Information Retrieval Journal最新文献

In a wide variety of applications, such as indoor position selection for advertising and setting rents of different shops in a shopping mall, it is better to get the passenger flow of each room. In the indoor space, the positions of users are commonly captured by the indoor positioning system consisting of static positioning devices. And the sequence of all tracking events with the same user ordered by the corresponding time is the indoor trajectory of this user. Thus, in this paper, we define and study two essential queries named Rooms with top-k passenger flows at a Timestamp query (RkT for short) and Rooms with top-k passenger flows within a time Interval query (RkI for short), i.e., how to search rooms with top-k passenger flows at a timestamp and within a time interval in the past using indoor trajectories, respectively. For the indoor positioning system, there are only limited static positioning devices deployed in the indoor space on account of the cost. And the detection ranges of these static positioning devices only cover a small part of the indoor space. When a user is in the undetected state, there is uncertainty in its position combined with the quite complex indoor topology. Such uncertainty brings great challenges to determining the passenger flow in each room. Considering the distribution of static positioning devices, we propose a new method about how to reasonably infer where a user is in the undetected state and the corresponding probability based on its indoor trajectory and the complex indoor topology. In order to quickly retrieve the set of indoor trajectories, we propose a full Binary tree indexing indoor trajectories divided by Time intervals (BiT for short), which is built on the given set of indoor trajectories. Based on the index BiT, we propose PAT Algorithm and PAI Algorithm to efficiently process RkT and RkI queries, respectively. Extensive experiment results demonstrate superior performance of PAT Algorithm and PAI Algorithm.

Polycystic ovary syndrome (PCOS) is an endocrine disorder affecting women of reproductive age characterized by the presence of multiple follicles in the ovaries that are detectable via ultrasound imaging. Early diagnosis of PCOS morphology can be challenging due to low resolution and increased speckle noise, making it difficult to identify smaller follicle boundaries. This article introduces a novel methodology, multiscale gradient-weighted oriented Otsu thresholding with sum of product fusion (MOT-SF), to address these challenges. The MOT-SF technique precisely recognizes smaller region boundaries even at lower resolutions by employing a pyramidal structure for image computation at multiple scales. Otsu's thresholding is used to segment the image, optimizing the threshold by minimizing the interclass variance at each stage. Incorporating gradient weights (λ) within classes enhances smaller boundary regions and reduces noise. Additionally, the MOT-SF method integrates a sum of product fusion strategies, combining segmented images from various scales to produce a final image that preserves both small and large PCOS structures while mitigating noise. The experimental results show that MOT-SF outperforms traditional methods such as Otsu’s thresholding and Chan-Vese models, as well as deep learning approaches such as R-CNN, in terms of computational efficiency and robustness to variations in ultrasound image quality. The MOT-SF methodology achieves an accuracy of nearly 85% and a precision of 94%, highlighting its potential to improve the detection and characterization of follicles in ultrasound images and advancing diagnostic tools in reproductive health.

Graphical Abstract

The key to improving the performance of graph convolutional networks (GCN) is to fully explore the correlation between neighboring and distant information. Aiming at the over-smoothing problem of GCN, in order to make full use of the relationship among features, graphs and labels, a graph residual generation network based on multi-information aggregation (MIA-GRGN) is proposed. Firstly, aiming at the defects of GCN, we design a deep initial residual graph convolution network (DIRGCN), which connects the initial input through residuals, so that each layer node retains part of the information of the initial features, ensuring the localization of the graph structure and effectively alleviating the problem of over-smoothing. Secondly, we propose a random graph generation method (RGGM) by utilizing graph edge sampling and negative edge sampling, and optimize the supervision loss function of DIRGCN in the form of generation framework. Finally, applying RGGM and DIRGCN as inference modules for modeling hypotheses and obtaining approximate posterior distributions of unknown labels, an optimized loss function is obtained, we construct a multi-information aggregation MIA-GRGN that combines graph structure, node characteristics and label joint distribution. Experiments on benchmark graph classification datasets show that MIA-GRGN achieves better classification results compared with the benchmark models and mainstream models, especially for datasets with less dense edge relationships between nodes.

