Artigos de revistas sobre o tema "Randomized sketches"

Quantiles are fundamental statistics in various data science tasks, but costly to compute, e.g., by loading the entire data in memory for ranking. With limited memory space, prevalent in end devices or databases with heavy loads, it needs to scan the data in multiple passes. The idea is to gradually shrink the range of the queried quantile till it is small enough to fit in memory for ranking the result. Existing methods use deterministic sketches to determine the exact range of quantile, known as deterministic filter, which could be inefficient in range shrinking. In this study, we propose to shrink the ranges more aggressively, using randomized summaries such as KLL sketch. That is, with a high probability the quantile lies in a smaller range, namely probabilistic filter, determined by the randomized sketch. Specifically, we estimate the expected passes for determining the exact quantiles with probabilistic filters, and select a proper probability that can minimize the expected passes. Analyses show that our exact quantile determination method can terminate in P passes with 1-δ confidence, storing O(N 1/P logP-1/2P (1/δ)) items, close to the lower bound Ømega(N1/P) for a fixed δ. The approach has been deployed as a function in an LSM-tree based time-series database Apache IoTDB. Remarkably, the randomized sketches can be pre-computed for the immutable SSTables in LSM-tree. Moreover, multiple quantile queries could share the data passes for probabilistic filters in range estimation. Extensive experiments on real and synthetic datasets demonstrate the superiority of our proposal compared to the existing methods with deterministic filters. On average, our method takes 0.48 fewer passes and 18% of the time compared with the state-of-the-art deterministic sketch (GK sketch).

Image segmentation process for high quality visual saliency map is very dependent on the existing visual saliency metrics. It is mostly only get sketchy effect of saliency map, and roughly based visual saliency map will affect the image segmentation results. The paper had presented the randomized visual saliency detection algorithm. The randomized visual saliency detection method can quickly generate the same size as the original input image and detailed results of the saliency map. The randomized saliency detection method can be applied to real-time requirements for image content-based scaling saliency results map. The randomization method for fast randomized video saliency area detection, the algorithm only requires a small amount of memory space can be detected detailed oriented visual saliency map, the presented results are shown that the method of visual saliency map used in image after the segmentation process can be an ideal segmentation results.

Measuring flow spread in real time from large, high-rate data streams has numerous practical applications, where a data stream is modeled as a sequence of data items from different flows and the spread of a flow is the number of distinct items in the flow. Past decades have witnessed tremendous performance improvement for single-flow spread estimation. However, when dealing with numerous flows in a data stream, it remains a significant challenge to measure per-flow spread accurately while reducing memory footprint. The goal of this paper is to introduce new multi-flow spread estimation designs that incur much smaller processing overhead and query overhead than the state of the art, yet achieves significant accuracy improvement in spread estimation. We formally analyze the performance of these new designs. We implement them in both hardware and software, and use real-world data traces to evaluate their performance in comparison with the state of the art. The experimental results show that our best sketch significantly improves over the best existing work in terms of estimation accuracy, data item processing throughput, and online query throughput.

CountSketch and Feature Hashing (the ``hashing trick'') are popular randomized dimensionality reduction methods that support recovery of l2 -heavy hitters and approximate inner products. When the inputs are not adaptive (do not depend on prior outputs), classic estimators applied to a sketch of size O(l / epsilon) are accurate for a number of queries that is exponential in l. When inputs are adaptive, however, an adversarial input can be constructed after O(l) queries with the classic estimator and the best known robust estimator only supports ~O(l^2) queries. In this work we show that this quadratic dependence is in a sense inherent: We design an attack that after O(l^2) queries produces an adversarial input vector whose sketch is highly biased. Our attack uses ``natural'' non-adaptive inputs (only the final adversarial input is chosen adaptively) and universally applies with any correct estimator, including one that is unknown to the attacker. In that, we expose inherent vulnerability of this fundamental method.

Background: Donald T. Campbell, a scientist, humanist, and generalist, left an indelible mark on the evaluation discipline through his methodological work. He is less well known within the evaluation community for his landmark contributions to biology and philosophy. Yet, the evolutionary epistemology that he pioneered has significant implications for evaluation. Purpose: This article examines the relevance of Donald T. Campbell’s blind variation and natural selection approach to evaluation theory, including an elucidation of its basic logic, its social remit as a discipline and trans-discipline, and its summative and formative functions. It also sketches in broad strokes the implications of evolutionary thinking for evaluation practice, including natural and artificial selection, ontogeny, phylogeny, co-evolution, and feedback. Finally, it comments on Campbell’s Experimenting Society vision and the ongoing craze for randomised evaluations in development through an evolutionary lens. Setting: Not applicable. Intervention: Not applicable. Research Design: Not applicable. Data Collection and Analysis: Not applicable. Findings: Not applicable.

Abstract We use randomized interest rates, offered across eighty geographically distinct regions for twenty-nine months by Mexico’s largest microlender, to sketch the adjustment from a price change to a new equilibrium. Demand is elastic, and more so over the longer run; e.g. the dollars-borrowed elasticity increases from $-$1.1 in Year one to $-$2.9 in Year three. Credit bureau data do not show evidence of crowd-out, although this and other null results are imprecisely estimated. The lender’s profits increase, albeit noisily, starting in Year two. But competitors do not respond by reducing rates. These findings, together with other results, suggest that informational frictions are important, and that cutting rates furthered Compartamos Banco’s “double bottom line” of improving social welfare subject to a profitability constraint.

Introduction: In children, acute otitis media (AOM) pain is undertreated. We sought to determine if video discharge instructions were associated with improved symptomatology, functional outcomes, and knowledge compared to a paper handout. Methods: We conducted a randomized controlled superiority trial comparing video discharge instructions (Easy Sketch Pro3TM) on management of pain to a paper handout detailing the same. We included caregivers of children 6 months to 5 years presenting to the emergency department (ED) with a clinical diagnosis of AOM. The primary outcome was symptomatology using the Acute Otitis Media Severity of Symptom (AOM SOS) score between 48 and 72 hours. The 7-item self-report AOM-SOS is scored from 0 to 13 with a higher score indicating more symptomatology. Secondary outcomes included knowledge gain using a 10-item survey, days of daycare/school/work missed, and recidivism. Assuming a minimal clinically important AOM-SOS difference of 2, 90% power, and 5% alpha, 60 individuals/group was needed. Results: 219 caregivers were randomized and 149 completed the 72-hour follow-up (72 paper and 77 video). The median (IQR) AOM-SOS score in the video group (adjusted for pre-intervention AOM-SOS, analgesic and antibiotic use) was significantly lower than paper [8 (7,11) versus 10 (7,13), respectively, p=0.004]. There were no significant differences between video and paper in the mean (SD) knowledge score [9.2 (1.3) versus 8.8 (1.8) correct answers, respectively, p=0.07], mean (SD) number that returned to a health provider [8/77 versus 10/72, respectively, p=0.49), mean (SD) number of daycare/schooldays missed [1.2 (1.5) versus 1.1 (2.1), respectively, p=0.62], and mean (SD) number of workdays missed by caregiver [0.5 (1) versus 0.8 (2), respectively, p=0.05]. Conclusion: Video discharge instructions are associated with less symptomatology compared to a paper handout, are effective for caregiver education in the ED, and should be used routinely.

