Significant Improvement In Recall Research Articles

Comparative evaluation of methods for the prediction of protein-ligand binding sites.

The accurate identification of protein-ligand binding sites is of critical importance in understanding and modulating protein function. Accordingly, ligand binding site prediction has remained a research focus for over three decades with over 50 methods developed and a change of paradigm from geometry-based to machine learning. In this work, we collate 13 ligand binding site predictors, spanning 30years, focusing on the latest machine learning-based methods such as VN-EGNN, IF-SitePred, GrASP, PUResNet, and DeepPocket and compare them to the established P2Rank, PRANK and fpocket and earlier methods like PocketFinder, Ligsite and Surfnet. We benchmark the methods against the human subset of our new curated reference dataset, LIGYSIS. LIGYSIS is a comprehensive protein-ligand complex dataset comprising 30,000 proteins with bound ligands which aggregates biologically relevant unique protein-ligand interfaces across biological units of multiple structures from the same protein. LIGYSIS is an improvement for testing methods over earlier datasets like sc-PDB, PDBbind, binding MOAD, COACH420 and HOLO4K which either include 1:1 protein-ligand complexes or consider asymmetric units. Re-scoring of fpocket predictions by PRANK and DeepPocket display the highest recall (60%) whilst IF-SitePred presents the lowest recall (39%). We demonstrate the detrimental effect that redundant prediction of binding sites has on performance as well as the beneficial impact of stronger pocket scoring schemes, with improvements up to 14% in recall (IF-SitePred) and 30% in precision (Surfnet). Finally, we propose top-N+2 recall as the universal benchmark metric for ligand binding site prediction and urge authors to share not only the source code of their methods, but also of their benchmark.Scientific contributionsThis study conducts the largest benchmark of ligand binding site prediction methods to date, comparing 13 original methods and 15 variants using 10 informative metrics. The LIGYSIS dataset is introduced, which aggregates biologically relevant protein-ligand interfaces across multiple structures of the same protein. The study highlights the detrimental effect of redundant binding site prediction and demonstrates significant improvement in recall and precision through stronger scoring schemes. Finally, top-N+2 recall is proposed as a universal benchmark metric for ligand binding site prediction, with a recommendation for open-source sharing of both methods and benchmarks.

NewsZoom (focus on the important): Text summarization of News Articles based on named entity recognition using Spacy library

The study discusses the importance of summarization in dealing with a large amount of data available on the internet. The study used a deep-learning algorithm based on functions from the spacy library in Python to summarize news articles and evaluated the impact of named entity recognition on the summarization process. The study assessed different datasets from CNN-DailyMail and the BBC (entertainment articles) and found that the proposed method based on named entity recognition showed significant improvement in recall, precision, and F-score compared to the word frequency method. The study also observed that the articles from CNN- DailyMail were longer, with an average of 551 words and 28 sentences, compared to the BBC (entertainment articles), which had an average of 190 words and 12 sentences. The evaluation results showed that the proposed method based onnamed entity recognition performed better on the shorter articles from the BBC, indicating that the method was more effective in summarizing shorter texts. In summary, the study highlighted the importance of summarization in dealing with a large amount of data available on the internet. It showed that named entity recognition can significantly improve the effectiveness of the summarization process. The study alsoobserved that the proposed method was more effective in summarizing shorter texts Keywords: Spacy library, entity recognition, Summarization, Deep-Learning, CNN-DailyMail

Synthetic Data-Driven Real-Time Detection Transformer Object Detection in Raining Weather Conditions

Images captured in rainy weather conditions often suffer from contamination, resulting in blurred or obscured objects, which can significantly impact detection performance due to the loss of identifiable texture and color information. Moreover, the quality of the detection model plays a pivotal role in determining detection outcomes. This study adopts a dual perspective, considering both pre-trained models and training data. It employs 15 image augmentation techniques, combined with neural style transfer (NST), CycleGAN, and an analytical method, to synthesize images under rainy conditions. The Real-Time Detection Transformer (RTDETR) and YOLOv8 pre-trained models are utilized to establish object detection frameworks tailored for rainy weather conditions. Testing is carried out using the DAWN (Detection in Adverse Weather Nature) dataset. The findings suggest compatibility between the pre-trained detection models and various data synthesis methods. Notably, YOLOv8 exhibits better compatibility with CycleGAN data synthesis, while RTDETR demonstrates a stronger alignment with the NST and analytical approaches. Upon the integration of synthesized rainy images into model training, RTDETR demonstrates significantly enhanced detection accuracy compared to YOLOv8, indicating a more pronounced improvement in performance. The proposed approach of combining RTDETR with NST in this study shows a significant improvement in Recall (R) and mAP50-95 by 16.35% and 15.50%, respectively, demonstrating the robust rainy weather resilience of this method. Additionally, RTDETR outperforms YOLOv8 in terms of inference speed and hardware requirements, making it easier to use and deploy in real-time applications.

