐MPNet (Multi‐Query Maximum Inner Product Search Pretraining) model is compared with other pretrained transformer models of Sentence‐Transformer. Outcomes & Results: The results show that the proposed approach outperforms the other models in capturing context level information, particularly semantic memory. Additionally, the study explores the suitability of different similarity measures to evaluate the relevance of uttered sequences of sentences. The experimentation reveals that cosine similarity is the most appropriate measure for this task. Conclusions & Implications: This finding has significant implications for the early warning signs of dementia, as it suggests that cosine similarity metrics can effectively capture the semantic relevance of spoken language. The persistent cognitive decline over time acts as one of the indicators for prevalence of dementia. Additionally early dementia could be recognised by analysis on other modalities like speech and brain images. WHAT THIS PAPER ADDS: What is already known on this subject: It is already known that speech‐ and language‐based detection methods can be useful for dementia diagnosis, as language difficulties are often early signs of the disease. Additionally, deep learning algorithms have shown promise in detecting and diagnosing dementia through analysing large datasets, particularly in speech‐ and language‐based detection methods. However, further research is needed to validate the performance of these algorithms on larger and more diverse datasets and to address potential biases and limitations. What this paper adds to existing knowledge: This study presents a unique and effective approach for cognitive decline assessment through analysing speech data. The study provides valuable insights into the importance of context and semantic memory in accurately detecting the potential in dementia and demonstrates the applicability of deep learning models for this purpose. The findings of this study have important clinical implications and can inform future research and development in the field of dementia detection and care. What are the potential or actual clinical implications of this work?: The proposed approach for cognitive decline assessment using speech data and deep learning models has significant clinical implications. It has the potential to improve the accuracy and efficiency of dementia diagnosis, leading to earlier detection and more effective treatments, which can improve patient outcomes and quality of life. [ABSTRACT FROM AUTHOR]

Background: Malnutrition is associated with increased morbidity, mortality, and healthcare costs. Early detection is important for timely intervention. This paper assesses the ability of a machine learning screening tool (MUST‐Plus) implemented in registered dietitian (RD) workflow to identify malnourished patients early in the hospital stay and to improve the diagnosis and documentation rate of malnutrition. Methods: This retrospective cohort study was conducted in a large, urban health system in New York City comprising six hospitals serving a diverse patient population. The study included all patients aged ≥ 18 years, who were not admitted for COVID‐19 and had a length of stay of ≤ 30 days. Results: Of the 7736 hospitalisations that met the inclusion criteria, 1947 (25.2%) were identified as being malnourished by MUST‐Plus‐assisted RD evaluations. The lag between admission and diagnosis improved with MUST‐Plus implementation. The usability of the tool output by RDs exceeded 90%, showing good acceptance by users. When compared pre‐/post‐implementation, the rate of both diagnoses and documentation of malnutrition showed improvement. Conclusion: MUST‐Plus, a machine learning‐based screening tool, shows great promise as a malnutrition screening tool for hospitalised patients when used in conjunction with adequate RD staffing and training about the tool. It performed well across multiple measures and settings. Other health systems can use their electronic health record data to develop, test and implement similar machine learning‐based processes to improve malnutrition screening and facilitate timely intervention. Key points/Highlights: Malnutrition is prevalent among hospitalised patients and frequently goes unrecognised, with the potential for severe sequelae. Accurate diagnosis, documentation and treatment of malnutrition have the potential of having a positive impact on morbidity rate, mortality rate, length of inpatient stay, readmission rate and hospital revenue. The tool's successful application highlights its potential to optimise malnutrition screening in healthcare systems, offering potential benefits for patient outcomes and hospital finances. [ABSTRACT FROM AUTHOR]

With the growing use of artificial intelligence, search engine providers are increasingly pushing advertisers to use automated bidding strategies based on machine learning. Such automated decision‐making systems leave advertisers in the dark about the data being used and how they can influence the outcome of the decision‐making process. Previous literature on artificial intelligence lacks an understanding of the dangers related to artificially intelligent systems and their lack of transparency. In response, our paper addresses the inherent risks of the automated optimization of advertisers' bidding strategies in search engine advertising. The selected empirical case of a service company therefore demonstrates how data availability can trigger a long‐term decline in advertising performance and how search engine advertising performance metrics develop before and after an event of data scarcity. Based on data collected for 525 days, difference‐in‐differences analysis shows that the algorithmic approach has a considerable and lasting negative impact on advertising performance. Furthermore, the empirical case indicates that self‐regulated learning can initialize a downward spiral that gradually impairs advertising performance. Thus, the aim of this study is to increase awareness regarding automated decision‐making dangers in search engine advertising and help advertisers take preventive measures to reduce the risks of algorithm missteps. [ABSTRACT FROM AUTHOR]