IntroductionDrawing represents mainly a motor activity of expression. Drawing represents a form of non-verbal language that is very important both cognitively and affectively. Therefore, it allows to hypothesise and evaluate the degree of neurodevelopment of individuals as well as their level of interaction with the environment.Family sketches can be evaluated in a projective, neurocognitive and affective way to provide insights on the attachment system, degree of bonding, communication, social and affect interaction as well as difficulties or problems that have motivated emergency consultation.ObjectivesThis study evaluate the possible correlation between family dysfunction and irritability as cause of request of consultation in an emergency department of mental health in child & adolescents through the analysis of family drawings.MethodsThis is a retrospective, observational study of correlation between the reasons of emergency consultations, dysfunctional irritability and family difficulties represented through family drawing. It is based on a randomised sample of 30 reports of emergency appointments of children between 8 to 13 years old that have been examined in the Child & Adolescent Psychiatry Emergency Department at the Pitié Salpêtrière Hospital during two years for Emotional or Irritability dysfunction.An adaptation of both Goodenough-Harris Drawing projective test and Corman test were used to evaluate findings from family drawings as well as neurocognitive parameters of drawing technics, sociodemographic dates, cognitive level and family dysfunction.ResultsThe degree of cohesion, identification and devaluation of adult figures have been important elements of interpretation in irritability dysfunction and family drawings.ConclusionsThe family environment could be a factor in the interpretation of chronic irritability and its manifestations on the child’s family drawings establish a clear correlation.The adapted assessment of the family drawing could be an important tool in the nosological exploration of children’s mental health in emergency, especially on relational systemic representation and symbolization.Disclosure of InterestNone Declared

Introduction. Laparoscopic greater curvature plication is an operation that is gaining ground in the treatment of morbid obesity, as it appears to replicate the results of laparoscopic sleeve gastrectomy with fewer complications.Aim. Review of current literature, especially results on weight loss and complications.Method. 11 (eleven) published articles on laparoscopic gastric plication, of which 1 preclinical study, 8 prospective studies for a total of 521 patients and 2 case reports of unusual complications.Results. Reported Paracentage of EWL in all studies is comparable to Laparoscopic Sleeve Gastrectomy (around 50% in 6 months, 60–65% in 12 months, 60–65% in 24 months) and total complication rate is at 15,1% with minor complications in 10,7%, major complications in 4,4%. Reoperation rate was 3%, conversion rate was 0,2%, and mortality was zero.Conclusion. Current literature on gastric plication and its modifications is limited and sketchy at times. Low cost, short hospital stay, absence of prosthetic material, and reversibility make it an attractive option. Initial data show that LGCP is effective for short- and medium-term weight loss, complication and reoperation rates are low, and GERD symptoms are unaffected. More data is required, and randomized control trials must be completed in order to reach safe conclusions.

We propose, implement, and evaluate a hopping proof approach for proving expectation-based properties of probabilistic programs. Our approach combines EHL, a syntax-directed proof system for reducing proof goals of a program to proof goals of simpler programs, with a "hopping" proof rule for reducing proof goals of an original program to proof goal of a different program which is suitably related (by means of pRHL, a relational program logic for probabilistic program) to the original program. We prove that EHL is sound for a core language with procedure calls and adversarial computations, and complete for the adversary-free fragment of the language. We also provide an implementation of EHL into EasyCrypt, a proof assistant tailored for reasoning about relational properties of probabilistic programs. We provide a tight integration of EHL with other program logics supported by EasyCrypt, and in particular probabilistic Relational Hoare Logic (pRHL). Using this tight integration, we give mechanized proofs of expected complexity of in-place implementations of randomized quickselect and skip lists. We also sketch applications of our approach to cryptographic proofs and discuss the broader impact of EHL in the EasyCrypt proof assistant.

Sustainability in irrigation is an essential step towards responsible water consumption. In recent years, many studies have sketched climate-resilient agricultural practices to fight drought and uncertain rainfall patterns. Major rain-fed crops such as paddy and wheat require aid when there are abnormal dry spells. To mitigate the loss of crops from such events, superabsorbent polymers can be used. Soils amended with hydrogel or Superabsorbent polymer (SAP) retain moisture during drought to prevent loss of water through evaporation and percolation. This allows the crop to grow with less shock from drought. This study compares rice (Oryza sativa L.) growth rate under application (treatment groups) and non-application (control groups) of hydrogel, considering their high-water requirement. NDLR07 (recently developed) and BPT5204 (local variety) rice varieties were chosen for the current study. Randomized controlled trials were performed for each variety on a control group (NC & BC) and 3 treatment groups with 20% (NT20 & BT20), 40% (NT40 & BT40), and 60% (NT60 & BT60) deficit water supplies respectively. N, T, C refers to seed type, treatment group, control group respectively. Intermittent drought condition was imposed for 14 days to assess the resilience of crops. The water retention capacity of the sandy loam soil was better for treatment groups by 20% than control groups even at an average temperature of 40 ℃. Treatment groups continued growing through the drought phase and after, while control groups showed stagnation. Among the tested treatment groups, NT20 had the highest growth among all trials. The results of the study suggested that hydrogel application can help to combat droughts and thereby contribute to sustainable agricultural production by restricting the involvement of climate changes.

Firefighter clothing is known to restrict movement and increase the risk of musculoskeletal injury. The aim of this study was to investigate differences between clothing variations and firefighter mobility. Data were collected from eight firefighters using a randomized counterbalanced, repeated measures, design. Three different clothing variants (V1–V3) in addition to current station wear (S) were trialed. Combinations of S and V1 (SV1) and V2 (SV2) were also trialed. Outcome measures included: standing reach height; the Functional Movement Screen [FMS]; vertical jump; a visual analogue scale [VAS] for the FMS, vertical jump, step ups and crawl; and a mannequin sketch to mark areas of discomfort. V3 was preferred over S in all tasks and performed better both objectively (FMS and vertical jump) and subjectively (VAS scores). While V1 was typically associated with poorer performance, the impacts of V1 and V2 when compared to each other and S varied depending on the tasks performed. This variation was greater when V1 or V2 were worn with S (SV1 and SV2). Areas of discomfort across all variations was the knees, followed by the thighs. Clothing that has the least impact on, or improves, mobility (especially around the lower limbs) may help reduce firefighter injury and limit subjective discomfort.

Abstract BACKGROUND Pain, the most identifiable feature of acute otitis media (AOM), is undertreated with 30% of children ≤ 2 years experiencing pain, fever, or both for up to 7 days, highlighting the importance of effective caregiver education. OBJECTIVES We sought to determine if video discharge instructions were associated with improved symptomatology, functional outcomes, and knowledge compared to a paper handout. DESIGN/METHODS We conducted a randomized controlled superiority trial comparing video discharge instructions (Easy Sketch Pro3TM) on management of pain and fever to a paper handout detailing the same. We included primary caregivers of children 6 months to 5 years presenting to the emergency department (ED) with a clinical diagnosis of AOM. The primary outcome was symptomatology using the Acute Otitis Media Severity of Symptom (AOM SOS) score between 48 and 72 hours. The 7-item self-report AOM-SOS is scored from 0 to 13 with a higher score indicating more symptomatology. Secondary outcomes included knowledge gain using a 10-item survey, days of daycare/school/work missed, and recidivism. RESULTS A total of 219 caregivers were randomized and 149 completed the 72-hour follow-up (72 paper and 77 video). Participants were primarily mothers (175/219, 79.9%); 136/219 (62.1%) completed post-secondary education and 147/219 (67.1%) had previously cared for a child with AOM. Children included 107/219 (48.6%) females with an overall mean (SD) age of 2.9 (2.8) years. Caregivers did not offer analgesia to 41/219 (18.7%) of children. The median (IQR) AOM-SOS score in the video group was significantly lower than the paper group, even after adjusting for pre-intervention AOM-SOS and medication (analgesics and antibiotics) given by caregivers [8 (7,11) versus 10 (7,13), respectively, p = 0.004]. There were no significant differences between video and paper in the mean (SD) number of correct answers given on the post-intervention survey [9.2 (1.3) versus 8.8 (1.8), respectively, p = 0.07], mean (SD) number of children that returned to a health provider [8/77 versus 10/72, respectively, p = 0.49), mean (SD) number of daycare/school missed by child [1.2 (1.5) versus 1.1 (2.1), respectively, p = 0.62, mean (SD) number of work missed by caregiver [0.5 (1) versus 0.8 (2), respectively, p = 0.05]. CONCLUSION Children of caregivers with AOM who received a five-minute video detailing the identification and management of pain and fever experienced less symptomatology compared to a paper handout. Our findings suggest that video discharge instructions in the ED are effective for caregiver education and should be used routinely.