Boost recall in quasi-stellar object selection from highly imbalanced photometric datasets

Context. The identification of bright quasi-stellar objects (QSOs) is of fundamental importance to probe the intergalactic medium and address open questions in cosmology. Several approaches have been adopted to find such sources in the currently available photometric surveys, including machine learning methods. However, the rarity of bright QSOs at high redshifts compared to other contaminating sources (such as stars and galaxies) makes the selection of reliable candidates a difficult task, especially when high completeness is required. Aims. We present a novel technique to boost recall (i.e., completeness within the considered sample) in the selection of QSOs from photometric datasets dominated by stars, galaxies, and low-z QSOs (imbalanced datasets). Methods. Our heuristic method operates by iteratively removing sources whose probability of belonging to a noninteresting class exceeds a user-defined threshold, until the remaining dataset contains mainly high-z QSOs. Any existing machine learning method can be used as the underlying classifier, provided it allows for a classification probability to be estimated. We applied the method to a dataset obtained by cross-matching PanSTARRS1 (DR2), Gaia (DR3), and WISE, and identified the high-z QSO candidates using both our method and its direct multi-label counterpart. Results. We ran several tests by randomly choosing the training and test datasets, and achieved significant improvements in recall which increased from ~50% to ~85% for QSOs with z > 2.5, and from ~70% to ~90% for QSOs with z > 3. Also, we identified a sample of 3098 new QSO candidates on a sample of 2.6 ×106 sources with no known classification. We obtained follow-up spectroscopy for 121 candidates, confirming 107 new QSOs with z > 2.5. Finally, a comparison of our QSO candidates with those selected by an independent method based on Gaia spectroscopy shows that the two samples overlap by more than 90% and that both selection methods are potentially capable of achieving a high level of completeness.

Chinese English language learners’ vocabulary retention: Investigating the effectiveness of neuro/metacognitive and socio-cultural strategies

BackgroundThe acquisition of a rich vocabulary is foundational to language proficiency. In the pedagogical pursuit of effective vocabulary teaching, educators explore diverse methodologies. Researchers investigated the impact of different neurocognitive, metacognitive, and socio-cultural strategies on enhancing vocabulary learning, particularly among Chinese English Language Learners.ObjectivesThe study aims to determine the effectiveness of techniques derived from these theories compared to traditional teaching methods in enhancing vocabulary recall and recognition among English language learners.MethodologyA quasi-experimental pre-test/post-test design was employed for the experimental and control groups, comprising 90 Chinese EFL learners selected from educational institutions in 2022–2023. The experimental group (n = 45) received instruction involving visual imagery, multisensory rotation, circle rotation, and mind mapping over eight sessions, while the control group (n = 45) received traditional teaching methods.FindingsStatistical analysis, utilizing covariance and analysis of variance with SPSS software version 22, revealed significant improvements in recall and vocabulary recognition within and between the experimental and control groups.ConclusionsThe results indicate that incorporating techniques based on Neuro-Cognitive, Multimedia, Socio-Cultural, and Metacognitive theories positively influences vocabulary recall and recognition. This suggests the efficacy of these innovative methods in enhancing English language learning, highlighting their potential for broader integration into EFL instruction.

Effects of Virtual Reality on Complex Building System Recall

Multidisciplinary design and construction teams are challenged to communicate and coordinate across complex building systems, including architectural, structural, mechanical, electrical, and piping (MEP). To support this coordination, disciplinary 3D models are combined and coordinated before installation. Studies show that besides the use of 3D models, industry professionals sketch building components to discuss coordination issues and find resolutions that require them to recall the building components in the model. In current practices, 3D models are explored with Building Information Modeling (BIM) tools presented on 2D screens, while Virtual Reality (VR) can provide users with an immersive environment to explore. This paper presents the results of an experiment that studied the effects of VR’s immersive environment on the participants’ complex MEP system recall compared to BIM via sketching. The comparison criteria were the 3D geometry properties of the piping system and the users’ self-awareness in the model categorized under color, shape, dimension, piping, and viewpoint. The results showed significant improvement in recall of shape, dimension, and piping when the model was explored in VR.