Background: Despite its obvious relevance to computer science, computational thinking (CT) is transdisciplinary with the potential of impacting one's analytical ability. Although countless efforts have been invested across K‐12 education, there is a paucity of research at the postsecondary level about the extent to which CT can contribute to sustainable learning outcomes. Objectives: The current study examines how a series of Arduino‐based robotics learning activities capture the fuller essence of concepts related to CT. Methods: College students (n = 50) completed a series of six robotics learning activities. Think‐alouds, student reflections and performance scores were used to assess students' CT through a robotics challenge in virtual and physical learning environments. Results and Conclusions: Students verbalized CT concepts related to algorithmic thinking much more than abstraction, problem decomposition and testing and debugging. Students exposed to active learning performed better in a virtual robotics challenge compared to their peers in a traditional‐oriented classroom. Students' scores on the physical robotics challenge increased as a function of the number of references they made to CT concepts during the think‐alouds. It is possible to design pedagogical experiences that tap into various dimensions of CT at incremental levels of complexity through a series of Arduino‐based robotics activities. With the integration of an online simulation, students can visualize and transfer their CT skills between a virtual and physical learning environment, thus leading to more sustainable learning outcomes. Lay Description: What is already known about this topic: Computational thinking (CT) is the conceptual foundation for problem‐solving.CT shifted from unplugged to plugged activities.CT concepts can be fostered through educational robotics. What this paper adds: A multi‐dimensional approach to assessing CT.Applications of CT provide insights into skill transfer.An integrative theoretical approach to study the 'how' and the 'what' of CT. Implications for practice: Educational robotics activities can cater to a fuller range of CT skills.Virtual simulations with constructive feedback are beneficial for students in traditional settings.Mixed learning environments contribute to making links between lectures and labs. [ABSTRACT FROM AUTHOR]

Background: This study is the first to address the topic of schematic versus realistic dynamic visualization with particular focus on the human movement effect (HME) when the content to be learned takes the form of a computer algorithm. An AlgoRythmics dance choreography illustration (HM‐realistic) was compared with an abstract computer animation (schematic). Previous research in the field of dynamic visualizations found schematic illustrations to be more effective, but the examined realistic representations did not include HME. Objectives: The objective of this study was to explore the process of comprehension of computer algorithms from these two types of representations (abstract animation vs. dance choreography illustration). Methods: The experiment (pre‐test, study phase, post‐test) involved 84 undergraduate students and included the following conditions (2 × 2 between‐participants‐design): HM‐realistic → HM‐realistic, schematic → schematic, HM‐realistic → schematic, schematic → HM‐realistic. Results and Conclusions: The results reveal that: (i) the group of participants who saw the dance choreography twice outperformed those who saw the computer animation twice; (ii) dance choreography illustration is only beneficial if it is presented as the second visualization. These findings are apparently contrary to the conclusions of some relevant previous research in the field, but they are in line with recent results regarding the HME. Takeaways: Realistic dynamic visualizations can be more effective than schematic ones if they involve human movement. To benefit maximally from realistic visualization, students need to have previously viewed it or to have previously viewed the isomorphic schematic presentation of it. Lay Description: What is already known about this topic: Studies that researched whether animation aids learners' understanding of dynamic phenomena have reported mixed results.When human movement is used to illustrate the learning content, dynamic visualizations have consistently proven to be superior to static visualizations.Various studies have explored dynamic visualizations under different combinations of schematic and/or realistic representations. Participants who were exposed exclusively to realistic animations exhibited significantly lower performance compared to their counterparts in the other conditions. What this paper adds: Realistic dynamic visualizations can be more effective than schematic ones if they involve human movement.This result is apparently contrary to the conclusions of some relevant previous research in the field of schematic versus realistic dynamic visualizations, but it is in line with recent findings regarding the so‐called human movement effect.To benefit maximally from realistic visualization, students need to have previously viewed it or to have previously viewed the isomorphic schematic presentation of it. Implications for practice and/or policy: The study provides a new insight into the research field of dynamic visualizations and valuable guidelines for instructional design. [ABSTRACT FROM AUTHOR]