A scientific paradigm consists of a set of shared rules, beliefs, values, methods, and instruments for addressing scientific problems. Currently, health care embraces the paradigm of evidence-based health care (EBH). This paradigm prompts health care institutions to base decisions on the best available evidence, which is commonly generated in large-scale randomized controlled trials. We illustrate the application of EBH via the evaluation of drugs. We show how EBH is challenged when it is applied to the evaluation of digital therapeutics, which refers to technology and data to prevent, manage, or treat a medical disorder or disease. We conclude that amid the growing application of digital therapeutics, the paradigm of EBH is challenged in four domains: population, intervention, comparison, outcome. In the second part of this viewpoint, we argue for a paradigm shift in health care so we can optimally evaluate and implement digital therapeutics, and we sketch out the contours of this novel paradigm. We address the need for considering design in health care and evaluation processes, studying user values so that health care can move from a focus on health to well-being, focusing on individual experiences rather than the average, addressing the need for evaluation in authentic use contexts, and stressing the need for continuous evaluation of the dynamic relations between users, context, and digital therapeutics. We conclude that the transition from EBH toward evidence-based well-being would improve the successful implementation of digital technologies in health care.

Chinese bamboo has advantages like other bamboos, it has many roots, produces dense clumps, is able to prevent soil erosion and is a source of oxygen. This bamboo plant can be used as handicrafts, one of which is as a sketch and accessories. The purpose of this study was to analyze the effect of giving the best Ecofert biofertilizer on the growth of Chinese bamboo with the parameters of survival percentage, shoot growth and shoot height increase. This research method was carried out using a completely randomized design (CRD) which consisted of 5 different treatments, namely 1) treatment A without the application of Ecofert biofertilizer; 2) treatment B application of 10 grams of Ecofert fertilizer; 3) treatment C with 20 grams of Ecofert fertilizer; 4) treatment D application of 30 grams of Ecofert fertilizer; 4) treatment E was the application of 40 grams of Ecofert fertilizer. From each treatment, the observation of the live percentage got a value ranging from 65% to 85%, the highest result at the end of the observation was the average increase in the number of shoots in treatment E (40 grams) with the results of getting 3.12 shoots and at the average increase in shoot height the results the highest treatment B (10 grams) with the result 159.92 cmBambu cina memiliki kelebihan seperti bambu lainnya memiliki perakaran yang banyak, menghasilkan rumpun yang rapat, mampu mencegah erosi tanah dan penghasil sumber oksigen. Tanaman bambu ini dapat dijadikan kerajinan tangan salah satunya sebagai sketsel dan aksesoris. Tujuan penelitian ini agar dapat menganalisis pengaruh pemberian pupuk hayati Ecofert terbaik terhadap pertumbuhan bambu cina dengan parameter persentase hidup, pertambahan tunas dan pertambahan tinggi tunas. Metode penelitian ini dilakukan dengan Rancangan Acak Lengkap (RAL) yang terdiri dari 5 perlakuan berbeda yaitu 1) perlakuan A tanpa pemberian pupuk hayati Ecofert; 2) perlakuan B pemberian pupuk Ecofert 10 gram; 3) perlakuan C pemberian pupuk Ecofert 20 gram; 4) perlakuan D pemberian pupuk Ecofert 30 gram; 4) perlakuan E pemberian pupuk Ecofert 40 gram. Dari setiap perlakuan pengamatan dari persentasi hidup mendapatkan nilai kisaran 65% sampai 85%, hasil tertinggi diakhir pengamatan dari rata-rata pertambahan jumlah tunas pada perlakuan E (40 gram) dengan hasil mendapatkan 3,12 tunas dan pada pertambahan rata-rata tinggi tunas hasil tertinggi perlakuan B (10 gram) dengan hasil 159,92 cm.

Estimating ranks, quantiles, and distributions over streaming data is a central task in data analysis and monitoring. Given a stream of n items from a data universe equipped with a total order, the task is to compute a sketch (data structure) of size polylogarithmic in n . Given the sketch and a query item y , one should be able to approximate its rank in the stream, i.e., the number of stream elements smaller than or equal to y . Most works to date focused on additive ε n error approximation, culminating in the KLL sketch that achieved optimal asymptotic behavior. This paper investigates multiplicative (1 ± ε)-error approximations to the rank. Practical motivation for multiplicative error stems from demands to understand the tails of distributions, and hence for sketches to be more accurate near extreme values. The most space-efficient algorithms due to prior work store either O (log (ε 2 n )/ε 2 ) or O (log 3 (ε n )/ε) universe items. We present a randomized sketch storing O (log 1.5 (ε n )/ε) items that can (1 ± ε)-approximate the rank of each universe item with high constant probability; this space bound is within an \(O(\sqrt {\log (\varepsilon n)}) \) factor of optimal. Our algorithm does not require prior knowledge of the stream length and is fully mergeable, rendering it suitable for parallel and distributed computing environments.

Abstract The sketch-and-project, as a general archetypal algorithm for solving linear systems, unifies a variety of randomized iterative methods such as the randomized Kaczmarz and randomized coordinate descent. However, since it aims to find a least-norm solution from a linear system, the randomized sparse Kaczmarz can not be included. This motivates us to propose a more general framework, called sketched Bregman projection (SBP) method, in which we are able to find solutions with certain structures from linear systems. To generalize the concept of adaptive sampling to the SBP method, we show how the progress, measured by Bregman distance, of single step depends directly on a sketched loss function. Theoretically, we provide detailed global convergence results for the SBP method with different adaptive sampling rules. At last, for the (sparse) Kaczmarz methods, a group of numerical simulations are tested, with which we verify that the methods utilizing sampling Kaczmarz-Motzkin rule demands the fewest computational costs to achieve a given error bound comparing to the corresponding methods with other sampling rules.

Abstract In this paper we consider large-scale smooth optimization problems with multiple linear coupled constraints. Due to the non-separability of the constraints, arbitrary random sketching would not be guaranteed to work. Thus, we first investigate necessary and sufficient conditions for the sketch sampling to have well-defined algorithms. Based on these sampling conditions we develop new sketch descent methods for solving general smooth linearly constrained problems, in particular, random sketch descent (RSD) and accelerated random sketch descent (A-RSD) methods. To our knowledge, this is the first convergence analysis of RSD algorithms for optimization problems with multiple non-separable linear constraints. For the general case, when the objective function is smooth and non-convex, we prove for the non-accelerated variant sublinear rate in expectation for an appropriate optimality measure. In the smooth convex case, we derive for both algorithms, non-accelerated and A-RSD, sublinear convergence rates in the expected values of the objective function. Additionally, if the objective function satisfies a strong convexity type condition, both algorithms converge linearly in expectation. In special cases, where complexity bounds are known for some particular sketching algorithms, such as coordinate descent methods for optimization problems with a single linear coupled constraint, our theory recovers the best known bounds. Finally, we present several numerical examples to illustrate the performances of our new algorithms.