Hierarchical User Intention-Preference for Sequential Recommendation with Relation-Aware Heterogeneous Information Network Embedding.

Existing recommender systems usually make recommendations by exploiting the binary relationship between users and items, and assume that users only have flat preferences for items. They ignore the users' intentions as an origin and driving force for users' performance. Cognitive science tells us that users' preference comes from an explicit intention. They first have an intention to possess a particular (type of) item(s) and then their preferences emerge when facing multiple available options. Most of the data used in recommender systems are composed of heterogeneous information contained in a complicated network's structure. Learning effective representations from these heterogeneous information networks (HINs) can help capture the user's intention and preferences, therefore, improving recommendation performance. We propose a hierarchical user's intention and preferences modeling for sequential recommendation based on relation-aware HIN embedding (HIP-RHINE). We first construct a multirelational semantic space of heterogeneous information networks to learn node embedding based on specific relations. We then model user's intention and preferences using hierarchical trees. Finally, we leverage the structured decision patterns to learn user's preferences and thereafter make recommendations. To demonstrate the effectiveness of our proposed model, we also report on the conducted experiments on three real data sets. The results demonstrated that our model achieves significant improvements in Recall and Mean Reciprocal Rank metrics compared with other baselines.

Translating medical image to radiological report: Adaptive multilevel multi-attention approach

Background and ObjectiveMedical imaging techniques are widely employed in disease diagnosis and treatment. A readily available medical report can be a useful tool in assisting an expert for investigating the patient’s health. A radiologist can benefit from an automatic medical image to radiological report translation system while preparing a final report. Previous attempts on automatic medical report generation task includes image captioning algorithms without taking domain-specific visual and textual contents into account, thus arises the question about credibility of generated report. MethodsIn this work, a novel Adaptive Multilevel Multi-Attention (AMLMA) approach is proposed by offering domain-specific visual-textual knowledge to generate a thorough and believable radiological report for any view of a human chest X-ray image. The proposed approach leverages the encoder-decoder framework incorporated with multiple adaptive attention mechanisms. The potential of a convolutional neural network (CNN) with residual attention module (RAM) is demonstrated as a strong visual encoder for multi-label abnormality detection. The multilevel visual features (local and global) are extracted from proposed visual encoder to retrieve regional-level and abstract-level radiology-based semantic information. The Word2Vec and FastText word embeddings are trained on medical reports to acquire radiological knowledge and further used as textual encoders, feeding as input to Bi-directional Long Short Term Memory (Bi-LSTM) network to learn the co-relationship between medical terminologies in radiological reports. The AMLMA employs a weighted multilevel association of adaptive visual-semantic attention and visual-based linguistic attention mechanisms. This association of adaptive attention is exploited as a decoder and produces significant improvements in the report generation task. ResultsThe proposed approach is evaluated on a publicly available Indiana University chest X-ray (IU-CXR) dataset. The CNN with RAM shows the significant improvement in recall (0.4423), precision (0.1803) and F1-score (0.2551) for prediction of multiple abnormalities in X-ray image. The results of language generation metrics for proposed variants were acquired using the COCO-caption evaluation Application Program Interface (API). The trained embeddings with AMLMA model generates the convincing radiology report and outperform state-of-the-art (SOTA) approaches with high evaluation metrics scores for Bleu-4 (0.172), Meteor (0.247), Rouge_L (0.376) and CIDEr (0.381). In addition, a new “Unique Index” (UI) statistic is introduced to highlight the model’s ability for generating unique reports. ConclusionThe overall architecture aids to the understanding of various X-ray image views and generating the relevant normal and abnormal radiography statements. The proposed model is emphasized on multi-level visual-textual knowledge with adaptive attention mechanism to balance visual and linguistic information for the generation of admissible radiology report.