Background: It is vital to understand students' Self‐Regulatory Learning (SRL) processes, especially in Blended Learning (BL), when students need to be more autonomous in their learning process. In studying SRL, most researchers have followed a variable‐oriented approach. Moreover, little has been known about the unfolding process of students' SRL profiles. Objectives: We present the insights derived from a study that measured motivation and the learning strategies used by 198 students of a university entry‐level, business school, BL course to develop an understanding of students' SRL processes. Methods: The Strategies for Learning Questionnaire (MSLQ) was used to survey 198 students three times during a semester to investigate SRL profiles and how they unfolded as the course progressed using a person‐oriented approach. Through a clustering approach, we focus on MSLQ's motivation aspects as its importance has been emphasised by different SRL theories, and extant research into motivation in learning analytics (LA) is still lacking. Results and Conclusions: Through the longitudinal clustering approach, we identified minimally, average, and highly SRL profiles. We acknowledged that students might change their SRL profiles as the course progressed as a result of feedback they received. What are the 1 or 2 Major Takeaways from the Study?: This study contributes to the SRL theory by examining students' SRL profiles adaptation longitudinally (addressing the challenge identified regarding the cyclical nature of SRL). This study contributes to LA by investigating motivational constructs currently lacking in the field and bringing forward theory based empirical evidence to inform theory and practice. Lay Description: What is currently known about the subject matter: Self‐regulated learning (SRL) is important for academic achievement, especially during online learning; more research is needed to understand students' SRL and its cyclical nature.Learning analytics (LA) is an advocate for gathering and analysing data for supporting students' learning. It lacks empirical evidence based on theoretical and students' motivational studies.Institutions will benefit from more research on evidence‐based practice in Blended Learning (BL). What this paper adds to that: Evidence for the understanding of students' SRL, identifying student subgroups (different SRL profiles), and observing the cyclical nature of SRL during BL courses.We enhanced LA through the use of empirical research into motivation and its mapping of students' SRL profiles.Contribute to knowledge within BL about the dynamics of students' motivation, strategy use, and SRL, and showing that successful students better self‐regulate their learning.Evidence to support the emerging role of LA in identifying at‐risk students during BL.Evidence shows the most important motivational and learning strategy constructs that are significantly correlated with the final score in BL. The implications of study findings for practitioners: Through LA, students, educators, and institutions can benefit when universities implement BL.Enabling educators to better understand their students' SRL and how students can adapt quite different SRL profiles over time. Through this information, they can help students adopt a better profile. They also can help students in their SRL process, for example, by applying appropriate interventions.Pedagogically helps lecturers when developing the instructional and interventional design. [ABSTRACT FROM AUTHOR]

Background: Polypharmacy in haemodialysis patients can not only manifest as the continued prescribing of unnecessary medications but also has the potential to increase medication‐related hospital admissions, morbidity, and mortality. A validated deprescribing algorithm has recently been developed, specifically targeted at dialysis patients. This paper audits the experience of implementing this algorithm in Australian dialysis settings. Aim: The aim of this paper was to evaluate the usefulness and applicability of the pharmacist‐led Toronto deprescribing tool in Australian haemodialysis settings. Method: The pharmacist‐led deprescribing algorithm was implemented across two metropolitan sites and one rural site. The audit focused on five medications that could potentially be deprescribed in the target patient group (diuretics, alpha blockers, statins, proton pump inhibitors [PPIs], and quinine). Between 1 and 12 months later, a reaudit was conducted, with patients followed up to confirm if medications remained deprescribed. Results: Two hundred and eleven patients across three sites were reviewed. Application of the algorithm resulted in 168 medications deemed appropriate to deprescribe. Of these 168 medications, 56 (33%) were initially deprescribed, with 50 medications (30%) remaining deprescribed on reaudit. The deprescribing rates varied between the three different services, with initial deprescribing rates ranging from 18% to 61%. After follow‐up, deprescribing changes across target medications were fairly static, with only a small number of patients restarting either their diuretic or PPI. Conclusion: The pharmacist‐led deprescribing algorithm resulted in substantial deprescribing across the three sites. Deprescribing rates varied between the sites due to differences in the team model that the pharmacist worked within and the method of the rollout. Further studies should draw on aspects such as finding enablers to overcome prescriber and patient concerns about deprescribing and the aspects of which team models lead to higher rates of successful deprescribing. [ABSTRACT FROM AUTHOR]