With the increase in collected data volumes, either from experimental measurements or high fidelity simulations, there is an ever-growing need to develop computationally efficient tools to process, analyze, and interpret these data sets. Modal analysis techniques have gained great interest due to their ability to identify patterns in the data and extract valuable information about the system being considered. Dynamic mode decomposition (DMD) relies on elements of Koopman approximation theory to compute a set of modes, each associated with a fixed oscillation frequency and decay/growth rate. Extracting these details from large data sets can be computationally expensive due to the need to implement singular value decomposition of the input data matrix. Sketching algorithms have become popular in numerical linear algebra where statistical theoretic approaches are utilized to reduce the cost of major operations. A sketch of a matrix is another matrix which is significantly smaller, but still sufficiently approximates the original system. We put forth an efficient DMD framework, SketchyDMD, based on a core sketching algorithm that captures information about the range and co-range (an their mutual relationship) of input data. The proposed sketching-based framework can accelerate various portions of the DMD routines, compared to classical methods that operate directly on the raw input data. We conduct numerical experiments using the spherical shallow water equations as a prototypical model in the context of geophysical flows. We show that the proposed SketchyDMD is superior to existing randomized DMD methods that are based on capturing only the range of the input data.

Abstract Low-rank matrix approximation is extremely useful in the analysis of data that arises in scientific computing, engineering applications, and data science. However, as data sizes grow, traditional low-rank matrix approximation methods, such as singular value decomposition (SVD) and column pivoting QR decomposition (CPQR), are either prohibitively expensive or cannot provide sufficiently accurate results. A solution is to use randomized low-rank matrix approximation methods such as randomized SVD, and randomized LU decomposition on extremely large data sets. In this paper, we focus on the randomized LU decomposition method. Then we propose a novel randomized LU algorithm, called SubspaceLU, for the fixed low-rank approximation problem. SubspaceLU is based on the sketch of the co-range of input matrices and allows for an arbitrary number of passes of the input matrix, v ≥ 2 {v\geq 2} . Numerical experiments on CPU show that our proposed SubspaceLU is generally faster than the existing randomized LU decomposition, while remaining accurate. Experiments on GPU shows that our proposed SubspaceLU can gain more speedup than the existing randomized LU decomposition. We also propose a version of SubspaceLU, called SubspaceLU_FP, for the fixed precision low-rank matrix approximation problem. SubspaceLU_FP is a post-processing step based on an efficient blocked adaptive rank determination Algorithm 5 proposed in this paper. We present numerical experiments that show that SubspaceLU_FP can achieve close results to SVD but faster in speed. We finally propose a single-pass algorithm based on LU factorization. Tests show that the accuracy of our single-pass algorithm is comparable with the existing single-pass algorithms.

This one-off, largely exploratory study seeks to answer questions that map several basic phenomena of online sociability, in order to be able to have a basis for a possible analysis of the external cultural reception of an entity formed in a theoretically different culture – the Romanian culture. By going through the empirical inventory of some common typologies of online gamers, by means of randomized, anonymous interviews with players, our final goal (and hope) was to sketch a relationship between social capital and group configuration, within an internal community of the online game (MMORPG) World of Warcraft.

AbstractPhotorealistic 3D models (PR3DM) have great potential to supplement anatomy education; however, there is evidence that realism can increase cognitive load and negatively impact anatomy learning, particularly in students with decreased spatial ability. These differing viewpoints have resulted in difficulty in incorporating PR3DM when designing anatomy courses. To determine the effects of spatial ability on anatomy learning and reported intrinsic cognitive load using a drawing assessment, and of PR3DM versus an Artistic colour-coded 3D model (A3DM) on extraneous cognitive load and learning performance. First-year medical students participated in a cross-sectional (Study 1) and a double-blind randomised control trial (Study 2). Pre-tests analysed participants' knowledge of anatomy of the heart (Study 1, N = 50) and liver (Study 2, N = 46). In Study 1, subjects were first divided equally using a mental rotations test (MRT) into low and high spatial ability groups. Participants memorised a 2D-labeled heart valve diagram and sketched it rotated 180°, before self-reporting their intrinsic cognitive load (ICL). For Study 2, participants studied a liver PR3DM or its corresponding A3DM with texture-homogenisation, followed by a liver anatomy post-test, and reported extraneous cognitive load (ECL). All participants reported no prior anatomy experience. Participants with low spatial ability (N = 25) had significantly lower heart drawing scores (p = 0.001) than those with high spatial ability (N = 25), despite no significant differences in reported ICL (p = 0.110). Males had significantly higher MRT scores than females (p = 0.011). Participants who studied the liver A3DM (N = 22) had significantly higher post-test scores than those who studied the liver PR3DM (N = 24) (p = 0.042), despite no significant differences in reported ECL (p = 0.720). This investigation demonstrated that increased spatial ability and colour-coding of 3D models are associated with improved anatomy performance without significant increase in cognitive load. The findings are important and provide useful insight into the influence of spatial ability and photorealistic and artistic 3D models on anatomy education, and their applicability to instructional and assessment design in anatomy.

AbstractIt’s a challenge for detecting the therapeutic targets of a polypharmacological drug from variations in the responsed networks in the differentiated populations with complex diseases, as stable coronary heart disease. Here, in an adaptive, 31-center, randomized, double-blind trial involving 920 patients with moderate symptomatic stable angina treated by 14-day Danhong injection(DHI), a kind of polypharmacological drug with high quality control, or placebo (0.9% saline), with 76-day following-up, we firstly confirmed that DHI could increase the proportion of patients with clinically significant changes on angina-frequency assessed by Seattle Angina Questionnaire (ΔSAQ-AF ≥ 20) (12.78% at Day 30, 95% confidence interval [CI] 5.86–19.71%, P = 0.0003, 13.82% at Day 60, 95% CI 6.82–20.82%, P = 0.0001 and 8.95% at Day 90, 95% CI 2.06–15.85%, P = 0.01). We also found that there were no significant differences in new-onset major vascular events (P = 0.8502) and serious adverse events (P = 0.9105) between DHI and placebo. After performing the RNA sequencing in 62 selected patients, we developed a systemic modular approach to identify differentially expressed modules (DEMs) of DHI with the Zsummary value less than 0 compared with the control group, calculated by weighted gene co-expression network analysis (WGCNA), and sketched out the basic framework on a modular map with 25 functional modules targeted by DHI. Furthermore, the effective therapeutic module (ETM), defined as the highest correlation value with the phenotype alteration (ΔSAQ-AF, the change in SAQ-AF at Day 30 from baseline) calculated by WGCNA, was identified in the population with the best effect (ΔSAQ-AF ≥ 40), which is related to anticoagulation and regulation of cholesterol metabolism. We assessed the modular flexibility of this ETM using the global topological D value based on Euclidean distance, which is correlated with phenotype alteration (r2: 0.8204, P = 0.019) by linear regression. Our study identified the anti-angina therapeutic module in the effective population treated by the multi-target drug. Modular methods facilitate the discovery of network pharmacological mechanisms and the advancement of precision medicine. (ClinicalTrials.gov identifier: NCT01681316).