Design of a Music Recommendation Model on the Basis of Multilayer Attention Representation

Current music recommendation systems can explore the general relationship between the users and songs to recommend music to the users; however, they cannot distinguish the different preferences of different users for the same song. For example, a user may like a song because of the singer, while another user will like it not for the singer but just because of the composition of the song or its melody. A recommender system that knows this difference would be more effective in recommending music to the users. To this end, this paper proposes a music recommendation model based on multilayer attention representation, which learns song representations from multidimensions using user-attribute information and song content information, and mines the preference relationship between users and songs. In order to distinguish the differences in user preferences for multidomain features of songs, a feature-dependent attention network is designed; in order to distinguish the differences in user preferences for different historical behaviors and to explore the temporal dependence of user behaviors, a song-dependent attention network is designed. Finally, the SoftMax function is used to calculate the distribution of users’ preferences for candidate songs and is used to generate recommendations. The experimental results on 30Music and MIGU datasets show that the proposed model achieves significant improvement in recall and MRR compared with the current recommendation models.

Low-light image enhancement using Gaussian Process for features retrieval

Low-light is a challenging environment for image processing and computer vision tasks, either in contrast enhancement for better visibility and quality, or application oriented tasks such as detection. We found that the current trend of low-light enhancement research is heavily on quality improvement. In this work, we aim to shift the focus towards a more functional direction, that is enhancement that prioritizes feature retrieval. For this reason, we first propose to model low-light enhancement as a set of localized functions using Gaussian Process (GP) that is trained at runtime using data from a simple Convolutional Neural Network (CNN) to provide the necessary feature information as reference. The CNN is in turn trained using large amount of synthetic data, based upon the luminance distribution of real world low-light images to learn the relationship between features and pixels. Secondly, we also proposed two new evaluation metrics to better assess enhancement algorithms to support high level computer vision tasks, namely, local features matching and intensity histogram similarity. In our experiments, our proposed low-light enhancement framework outperforms the state-of-the-arts with significant improvement in Recall, F1, and F2-score of SIFT features matching, and achieve comparable results for l1-norm distance of histograms as well as the PSNR. Moreover, our analysis of the performance showed that the PSNR quality metric is not only unable to assess the practicality of the results, but also inappropriately gives high assessment to low visual quality images.

Improving Regional and Teleseismic Detection for Single-Trace Waveforms Using a Deep Temporal Convolutional Neural Network Trained with an Array-Beam Catalog.

The detection of seismic events at regional and teleseismic distances is critical to Nuclear Treaty Monitoring. Traditionally, detecting regional and teleseismic events has required the use of an expensive multi-instrument seismic array; however in this work, we present DeepPick, a novel seismic detection algorithm capable of array-like detection performance from a single-trace. We achieve this performance through three novel steps: First, a high-fidelity dataset is constructed by pairing array-beam catalog arrival-times with single-trace waveforms from the reference instrument of the array. Second, an idealized characteristic function is created, with exponential peaks aligned to the cataloged arrival times. Third, a deep temporal convolutional neural network is employed to learn the complex non-linear filters required to transform the single-trace waveforms into corresponding idealized characteristic functions. The training data consists of all arrivals in the International Seismological Centre Database for seven seismic arrays over a five year window from 1 January 2010 to 1 January 2015, yielding a total training set of 608,362 detections. The test set consists of the same seven arrays over a one year window from 1 January 2015 to 1 January 2016. We report our results by training the algorithm on six of the arrays and testing it on the seventh, so as to demonstrate the generalization and transportability of the technique to new stations. Detection performance against this test set is outstanding, yielding significant improvements in recall over existing techniques. Fixing a type-I error rate of 0.001, the algorithm achieves an overall recall (true positive rate) of 56% against the 141,095 array-beam arrivals in the test set, yielding 78,802 correct detections. This is more than twice the 37,572 detections made by an STA/LTA detector over the same period, and represents a 35% improvement over the 58,515 detections made by a state-of-the-art kurtosis-based detector. Furthermore, DeepPick provides at least a 4 dB improvement in detector sensitivity across the board, and is more computationally efficient, with run-times an order of magnitude faster than either of the other techniques tested. These results demonstrate the potential of our algorithm to significantly enhance the effectiveness of the global treaty monitoring network.