Background: Readability metrics provide us with an objective and efficient way to assess the quality of educational texts. We can use the readability measures for finding assessment items that are difficult to read for a given grade level. Hard‐to‐read math word problems can put some students at a disadvantage if they are behind in their literacy learning. Despite their math abilities, these students can perform poorly on difficult‐to‐read word problems because of their poor reading skills. Less readable math tests can create equity issues for students who are relatively new to the language of assessment. Less readable test items can also affect the assessment's construct validity by partially measuring reading comprehension. Objectives: This study shows how large language models help us improve the readability of math assessment items. Methods: We analysed 250 test items from grades 3 to 5 of EngageNY, an open‐source curriculum. We used the GPT‐3 AI system to simplify the text of these math word problems. We used text prompts and the few‐shot learning method for the simplification task. Results and Conclusions: On average, GPT‐3 AI produced output passages that showed improvements in readability metrics, but the outputs had a large amount of noise and were often unrelated to the input. We used thresholds over text similarity metrics and changes in readability measures to filter out the noise. We found meaningful simplifications that can be given to item authors as suggestions for improvement. Takeaways: GPT‐3 AI is capable of simplifying hard‐to‐read math word problems. The model generates noisy simplifications using text prompts or few‐shot learning methods. The noise can be filtered using text similarity and readability measures. The meaningful simplifications AI produces are sound but not ready to be used as a direct replacement for the original items. To improve test quality, simplifications can be suggested to item authors at the time of digital question authoring. Lay Description: What is Known About the Subject: Difficult to read math assessment items cause measurement and equity issuesReadability of math word problems negatively correlated with outcomes What Our Paper Adds: GPT‐3 AI system is capable to simplify math word problemsPrompt‐based and few‐shot learning‐based approaches are able to create meaningful simplifications, but with a very low accuracy rateText similarity and readability measures can be used to filter out noisy outputs and discover interesting simplifications What are the Implications for Practitioners: GPT‐3 AI is capable of generating relevant math world problem simplificationsSimplified text can be used as suggestions to question authors during the digital item authoring process [ABSTRACT FROM AUTHOR]

Background: Assignments that involve writing based on several texts are challenging to many learners. Formative feedback supporting learners in these tasks should be informed by the characteristics of evolving written product and by the characteristics of learning processes learners enacted while developing the product. However, formative feedback in writing tasks based on multiple texts has almost exclusively focused on essay product and rarely included SRL processes. Objectives: We explored the viability of using product and process features to develop machine learning classifiers that identify low‐ and high‐performing essays in a multi‐text writing task. Methods: We examined learning processes and essay submissions of 163 graduate students working on an authentic multi‐text writing assignment. We utilised learners' trace data to obtain process features and state‐of‐the‐art natural language processing methods to obtain product features for our classifiers. Results and Conclusions: Of four popular classifiers examined in this study, Random Forest achieved the best performance (accuracy = 0.80 and recall = 0.77). The analysis of important features identified in the Random Forest classification model revealed one product (coverage of reading topics) and three process (elaboration/organisation, re‐reading and planning) features as important predictors of writing quality. Major Takeaways: The classifier can be used as a part of a future automated writing evaluation system that will support at scale formative assessment in writing tasks based on multiple texts in different courses. Based on important predictors of essay performance, a guidance can be tailored to learners at the outset of a multi‐text writing task to help them do well in the task. Lay Description: What is already known about this topic?: Both product and process features should be used to inform formative feedback on writing.Providing product‐ and process‐oriented feedback to learners is challenging.Automatic writing evaluation systems have mainly relied upon product features.Automated analysis of learners' trace data and their essay drafts is a promising venue. What this paper adds?: An accurate machine learning classifier that identifies low‐ and high‐scoring essays.The classifier utilized both product and process features.We obtained process features from learners' trace data in digital learning environment.We computed product features using state‐of‐the‐art text analytical methods. Implications for practice and/or policy: The classifier can be used as a part of a future automated writing evaluation system.We revealed learning processes and essay characteristics that influence performance.Based on important predictors of performance, formative feedback can be given to learners. [ABSTRACT FROM AUTHOR]