Abstract Background Primary aim of the current randomized controlled trial was to test the effectiveness of the parenting intervention ‘Video-feedback to promote Positive Parenting and Sensitive Discipline’ (VIPP-SD) in a sample of parents of preschool-aged twins, as well as differential susceptibility to intervention efforts, that is, whether more temperamentally reactive parents would profit more from the VIPP-SD than parents with lower reactivity. Methods The sample consisted of 202 families with same-sex twins [N = 404 children, mean age 45 months (SD = 6.81)]. Randomization was done at the family level in a 2:3 ratio, with 83 families (41%) randomized to the VIPP-SD group, and 119 families (59%) to the control group. After two pre-tests in year 1 and year 2 of the study, the VIPP-SD was implemented in the third year, with a post-test assessment 1 month after the five intervention sessions. Parental sensitivity was observed during structured play in which parent and child copied a drawing together in a computerized Etch-A-Sketch paradigm. Parental limit-setting was observed in a ‘don’t touch’ task in which the parent required from the child to abstain from playing with attractive toys. Parents interacted with each of their twins in separate sessions. Results The VIPP-SD intervention had a positive impact on the level of parents’ positive limit-setting in interaction with their preschool twins, and this positive effect was most pronounced when the parents completed at least five intervention sessions. However, the intervention did not enhance parental sensitivity during structured play. Parents with higher reactivity were not more open to the impact of the intervention, thus for this temperamental marker differential susceptibility in adults was not supported. Conclusions The current study is unique in targeting families with twin preschoolers, providing proof of principle that coaching parents with video-feedback promotes parental sensitive limit-setting to both children. It remains to be seen whether this finding can be replicated in families with non-twin siblings, or other parental susceptibility markers. Trial registration Trial NL5172 (NTR5312), 2015-07-20.

Suppose that party A collects private information about its users, where each user's data is represented as a bit vector. Suppose that party B has a proprietary data mining algorithm that requires estimating the distance between users, such as clustering or nearest neighbors. We ask if it is possible for party A to publish some information about each user so that B can estimate the distance between users without being able to infer any private bit of a user. Our method involves projecting each user's representation into a random, lower-dimensional space via a sparse Johnson-Lindenstrauss transform and then adding Gaussian noise to each entry of the lower-dimensional representation. We show that the method preserves differential privacy---where the more privacy is desired, the larger the variance of the Gaussian noise. Further, we show how to approximate the true distances between users via only the lower-dimensional, perturbed data. Finally, we consider other perturbation methods such as randomized response and draw comparisons to sketch-based methods. While the goal of releasing user-specific data to third parties is more broad than preserving distances, this work shows that distance computations with privacy is an achievable goal.

The use of blood‐derived products (Eye Drop of Human Origin—EDHO; in Italian: Emocomponents for Not Transfusional Use, EUNT) for the treatment of various ocular surface diseases has become increasingly utilized in ophthalmic practice over the years. In particular, they are recommended as a third line option, when conventional tear substitutes, or active drugs, such as corticosteroids or Cyclosporine A, result unsuccessful in controlling either signs and, perhaps most importantly, subjective symptoms of discomfort and pain. The rationale for their use is based on the stimulation of cell proliferation and migration through the supply of metabolically active substances, in particular growth factors, and definitely mimicking the function of natural tears.EDHO have been used for the treatment of various ocular surface diseases, including dry eye, persistent corneal epithelial defect, corneal ulcer, neurotrophic keratitis, ocular surface burn, recurrent corneal erosion and limbal stem cell deficiency. However, the therapeutic potential of these products is still to be fully explored, as randomized clinical trials and internationally recognized harmonized guidelines are still needed to provide better evidence, improve quality of the final products and lead to a more widespread use of these therapies in daily clinical practice. The autologous source is still quite popular, although allogeneic sources appear to be a promising innovation. These might offer the solution of obtaining in advance products tailored for each patient and ocular surface disease, at scalable concentrations based on the severity of the disease and clinical parameters. There remain many unanswered questions regarding what might be considered optimal treatment, sketched as “The 5 Ws (and 2 H) for blood based eye drops”, to quote a review from our group (doi: 10.3390/jcm8091478). This presentation will discuss: 1. who is the patient to be treated, in terms of disease type, severity, and stage?; 2. Why is an EDHO needed, in terms of a target indication?; 3. When is it appropriate to prescribe an EDHO therapy?; 4. Where are EDHO be dispensed, and is a national/regional program a feasible solution to optimize resources?; 5. What is the EDHO of choice, which source and preparation are targeted for a given patient? And further: 1. how is EDHO standardized in terms of processing to ensure an optimal dilution, solvent, dispenser, storage time?; 2. How is EDHO delivered to the ocular surface, in terms of form, dose‐size modulation, length of treatment?An interdisciplinary work with transfusional medicine specialists is imperative to share information and competences, to achieve a better clinical management.

This review provides a rough sketch of the evidence, gaps and promising directions in e-learning from 2000 onwards, with a particular focus on Canada. We searched a wide range of sources and document types to ensure that we represented, comprehensively, the arguments surrounding e-learning. Overall, there were 2,042 entries in our database, of which we reviewed 1,146, including all the Canadian primary research and all scholarly reviews of the literature. In total, there were 726 documents included in our review: 235 – general public opinion; 131 – trade/practitioners’ opinion; 88 – policy documents; 120 – reviews; and 152 – primary empirical research. The Argument Catalogue codebook included the following eleven classes of variables: 1) Document Source; 2) Areas/Themes of e-learning; 3) Value/Impact; 4) Type of evidence; 5) Research design; 6) Area of applicability; 7) Pedagogical implementation factors; 8) A-priori attitudes; 9) Types of learners; 10) Context; and 11) Technology Factors. We examined the data from a number of perspectives, including their quality as evidence. In the primary research literature, we examined the kinds of research designs that were used. We found that over half of the studies conducted in Canada are qualitative in nature, while the rest are split in half between surveys and quantitative studies (correlational and experimental). When we looked at the nature of the research designs, we found that 51% are qualitative case studies and 15.8% are experimental or quasi-experimental studies. It seems that studies that can help us understand “what works” in e-learning settings are underrepresented in the Canadian research literature. The documents were coded to provide data on outcomes of e-learning (we also refer to them as “impacts” of e-learning). Outcomes/impacts are the perceived or measured benefits of e-learning, whereas predictors are the conditions or features of e-learning that can potentially affect the outcomes/impacts. The impacts were coded on a positive to negative scale and included: 1) achievement; 2) motivation/satisfaction; 3) interactivity/ communication; 4) meeting social demands; 5) retention/attrition; 6) learning flexibility; and 7) cost. Based on an analysis of the correlations among these impacts, we subsequently collapsed them (all but cost) into a single impact scale ranging from –1 to +1. We found, generally, that the perception of impact or actual measured impact varies across the types of documents. They appear to be lower in general opinion documents, practitioner documents and policy making reports than in scholarly reviews and primary research. While this represents an expression of hope for positive impact, on the one hand, it possibly represents reality, on the other. Where there were sufficient documents to examine and code, impact was high across each of the CCL Theme Areas. Health and Learning was the highest, with a mean of 0.80 and Elementary/Secondary was the lowest, with a mean of 0.77. However, there was no significant difference between these means. The impact of e-learning and technology use was highest in distance education, where its presence is required (Mean = 0.80) and lowest in face-to-face instructional settings (Mean = 0.60) where its presence is not required. Network-based technologies (e.g., Internet, Web-based, CMC) produced a higher impact score (Mean = 0.72) than straight technology integration in educational settings (Mean = 0.66), although this difference was considered negligible. Interestingly, among the Pedagogical Uses of Technology, student applications (i.e., students using technology) and communication applications (both Mean = 0.78) had a higher impact score than instructional or informative uses (Mean = 0.63). This result suggests that the student manipulation of technology in achieving educational goals is preferable to teacher manipulation of technology. In terms of predictor variables (professional training, course design, infrastructure/ logistics, type of learners [general population, special needs, gifted], gender issues and ethnicity/race/religion/aboriginal, location, school setting, context of technology use, type of tool used and pedagogical function of technology) we found the following: professional development was underrepresented compared to issues of course design and infrastructure/ logistics; most attention is devoted to general population students, with little representation of special needs, the gifted students, issues of gender or ethnic/race/religious/aboriginal status; the greatest attention is paid to technology use in distance education and the least attention paid to the newly emerging area of hybrid/blended learning; the most attention is paid to networked technologies such as the Internet, the WWW and CMC and the least paid to virtual reality and simulations. Using technology for instruction and using technology for communication are the two highest categories of pedagogical use. In the final stage, the primary e-learning studies from the Canadian context that could be summarized quantitatively were identified. We examined 152 studies and found a total of 7 that were truly experimental (i.e., random assignment with treatment and control groups) and 10 that were quasi-experimental (i.e., not randomized but possessing a pretest and a posttest). For these studies we extracted 29 effect sizes or standardized mean differences, which were included in the composite measure. The mean effect size was +0.117, a small positive effect. Approximately 54% of the e-learning participants performed at or above the mean of the control participants (50 th percentile), an advantage of 4%. However, the heterogeneity analysis was significant, indicating that the effect sizes were widely dispersed. It is clearly not the case that e-learning is always the superior condition for educational impact. Overall, we know that research in e-learning has not been a Canadian priority; the culture of educational technology research, as distinct from development, has not taken on great import. In addition, there appears to have been a disproportionate emphasis on qualitative research in the Canadian e-learning research culture. We noted that there are gaps in areas of research related to early childhood education and adult education. Finally, we believe that more emphasis must be placed on implementing longitudinal research, whether qualitative or quantitative (preferably a mixture of the two), and that all development efforts be accompanied by strong evaluation components that focus on learning impact. It is a shame to attempt innovation and not be able to tell why it works or doesn’t work. In this sense, the finest laboratories for e-learning research are the institutions in which it is being applied. Implications for K-12 Practitioners When implemented appropriately, technology tools are beneficial to students’ learning, and may facilitate the development of higher order thinking skills. Student manipulation of technology in achieving the goals of education is preferable to teacher manipulation of technology. Teachers need to be aware of differences between instructional design for e-learning as compared to traditional face-to-face situations. Immediate, extensive, and sustained support should be offered to teachers in order to make the best out of e-learning. Implications for Post-Secondary Some educators suggest that e-learning has the potential to transform learning, but there is limited empirical research to assess the benefits. Post-secondary education would benefit from a Pan-Canadian plan to assess the impact of e-learning initiatives. It is important that instructional design match the goals and potential of e-learning. Research is needed to determine the feasibility and effectiveness of such things as learning objects and multimedia applications. Properly implemented computer mediated communication can enrich the learning environment; help reduce low motivation and feelings of isolation in distance learners. E-learning appears to be more effective in distance education, where technology use is required than in face-to-face instructional settings. Implications for Policy Makers Effective and efficient implementation of e-learning technologies represents new, and difficult, challenges to practitioners, researchers, and policymakers. The term e-learning has been used to describe many different applications of technology, which may be implemented in a wide variety of ways (some of which are much more beneficial than others). School administrators must balance the needs of all stakeholders, and the cost-benefit ratios of technology tools, in deciding not only which technologies to use, but also when and how to implement new technologies. Traditional methods of instructional design and school administration must be adjusted to deal with the demands of distance education and other contexts of technology use. Professional education, development, and training for educators must ensure that teachers will be equipped to make optimal pedagogical use of new methods.