Randomized Controlled Trial of a Computer-Based Education Program in the Home for Solid Organ Transplant Recipients

De novo solid organ transplant recipients (SOTR) have a steep learning curve to acquire medication knowledge. Without adequate knowledge, SOTR are at risk of nonadherence and poor transplant outcomes. In this nonblinded, randomized controlled trial, de novo SOTR received standard teaching with or without postdischarge computer-based education (CBE) at home. Primary outcomes were change in knowledge (quiz and recall) and satisfaction, assessed by questionnaires at baseline and 3 months. Adherence was evaluated via self-report and immunosuppressant levels. Two hundred forty-six patients were randomized and 209 completed the 3-month analysis. In the intervention arm, 73 (57.9%) used the CBE program. Change in knowledge quiz score did not differ between groups (4.9% vs 0.6%; P = 0.084), despite a significant increase within the intervention (72.4% vs 77.3%, P = 0.007) but not the control (76.0% vs 76.6%, P = 0.726) arms. Both groups had a significant improvement in recall (intervention, 56.7% vs 82.1%, P < 0.001; control, 51.3% vs 79.7%, P < 0.001), with similar changes in scores (25.4% vs 28.4%, P = 0.55). Change in satisfaction differed between groups (intervention, 1.2% vs control, -4.9%; P = 0.050). There was a significant decline in satisfaction within the control group (88.4% vs 83.5%, P = 0.035), whereas satisfaction was maintained with the intervention (85.6% vs 86.8%, P = 0.55). Adherence was similar in both groups. Knowledge improved over the study period in both groups, with no incremental benefit for the intervention. Patient satisfaction was maintained with the CBE program. More research is needed to identify barriers to uptake of CBE at home and to develop effective strategies for posttransplant education.

Learning Genetic Regulatory Network Connectivity from Time Series Data

Recent experimental advances facilitate the collection of time series data that indicate which genes in a cell are expressed. This information can be used to understand the genetic regulatory network that generates the data. Typically, Bayesian analysis approaches are applied which neglect the time series nature of the experimental data, have difficulty in determining the direction of causality, and do not perform well on networks with tight feedback. To address these problems, this paper presents a method to learn genetic network connectivity which exploits the time series nature of experimental data to achieve better causal predictions. This method first breaks up the data into bins. Next, it determines an initial set of potential influence vectors for each gene based upon the probability of the gene's expression increasing in the next time step. These vectors are then combined to form new vectors with better scores. Finally, these influence vectors are competed against each other to determine the final influence vector for each gene. The result is a directed graph representation of the genetic network's repression and activation connections. Results are reported for several synthetic networks with tight feedback showing significant improvements in recall and runtime over Yu's dynamic Bayesian approach. Promising preliminary results are also reported for an analysis of experimental data for genes involved in the yeast cell cycle.

Pilot study using an Internet-based program in informed consent

Purpose To examine the effect of an internet-based aid to informed consent on parent recall of potential surgical complications. Methods Parents of children scheduled for elective inguinal hernia repair were assigned to a control group or were enrolled in an internet-based program designed to aid in the consent process. Nine potential surgical complications were presented to the parent(s) in the consent discussion and in the Internet program. Parent recall of potential surgical complications was assessed immediately after the consent discussion and on the day of surgery. Results Overall recall of complications was poor in both groups, both immediately and on the day of surgery. Parents in the control group (n = 13) recalled a mean of 2.9 complications immediately and 1.5 on the day of surgery, approaching statistical significance ( P = .056). The parents in the internet program group (n = 17) recalled a mean of 2.6 complications immediately and 2.9 on the day of surgery ( P = NS). There was no significant difference in immediate recall between the two groups, but there was a trend towards statistically significant improvement in recall in the study group the day of surgery vs. controls ( P = .06). Conclusion Although overall recall of potential surgical complications was poor in both groups, there was a trend towards a significant improvement in recall in the study group after viewing the Internet-based program.

A Theatrical Intervention to Improve Cognition in Intact Residents of Long Term Care Facilities

This study explored a theatrically-based intervention for old-olds (average age: 82.3) living in a continuing care facility. Approximately half of the participants used hearing aids and/or walkers, canes or wheelchairs, and most had self-reported memory concerns although they did not suffer from serious mental impairment as assessed by the SPMSQ. A distinguishing feature of the intervention was that no element of the theatre instruction was targeted to the test instruments. Instead, participants were trained to create life-like characters who could spontaneously interact with others. The design consisted of a one group pretest-posttest with double pretest. After four weeks of twice-weekly instructional sessions, the researchers found significant improvement in recall and problem-solving ability. Reasons for the efficacy of theatrical training with this population are discussed. Also, evidence is presented for the self-reinforcing nature of the program.