Background: Auditory–perceptual assessment of voice is a subjective procedure. Artificial intelligence with deep learning (DL) may improve the consistency and accessibility of this task. It is unclear how a DL model performs on different acoustic features. Aims: To develop a generalizable DL framework for identifying dysphonia using a multidimensional acoustic feature. Methods & Procedures: Recordings of sustained phonations of /a/ and /i/ were retrospectively collected from a clinical database. Subjects contained 238 dysphonic and 223 vocally healthy speakers of Chinese Mandarin. All audio clips were split into multiple 1.5‐s segments and normalized to the same loudness level. Mel frequency cepstral coefficients and mel‐spectrogram were extracted from these standardized segments. Each set of features was used in a convolutional neural network (CNN) to perform a binary classification task. The best feature was obtained through a five‐fold cross‐validation on a random selection of 80% data. The resultant DL framework was tested on the remaining 20% data and a public German voice database. The performance of the DL framework was compared with those of two baseline machine‐learning models. Outcomes & Results: The mel‐spectrogram yielded the best model performance, with a mean area under the receiver operating characteristic curve of 0.972 and an accuracy of 92% in classifying audio segments. The resultant DL framework significantly outperformed both baseline models in detecting dysphonic subjects on both test sets. The best outcomes were achieved when classifications were made based on all segments of both vowels, with 95% accuracy, 92% recall, 98% precision and 98% specificity on the Chinese test set, and 92%, 95%, 90% and 89%, respectively, on the German set. Conclusions & Implications: This study demonstrates the feasibility of DL for automatic detection of dysphonia. The mel‐spectrogram is a preferred acoustic feature for the task. This framework may be used for vocal health screening and facilitate automatic perceptual evaluation of voice in the era of big data. WHAT THIS PAPER ADDS: What is already known on this subject: Auditory–perceptual assessment is the current gold standard in clinical evaluation of voice quality, but its value may be limited by the rater's reliability and accessibility. DL is a new method of artificial intelligence that can overcome these disadvantages and promote automatic voice assessment. This study explored the feasibility of a DL approach for automatic detection of dysphonia, along with a quantitative comparison of two common sets of acoustic features. What this study adds to existing knowledge: A CNN model is excellent at decoding multidimensional acoustic features, outperforming the baseline parameter‐based models in identifying dysphonic voices. The first 13 mel‐frequency cepstral coefficients (MFCCs) are sufficient for this task. The mel‐spectrogram results in greater performance, indicating the acoustic features are presented in a more favourable way than the MFCCs to the CNN model. What are the potential or actual clinical implications of this work?: DL is a feasible method for the detection of dysphonia. The current DL framework may be used for remote vocal health screening or documenting voice recovery after treatment. In future, DL models may potentially be used to perform auditory–perceptual tasks in an automatic, efficient, reliable and low‐cost manner. [ABSTRACT FROM AUTHOR]

This study investigates scholars' citation behaviors from a fine‐grained perspective. Specifically, each scholarly citation is considered multidimensional rather than logically unidimensional (i.e., present or absent). Thirty million articles from PubMed were accessed for use in empirical research, in which a total of 15 interpretable features of scholarly citations were constructed and grouped into three main categories. Each category corresponds to one aspect of the reasons and motivations behind scholars' citation decision‐making during academic writing. Using about 500,000 pairs of actual and randomly generated scholarly citations, a series of Random Forest‐based classification experiments were conducted to quantitatively evaluate the correlation between each constructed citation feature and citation decisions made by scholars. Our experimental results indicate that citation proximity is the category most relevant to scholars' citation decision‐making, followed by citation authority and citation inertia. However, big‐name scholars whose h‐indexes rank among the top 1% exhibit a unique pattern of citation behaviors—their citation decision‐making correlates most closely with citation inertia, with the correlation nearly three times as strong as that of their ordinary counterparts. Hopefully, the empirical findings presented in this paper can bring us closer to characterizing and understanding the complex process of generating scholarly citations in academia. [ABSTRACT FROM AUTHOR]