Introduction Medical misinformation and conspiracies have thrived during the current infodemic as a result of the volume of information people have been exposed to during the disease outbreak. Given that SARS-CoV-2 (COVID-19) is a novel coronavirus discovered in 2019, much remains unknown about the disease. Moreover, a considerable amount of what was originally thought to be known has turned out to be inaccurate, incomplete, or based on an obsolete knowledge of the virus. It is in this context of uncertainty and confusion that conspiracies flourish. Michael Golebiewski and danah boyd’s work on ‘data voids’ highlights the ways that actors can work quickly to produce conspiratorial content to fill a void. The data void absent of high-quality data surrounding COVID-19 provides a fertile information environment for conspiracies to prosper (Chou et al.). Conspiracism is the belief that society and social institutions are secretly controlled by a powerful group of corrupt elites (Douglas et al.). Michael Barkun’s typology of conspiracy reveals three components: 1) the belief that nothing happens by accident or coincidence; 2) nothing is as it seems: the "appearance of innocence" is to be suspected; 3) the belief that everything is connected through a hidden pattern. At the heart of conspiracy theories is narrative storytelling, in particular plots involving influential elites secretly colluding to control society (Fenster). Conspiracies following this narrative playbook have flourished during the pandemic. Pharmaceutical corporations profiting from national vaccine rollouts, and the emergency powers given to governments around the world to curb the spread of coronavirus, have led some to cast these powerful commercial and State organisations as nefarious actors – 'big evil' drug companies and the ‘Deep State’ – in conspiratorial narratives. Several drugs believed to be potential treatments for COVID-19 have become entangled with conspiracy. At the start of the pandemic scientists experimented with repurposing existing drugs as potential treatments for COVID-19 because safe and effective vaccines were not yet available. A series of antimicrobials with potential activity against SARS-CoV-2 were tested in clinical trials, including lopinavir/ritonavir, favipiravir and remdesivir (Smith et al.). Only hydroxychloroquine and ivermectin transformed from potential COVID treatments into conspiracy objects. This article traces how the hydroxychloroquine and ivermectin conspiracy theories were amplified in the news media and online. It highlights how debunking processes contribute to amplification effects due to audience segmentation in the current media ecology. We conceive of these amplification and debunking processes as key components of a ‘Conspiracy Course’ (Baker and Maddox), identifying the interrelations and tensions between amplification and debunking practices as a conspiracy develops, particularly through mainstream news, social media and alternative media spaces. We do this in order to understand how medical claims about potential treatments for COVID-19 succumb to conspiracism and how we can intervene in their development and dissemination. In this article we present a commentary on how public discourse and actors surrounding two potential treatments for COVID-19: the anti-malarial drug hydroxychloroquine and the anti-parasitic drug ivermectin became embroiled in conspiracy. We examine public discourse and events surrounding these treatments over a 24-month period from January 2020, when the virus gained global attention, to January 2022, the time this article was submitted. Our analysis is contextually informed by an extended digital ethnography into medical misinformation, which has included social media monitoring and observational digital field work of social media sites, news media, and digital media such as blogs, podcasts, and newsletters. Our analysis focusses on the role that public figures and influencers play in amplifying these conspiracies, as well as their amplification by some wellness influencers, referred to as “alt.health influencers” (Baker), and those affiliated with the Intellectual Dark Web, many of whom occupy status in alternative media spaces. The Intellectual Dark Web (IDW) is a term used to describe an alternative influence network comprised of public intellectuals including the Canadian psychologist Jordan Peterson and the British political commentator Douglas Murray. The term was coined by the American mathematician and podcast host Eric Weinstein, who described the IDW as a group opposed to “the gated institutional narrative” of the mainstream media and the political establishment (Kelsey). As a consequence, many associated with the IDW use alternative media, including podcasts and newsletters, as an "eclectic conversational space" where those intellectual thinkers excluded from mainstream conversational spaces in media, politics, and academia can “have a much easier time talking amongst ourselves” (Kelsey). In his analysis of the IDW, Parks describes these figures as "organic" intellectuals who build identification with their audiences by branding themselves as "reasonable thinkers" and reinforcing dominant narratives of polarisation. Hence, while these influential figures are influencers in so far as they cultivate an online audience as a vocation in exchange for social, economic and political gain, they are distinct from earlier forms of micro-celebrity (Senft; Marwick) in that they do not merely achieve fame on social media among a niche community of followers, but appeal to those disillusioned with the mainstream media and politics. The IDW are contrasted not with mainstream celebrities, as is the case with earlier forms of micro-celebrity (Abidin Internet Celebrity), but with the mainstream media and politics. A public figure, on the other hand, is a “famous person” broadcast in the media. While celebrities are public figures, public figures are not necessarily celebrities; a public figure is ‘a person of great public interest or familiarity’, such as a government official, politician, entrepreneur, celebrity, or athlete. Analysis In what follows we explore the role of influencers and public figures in amplifying the hydroxychloroquine and ivermectin conspiracy theories during the pandemic. As part of this analysis, we consider how debunking processes can further amplify these conspiracies, raising important questions about how to most effectively respond to conspiracies in the current media ecology. Discussions around hydroxychloroquine and ivermectin as potential treatments for COVID-19 emerged in early 2020 at the start of the pandemic when people were desperate for a cure, and safe and effective vaccines for the virus were not yet publicly available. While claims concerning the promising effects of both treatments emerged in the mainstream, the drugs remained experimental COVID treatments and had not yet received widespread acceptance among scientific and medical professionals. Much of the hype around these drugs as COVID “cures” emerged from preprints not yet subject to peer review and scientific studies based on unreliable data, which were retracted due to quality issues (Mehra et al.). Public figures, influencers, and news media organisations played a key role in amplifying these narratives in the mainstream, thereby extending the audience reach of these claims. However, their transformation into conspiracy objects followed different amplification processes for each drug. Hydroxychloroquine, the “Game Changer” Hydroxychloroquine gained public attention on 17 March 2020 when the US tech entrepreneur Elon Musk shared a Google Doc with his 40 million followers on Twitter, proposing “maybe worth considering chloroquine for C19”. Musk’s tweet was liked over 50,200 times and received more than 