Effects of a knowledge base manipulation on children's recall

Developmental researchers have demonstrated the potent role of a knowledge base on children's memory performance. However, the only studies to experimentally assess a causal connection between children's knowledge and memory have failed to find such connection (DeMarie-Dreblow, 1991; Muir-Broaddus, Rorer, Branden, & George, 1995). In this study, we extended the length of training relative to earlier studies with samples of first- and seventh-grade children, incorporating knowledge training into their school curriculum. Both groups of children displayed an increase in knowledge as a result of training and corresponding significant improvements in recall, consistent with a causal relation between knowledge and memory. Although we believe that some of the benefits for the older children were associated with improvements in strategic organization, most of the gains, for both the older and younger children, could be attributed to item-specific effects of knowledge on recall.

Effects of Cues on Memory in Alcoholics and Controls

This study was conducted to determine whether alcoholics' memory difficulties are due, in part, to access (retrieval) or to availability (retention) deficits. Forty-four alcoholics (n = 20 females) and 44 controls (n = 22 females) learned a paired associate list consisting of 12 adjective-CCC trigram pairs. Half of the subjects in each group learned the list to a low degree of learning (DOL; 4/12 pairs); the remainder to a high DOL (8/12 pairs). Two distinct environmental contexts (providing implicit cues) were used during acquisition. Subjects then completed a cued recall (an explicit cue) test in either the same or a different room. Alcoholics were significantly inferior in the acquisition phase on trials required to reach criterion, regardless of DOL required [F(1,68) = 10.92, p = 0.002]. The main effect for implicit cuing was not significant; similarly, there were no significant interactions. In contrast, the explicit cue manipulation produced a significant group x DOL interaction on the number of trigrams correctly recalled [F = (1,77) = 6.38, p = 0.01]; alcoholics' recall did not benefit from the higher DOL in contrast to a significant improvement in recall by controls. The failure of alcoholics to demonstrate improvement with higher levels of learning is consistent with a deficit in the availability of information. The results confirm previous reports of recovering alcoholics' verbal learning and memory dysfunction, and suggest that these deficits may be attributed, in part, to a deficit in the availability of information (retention).

The effectiveness of document neighboring in search enhancement

We consider two kinds of queries that may be applied to a database. The first is a query written by a searcher to express an information need. The second is a request for documents most similar to a document already judged relevant by the searcher. We examine the effectiveness of these two procedures and show that in important cases the latter query type is more effective than the former. This provides a new view of the cluster hypothesis and a justification for document neighboring procedures (precomputation of closely related documents). If all the documents in a database have readily available precomputed nearest neighbors, a new search algorithm, which we call parallel neighborhood searching, is conveniently used. We show that this feedback-based method provides significant improvement in recall over traditional linear searching methods and even appears superior to traditional feedback methods in overall performance.

Optical disc technology for records management: A user perspective

The United States Nuclear Regulatory Commission is completing a pilot project which demonstrates a unique application of Worm‐based optical disc technology for the storage and retrieval of government records. Information which is captured at the source or from an optical character reader is stored on the disc in the form of the ASCII text and the original image of the complete document. The full text search software and indexes to the images are all stored on the optical disc. In this way a user can perform a full text search on the records and retrieve either the ASCII or the original document. By combining full text search and image retrieval, significant improvement in recall and precision are achieved when compared to traditional simple indexing techniques.

Long-Term Memory of Operativity Figures

Summary Ninety-two first and second graders (45 girls and 47 boys) were assessed for operativity with conservation (number, length, amount) seriation, and horizontality tasks and tested immediately, one week later, and five months later for their recall and recognition of the operativity stimuli in the tasks and a nonoperativity stimulus (the numeral 3). With very few exceptions there were no significant relationships between operativity and memory at any time. After five months there were large and significant amounts of forgetting for nearly all task stimuli for recall and recognition. However, with more lenient retention criteria based upon only operativity and not figurative aspects of the stimuli, there was evidence of significant improvement in recall for some children over the five months.