The amount of information is multiplying, one of the popular and widely used formats is short videos. Therefore, maintaining the copyright protection of this information, preventing it from being disclosed without authorization, is a challenge. This work presents a way to rank a set of short videos based on a video profile similarity metric, finding a set of reference videos, using a self-supervised method, without the need for human tagging. The self-supervised method uses a search based on a Genetic Algorithm, of a subgroup of the most similar videos. Similarities are calculated using the SMAPE metric on video signatures vectors, generated with a fixed size, using Structural Tensor, maximum sub matrix and T-SNE.

This scoping review aimed to improve the understanding of important factors in digital accessibility for people with visual impairment and blindness, focusing on the first-hand experiences and challenges faced by this target group while using digital technologies. Keywords related to ‘digital technologies,’ ‘accessibility,’ ‘visual impairment,’ and ‘blindness’ were used in searching two databases (n = 683), with additional articles identified by means of manual searches (n = 60). Two reviewers independently screened the titles and abstracts to select 97 articles for full-text screening, of which 49 articles met the inclusion criteria and were selected for review based on the WCAG guidelines, highlighting details for consideration and improvement of the guidelines. The analysis revealed that users suffered from inaccessibility in several ways. For example, many applications or websites are developed for sighted users, where information is communicated through visual content without providing alternatives for assistive technology users. In addition, the lack of keyboard accessibility, shortcuts, or compatibility with different assistive technologies remains a consistent challenge. Furthermore, it was highlighted that simple accessibility measures are not followed adequately or consistently, such as providing alternative text for images or labels for links and buttons. This review highlighted the challenges and consequences of the inaccessibility of digital technologies, providing a detailed explanation regarding the elements that should be considered in the development of digital technologies. It is recommended that people with disabilities should be involved in the design of technology and its accessibility assessment.

In an era marked by rapid digital transformation, understanding the factors that influence digital engagement is crucial for bridging the digital divide. This study aims to explore the impact of individual factors such as networking motive, social media use, content service usage, and economic activity on digital transformation awareness and satisfaction. Utilizing Partial Least Squares Structural Equation Modeling (PLS-SEM) to analyze data from 7,000 respondents of the National Information Society Agency (NIA)'s 2022 Digital Divide Survey, this research provides empirical insights into the dynamics of digital engagement. The findings reveal that networking motive significantly predicts social media use, which in turn slightly enhances digital transformation awareness but not satisfaction. Conversely, economic activity positively influences both awareness and satisfaction with digital transformation, underscoring the tangible benefits of digital economic engagement. Life service utilization emerged as a crucial factor, significantly impacting both awareness and satisfaction. These results offer critical implications for policymakers, educators, and digital platform developers, suggesting the need for targeted strategies to enhance digital literacy, promote inclusive digital services, and foster economic opportunities in the digital domain.

Zigbee is a wireless network technology that operates on a community-based infrastructure. The primary objective of this system is to allow for the effective and inexpensive transmission of information across intelligent devices, focusing on minimizing latency and addressing potential challenges. The prominence of security risks related to the Zigbee community has escalated due to the expanded implementation of Zigbee technology. The Zigbee community has implemented measures to mitigate network disruptions, prevent unauthorized access, and protect against the interception of statistical data. A plethora of encryption algorithms have been developed. The encryption technologies employed encompass a 128-bit Advanced Encryption Standard (AES) to encrypt data during transfer. Cryptographically safe protocols (CSPs) are implemented to ensure safe connections and authentication. This research introduces an innovative methodology for assessing the security of Zigbee community-based Internet of Things (IoT) systems. This article thoroughly examines the security concerns associated with Zigbee technology. Adopting intelligent design concepts in the construction of IoT devices facilitates the implementation of robust security measures and reliable detection systems to mitigate potential assaults. The system can implement access control mechanisms based on authentication, guaranteeing the secure management of devices, networks, and protocols and facilitating the secure installation of software updates.