13,500 retweets. The tweet was followed by several other tweets that day in which Musk shared a series of graphs and a paper alluding to the “potential benefit” of hydroxychloroquine in in vitro and early clinical data. Although Musk is not a medical expert, he is a public figure with status and large online following, which contributed to the hype around hydroxychloroquine as a potential treatment for COVID-19. Following Musk’s comments, search interest in chloroquine soared and mainstream media outlets covered his apparent endorsement of the drug. On 19 March 2020, the Fox News programme Tucker Carlson Tonight cited a study declaring hydroxychloroquine to have a “100% cure rate against coronavirus” (Gautret et al.). Within hours another public figure, the then-US President Donald Trump, announced at a White House Coronavirus Task Force briefing that the FDA would fast-track approval of hydroxychloroquine, a drug used to treat malaria and arthritis, which he said had, “tremendous promise based on the results and other tests”. Despite the Chief Medical Advisor to the President, Dr Anthony Fauci, disputing claims concerning the efficacy of hydroxychloroquine as a potential therapy for coronavirus as “anecdotal evidence”, Trump continued to endorse hydroxychloroquine describing the drug as a “game changer”: HYDROXYCHLOROQUINE & AZITHROMYCIN, taken together, have a real chance to be one of the biggest game changers in the history of medicine. He said that the drugs should be put in use IMMEDIATELY. PEOPLE ARE DYING, MOVE FAST, and GOD BLESS EVERYONE! Trump’s tweet was shared over 102,800 times and liked over 384,800 times. His statements correlated with a 2000% increase in prescriptions for the anti-malarial drugs hydroxychloroquine and chloroquine in the US between 15 and 21 March 2020, resulting in many lupus patients unable to source the drug. There were also reports of overdoses as individuals sought to self-medicate with the drug to treat the virus. Once Trump declared himself a proponent of hydroxychloroquine, scientific inquiry into the drug was eclipsed by an overtly partisan debate. An analysis by Media Matters found that Fox News had promoted the drug 109 times between 23 and 25 March 2020, with other right wing media outlets following suit. The drug was further amplified and politicised by conservative public figures including Trump’s attorney Rudy Giuliani, who claimed on 27 March 2020 that “hydroxychloroquine has been shown to have a 100% effective rate in treating COVID-19”, and Brazil’s President, Jair Bolsonaro, who shared a Facebook post on 8 July 2020 admitting to taking the drug to treat the virus: “I’m one more person for whom this is working. So I trust hydroxychloroquine”. In addition to these conservative political figures endorsing hydroxychloroquine, on 27 July 2020 the right-wing syndicated news outlet Breitbart livestreamed a video depicting America’s Frontline Doctors – a group of physicians backed by the Tea Party Patriots, a conservative political organisation supportive of Trump – at a press conference outside the US Supreme Court in Washington. In the video, Stella Immanuel, a primary care physician in Texas, said “You don’t need masks…There is prevention and there is a cure!”, explaining that Americans could resume their normal lives by preemptively taking hydroxychloroquine. The video was retweeted by public figures including President Trump and Trump’s son Donald Trump Jr., before going viral reaching over 20 million users on Facebook. The video explicitly framed hydroxychloroquine as an effective “cure” for COVID-19 suppressed by “fake doctors”, thereby transferring it from potential treatment to a conspiracy object. These examples not only demonstrate the role of prominent public figures in amplifying conspiratorial claims about hydroxychloroquine as an effective cure for COVID-19, they reveal how these figures converted the drug into an “article of faith” divorced from scientific evidence. Consequently, to believe in its efficacy as a cure for COVID-19 demonstrated support for Trump and ideological skepticism of the scientific and medical establishment. Ivermectin, the “Miracle Cure” Ivermectin followed a different amplification trajectory. The amplifying process was primarily led by influencers in alternative media spaces and those associated with the IDW, many of whom position themselves in contrast to the mainstream media and politics. Despite scientists conducting clinical trials for ivermectin in early 2020, the ivermectin conspiracy peaked much later that year. On 8 December 2020, the pulmonary and ICU specialist Dr. Pierre Kory testified to the US Senate Committee about I-MASK: a prevention and early outpatient treatment protocol for COVID-19. During the hearing, Kory claimed that “ivermectin is effectively a ‘miracle drug’ against COVID-19”, which could end the pandemic. Kory’s depiction of ivermectin as a panacea, and the subsequent media hype, elevated him as a public figure and led to an increase in public demand for ivermectin in early 2021. This resulted in supply issues and led some people to seek formulations of the drug designed for animals, which were in greater supply and easier to access. Several months later in June 2021, Kory’s description of ivermectin as a “miracle cure” was amplified by a series of influencers, including Bret Weinstein and Joe Rogan, both of whom featured Kory on their podcasts as a key public figure in the fight against COVID Conspiratorial associations with ivermectin were further amplified on 9 July 2021 when Bret Weinstein appeared on Fox Nation's Tucker Carlson Today claiming he had “been censored for raising concerns about the shots and the medical establishment's opposition to alternative treatments”. The drug was embroiled in further controversy on 1 September 2021 when Joe Rogan shared an Instagram post explaining that he had taken ivermectin as one of many drugs to treat the virus. In the months that followed, Rogan featured several controversial scientists on his podcast who implied that ivermectin was an effective COVID “cure” suppressed as part of a global agenda to promote vaccine uptake. These public figures included Dr Robert Malone, an American physician who contributed to the development of mRNA technology, and Dr Peter McCullough, an American cardiologist with expertise in vaccines. As McCullough explained to Rogan in December 2021: it seemed to me early on that there was an intentional very comprehensive suppression of early treatment in order to promote fear, suffering, isolation, hospitalisation and death and it seemed to be completely organised and intentional in order to create acceptance for and then promote mass vaccination. McCullough went on to imply that the pandemic was planned and that vaccine manufacturers were engaged in a coordinated response to profit from mass vaccination. Consequently, whereas conservative public figures, such as Trump and Bolsonaro, played a primary role in amplifying the hype around hydroxychloroquine as a COVID cure and embroiling it in a political and conspiratorial narrative of collusion, influencers, especially those associated with alternative media and the IDW, were crucial in amplifying the ivermectin conspiracy online by platforming controversial scientists who espoused the drug as a “miracle cure”, which could allegedly end the pandemic but was being suppressed by the government and medical establishment. Debunking Debunking processes refuting the efficacy of these drugs as COVID “cures” contributed to the amplification of these conspiracies. In April 2020 the paper endorsing hydroxychloroquine that Trump tweeted about a week earlier was debunked. The debunking process for hydroxychloroquine involved a series of statements, papers, randomised clinical trials and retractions not only