Despite the potential importance of emotional aspects in information seeking, there is a lack of adequate attention to emotions' role in facilitating serendipitous information encountering. This paper contributes to this research gap by investigating the role of emotions during the process of perceiving and experiencing serendipitous information encountering in a controlled laboratory setting. The results show that applying a sketch game can stimulate participants' emotions. Our findings indicate that participants are more likely to experience serendipitous information encountering under the influence of positive emotions. This study contributes to an understanding of the relationship between emotions and the perception of serendipitous information encountering. The implications of the possibilities of facilitating positive emotions to induce serendipitous information encountering are discussed. [ABSTRACT FROM AUTHOR]

Machine learning methods, especially deep learning models, have achieved impressive performance in various natural language processing tasks including sentiment analysis. However, deep learning models are more demanding for training data. Data augmentation techniques are widely used to generate new instances based on modifications to existing data or relying on external knowledge bases to address annotated data scarcity, which hinders the full potential of machine learning techniques. This paper presents our work using part‐of‐speech (POS) focused lexical substitution for data augmentation (PLSDA) to enhance the performance of machine learning algorithms in sentiment analysis. We exploit POS information to identify words to be replaced and investigate different augmentation strategies to find semantically related substitutions when generating new instances. The choice of POS tags as well as a variety of strategies such as semantic‐based substitution methods and sampling methods are discussed in detail. Performance evaluation focuses on the comparison between PLSDA and two previous lexical substitution‐based data augmentation methods, one of which is thesaurus‐based, and the other is lexicon manipulation based. Our approach is tested on five English sentiment analysis benchmarks: SST‐2, MR, IMDB, Twitter, and AirRecord. Hyperparameters such as the candidate similarity threshold and number of newly generated instances are optimized. Results show that six classifiers (SVM, LSTM, BiLSTM‐AT, bidirectional encoder representations from transformers [BERT], XLNet, and RoBERTa) trained with PLSDA achieve accuracy improvement of more than 0.6% comparing to two previous lexical substitution methods averaged on five benchmarks. Introducing POS constraint and well‐designed augmentation strategies can improve the reliability of lexical data augmentation methods. Consequently, PLSDA significantly improves the performance of sentiment analysis algorithms. [ABSTRACT FROM AUTHOR]

The impact of algorithms on everyday life is ever increasing. Medicine and public health are not excluded from this development – algorithms in medicine do not only challenge, change and inform research (methods) but also clinical situations. Given this development, questions arise concerning the competency level of prospective physicians, thus medical students, on algorithm related topics. This paper, based on a master's thesis in library and information science written at Humboldt‐Universität zu Berlin, gives an insight into this topic by presenting and analysing the results of a knowledge test conducted among medical students in Germany. F. J. [ABSTRACT FROM AUTHOR]

Developing knowledge‐transforming skills in writing may help students increase learning by actively building knowledge, regardless of the domain. However, many undergraduate students struggle to transform knowledge when drafting essays based on multiple sources. Writing analytics can be used to scaffold knowledge transforming as writers bring evidence to bear in supporting claims. We investigated how to automatically identify sentences representing knowledge transformation in argumentative essays. A synthesis of cognitive theories of writing and Bloom's typology identified 22 linguistic features to model processes of knowledge transforming in a corpus of 38 undergraduates' essays. Findings indicate undergraduates mostly paraphrase or copy information from multiple sources rather than engage deeply with sources' content. Eight linguistic features were important for discriminating evidential sentences as telling versus transforming source knowledge. We trained a machine learning algorithm that accurately classified nearly three of four evidential sentences as knowledge‐telling or knowledge‐transforming, offering potential for use in future research. Lay Description: What is already known about this topic: Engagement in knowledge transforming in multi‐source writing benefits learning.Post‐secondary writers rarely succeed in knowledge transforming.Computer‐generated formative feedback may promote knowledge‐transforming processes.Computational tools so far developed focused on a composition's rhetorical features only. What this paper adds: We proposed a novel methodology to identify knowledge transforming in argumentative essays.Computational approach involved both rhetorical and content characteristics of evidential sentences.Machine learning algorithm was developed to classify text as knowledge‐transforming versus telling.Eight linguistic features predicted evidential sentences as telling or transforming source knowledge. Implications for practice: There is a clear need to teach post‐secondary students tactics for knowledge transforming.The results can inform the development of writing analytics tool to scaffold knowledge‐transforming revisions. [ABSTRACT FROM AUTHOR]