This study investigates the transformative influence of artificial intelligence (AI) on accounting procedures in Saudi Arabia by examining accounting professionals’ attitudes, understanding, and practices regarding AI implementation. Using a questionnaire-based survey distributed among accounting professionals in Saudi Arabia, data analysis was conducted using the partial least squares (PLS) technique. We find significant direct relationships between AI awareness and usage, AI engagement and accountants, and the impact of AI and accounting procedures. This finding suggests that accountants who are knowledgeable about and utilize AI are more likely to be engaged in AI, leading to positive changes in accounting procedures. Moreover, the robust positive relationship between AI’s impact on accounting procedures and accounting efficiency indicates a significant positive influence. The outcomes revealed that AI engagement and impact played significant mediating roles in these relationships. These findings suggest that, while AI awareness and usage alone can lead to improved accounting outcomes, the effect is mediated by the level of engagement with AI and its impact. We provide compelling evidence that AI positively affects the accounting profession. Accountants who are aware of and use AI are more likely to engage in and experience positive changes in their accounting procedures. We provide theoretical insights into the influence of AI on accounting procedures, offer valuable resources to academics, and suggest areas for future research. It equips practitioners with effective strategies for incorporating AI into their practices, highlighting the harnessing of AI’s transformative potential of AI while addressing potential challenges.

The task of aspect-level sentiment analysis is intricately designed to determine the sentiment polarity directed towards a specific target within a sentence. With the increasing availability of online reviews and the growing importance of healthcare decisions, analyzing drug reviews has become a critical task. Traditional sentiment analysis, which categorizes a whole review as positive, negative, or neutral, provides limited insights for consumers and healthcare professionals. Aspect-based sentiment analysis (ABSA) aims to overcome these limitations by identifying and evaluating the sentiment associated with specific aspects or attributes of drugs mentioned in the reviews. Various fields, including business, politics, and medicine, have been explored in the context of sentiment analysis. Automation of online user reviews allows pharmaceutical companies to assess large amounts of user feedback. This helps extract pharmacological efficacy and side effect insights. The data collected could improve pharmacovigilance. Reviewing user comments can provide valuable data that can be used to improve drug safety and efficacy monitoring procedures. This improves pharmacovigilance processes, improving pharmaceutical outcomes understanding and corporate decision-making. Therefore, we propose a pre-trained RoBERTa with a Bi-LSTM model to categorise drug reviews from online sources and pre-process the text data. Ant Colony Optimization can be used in feature selection for ABSA, helping to identify the most relevant aspects and sentiments. Further, RoBERTa is fine-tuned to perform ABSA on the dataset, enabling the system to categorize aspects and determine the associated sentiment. The outcomes reveal that the suggested framework has achieved higher accuracy (96.78%) and F1 score (98.29%) on druglib.com, and 95.02% on the drugs.com dataset, than several prior state-of-the-art methods.

The development of threshold protocols based on lattice-signature schemes has been of increasing interest in the past several years. The main research focus has been towards protocols constructed for various variants of Crystals-Dilithium, future NIST digital signature standard known as ML-DSA. In this work, we propose TOPCOAT, a two-party lattice-based signature algorithm that embodies Dilithium’s compression techniques. The aforesaid result is achieved by introducing a new hinting mechanism that allows parties to collaboratively calculate (textsf {HighBits}). Our hinting mechanism allows public key compression similar to Dilithium. Additionally, we suggest an optimization technique to minimize number of restarts both parties need to produce a valid signature. Our approach allows to produce (approx 10) KB signatures within 3 rounds of communication. We prove security of our scheme under MLWE and MSIS assumptions in ROM, and provide implementation of our proposed scheme. As additional contribution, we present vulnerabilities and inconsistencies found in Liu et al. work (Future Generation Computer Systems 2023) which aimed to construct distributed lattice-based signature protocol.