rejecting the efficacy of hydroxychloroquine, but suggesting it was unsafe and had the potential to cause harm (Boulware et al.; Mehra; Voss). In April 2020, the FDA released a statement cautioning against the use of hydroxychloroquine for COVID-19 outside of a hospital setting or a clinical trial due to risk of heart rhythm problems, and in June the FDA revoked its emergency use authorisation to treat COVID-19 in certain hospitalised patients. The debunking process was not limited to fact-based claims, it also involved satire and ridicule of those endorsing the drug as a treatment for COVID-19. Given the politicisation of the drug, much of this criticism was directed at Trump, as a key proponent of the drug, and Republicans in general, both of whom were cast as scientifically illiterate. The debunking process for ivermectin was similarly initiated by scientific and medical authorities who questioned the efficacy of ivermectin as a COVID-19 treatment due to reliability issues with trials and the quality of evidence (Lawrence). In response to claims that supply issues led people to seek formulations of the drug designed for animals, in April 2021 the FDA released a statement cautioning people not to take ivermectin to prevent or treat COVID-19: While there are approved uses for ivermectin in people and animals, it is not approved for the prevention or treatment of COVID-19 … . People should never take animal drugs … . Using these products in humans could cause serious harm. The CDC echoed this warning, claiming that “veterinary formulations intended for use in large animals such as horses, sheep, and cattle can be highly concentrated and result in overdoses when used by humans”. Many journalists and Internet users involved in debunking ivermectin reduced the drug to horse paste. Social media feeds debunking ivermectin were filled with memes ridiculing those consuming “horse dewormer”. Mockery of those endorsing ivermectin extended beyond social media, with the popular US sketch comedy show Saturday Night Live featuring a skit mocking Joe Rogan for consuming “horse medicine” to treat the virus. The skit circulated on social media in the following days, further deriding advocates of the drug as a COVID cure as not only irresponsible, but stupid. This type of ridicule, visually expressed in videos and Internet memes, fuelled polarisation. This polarisation was then weaponised by influencers associated with the IDW to sell ivermectin as a “miracle drug” suppressed by the medical and political establishment, thereby embroiling the drug further in conspiracy (Baker and Maddox). This type of opportunistic marketing is not intended for a mass audience. Instead, audiences are taking advantage of what Crystal Abidin refers to as “silosociality”, wherein content is tailored for specific subcommunities, which are not necessarily “accessible” or “legible” to outsiders (Abidin Refracted Publics 4). This dynamic both reflects and reinforces the audience segmentation that occurs in the current media ecology by virtue of alternative media with mockery and ridicule strengthening in- and out-group dynamics. Conclusion In this article we have traced how hydroxychloroquine and ivermectin moved from promising potential COVID-19 treatments to objects tainted by conspiracy. Despite common associations of conspiracy theories with the fringe, both the hydroxychloroquine and ivermectin conspiracy theories emerged in the mainstream, amplified across mainstream social networks with the help of influencers and public figures whose claims were further amplified by the news media commenting on their apparent endorsement of these drugs as COVID cures. Whereas hydroxychloroquine was politicised as a result of controversial public figures and right-wing media outlets endorsing the drug and the conspiratorial narrative espoused by America’s Frontline Doctors, notably much of the conspiracy around ivermectin shifted to alternative media spaces amplified by influencers disillusioned with the mainstream media. We have demonstrated how debunking processes, which sought to discredit these drugs as potential treatments for COVID-19, often ridiculed those who endorsed them, further polarising discussions involving these treatments and pushing advocates to the extreme. By encouraging proponents of these treatments to retreat to alternative media spaces, such as podcasts and newsletters, polarisation strengthened in-group dynamics, assisting the ability for opportunistic influencers to weaponise these conspiracies for social, economic, and political gain. These findings raise important questions about how to effectively counter conspiracies. When debunking not only refutes claims but ridicules advocates, debunking can have unintended consequences by strengthening in-group dynamics and fuelling the legitimacy of conspiratorial narratives. References Abidin, Crystal. Internet Celebrity: Understanding Fame Online. Emerald Group Publishing, 2018. Abidin, Crystal. "From ‘Networked Publics’ to ‘Refracted Publics’: A Companion Framework for Researching ‘below the Radar’ Studies." Social Media + Society 7.1 (2021). Baker, Stephanie Alice. "Alt.Health Influencers: How Wellness Culture and Web Culture Have Been Weaponised to Promote Conspiracy Theories and Far-Right Extremism during the COVID-19 Pandemic." European Journal of Cultural Studies 25.1 (2022): 3-24. Baker, Stephanie Alice, and Alexia Maddox. “COVID-19 Treatment or Miracle 'Cure'?: Tracking the Hydroxychloroquine, Remdesivir and Ivermectin Conspiracies on Social Media.” Paper presented at the BSA Annual Conference 2022: Building Equality and Justice Now, 20-22 April 2022. <https://www.britsoc.co.uk/media/25695/ac2022_draft_conf_prog.pdf>. Barkun, Michael. A Culture of Conspiracy. University of California Press, 2013. Boulware, David R., et al. "A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for Covid-19." New England Journal of Medicine 383.6 (2020): 517-525. Chou, Wen-Ying Sylvia, Anna Gaysynsky, and Robin C. Vanderpool. "The COVID-19 Misinfodemic: Moving beyond Fact-Checking." Health Education & Behavior 48.1 (2021): 9-13. Douglas, Karen M., et al. "Understanding Conspiracy Theories." Political Psychology 40 (2019): 3-35. Fenster, Mark. Conspiracy Theories: Secrecy and Power in American Culture. University of Minnesota Press, 1999. Gautret, Philippe, et al. "Hydroxychloroquine and Azithromycin as a Treatment of COVID-19: Results of an Open-Label Non-Randomized Clinical Trial." International Journal of Antimicrobial Agents 56.1 (2020): 105949. Golebiewski, Michael, and danah boyd. "Data Voids: Where Missing Data Can Easily Be Exploited." Data & Society (2019). Kelsey, Darren. "Archetypal Populism: The ‘Intellectual Dark Web’ and the ‘Peterson Paradox’." Discursive Approaches to Populism across Disciplines. Cham: Palgrave Macmillan, 2020. 171-198. Lawrence, Jack M., et al. "The Lesson of Ivermectin: Meta-Analyses Based on Summary Data Alone Are Inherently Unreliable." Nature Medicine 27.11 (2021): 1853-1854. Marwick, Alice E. Status Update. Yale University Press, 2013. Mehra, Mandeep R., et al. "RETRACTED: Hydroxychloroquine or Chloroquine with or without a Macrolide for Treatment of COVID-19: A Multinational Registry Analysis." (2020). Parks, Gabriel. "Considering the Purpose of ‘an Alternative Sense-Making Collective’: A Rhetorical Analysis of the Intellectual Dark Web." Southern Communication Journal 85.3 (2020): 178-190. Senft, Theresa M. Camgirls: Celebrity and Community in the Age of Social Networks. Peter Lang, 2008. Smith, Tim, et al. "COVID-19 Drug Therapy." Elsevier (2020). Voss, Andreas. “Official Statement from International Society of Antimicrobial Chemotherapy (ISAC).” International Society of Antimicrobial Chemotherapy 3 Apr. 2020. <https://www.isac.world/news-and-publications/official-isac-statement>.