Ambiguity Problem Research Articles

Different types of approximation operators on Gn-CAS via ideals

A mathematical approach to dealing with the problems of ambiguity and indeterminacy in knowledge is called a rough set theory. It begins by using an equivalence relation to divide the universe into parts. Numerous generalized rough set models have been developed and investigated to increase their adaptability and extend their range of applications. In this context, we introduce new generalized rough set models that are inspired by covering-based rough sets and ideals. In this paper, lower and upper approximations of new types of covering rough sets based on j-neighborhoods, complementary j-neighborhoods, and j-adhesions are defined via ideals. The main features of these approximations are examined. The relationships among them are given by various examples and propositions. Some comparisons between our methods and others? methods such as Abd El-Monsef et al.?s method [2] and Nawar et al.?s method [22] are given. A practical example is given to illustrate one of our methods is more precise.

A Deep Dive into an AI Startup Company in the Pandemic

ABSTRACT This case allows students to apply managerial accounting concepts and analytical skills to solve a new, relevant, and ambiguous problem. Using the COVID-19 pandemic as the backdrop setting, this case takes students on a roller-coaster ride as they help a CEO of a high-tech startup company cut costs and increase sales to weather a highly uncertain and financially precarious environment. Based on a past budget report, students incorporate uncertainty into the planning process and provide managerial accounting insights to the CEO to make decisions as the company changes its business model in response to the pandemic. They analyze costs, revise the budget, predict cash runway under different scenarios, and integrate nonfinancial considerations with quantitative analysis in the decision-making process. JEL Classifications: L21; L26; M13; M41.

Probabilistic Intersection-Over-Union for Training and Evaluation of Oriented Object Detectors.

Oriented object detection is a challenging and relatively new problem. Most existing approaches are based on deep learning and explore Oriented Bounding Boxes (OBBs) to represent the objects. They are typically based on adaptations of traditional detectors that work with Horizontal Bounding Boxes (HBBs), which have been exploring IoU-like loss functions to regress the HBBs. However, extending this idea for OBBs is challenging due to complex formulations or requirement for customized backpropagation implementations. Furthermore, using OBBs presents limitations for irregular or roughly circular objects, since the definition of the ideal OBB is an ambiguous and ill-posed problem. In this work, we jointly tackle the problem of training, representing, and evaluating oriented detectors. We explore Gaussian distributions-called Gaussian Bounding Boxes (GBBs)-as fuzzy representations for oriented objects and propose using a similarity metric between two GBBs based on the Hellinger distance. We show that this metric leads to a differentiable closed-form expression that can be directly used as a localization loss term to train OBB object detectors. We also show that GBBs present a natural representation as elliptical regions (called EBBs), which inherently mitigate ambiguity representation for circular objects. Finally, we empirically show that the proposed similarity metric computed between two GBBs strongly correlates with the IoU between the corresponding EBBs, motivating the name Probabilistic Intersection-over-Union (ProbIoU). Our experiments show that results using ProbIoU as a regression loss are competitive with state-of-the-art alternatives without requiring additional hyperparameters or customized implementations, and that ProbIoU is a promising alternative to evaluate oriented object detectors. Our code is available at https://github.com/ProbIOU/.

Promoting Student Success with Neutrosophic Sets: Artificial Intelligence and Student Engagement in Higher Education Context

Several higher education institutions recommended active learning many decades ago. In many fields, artificial intelligence (AI) has turned out to be a dominant aspect of people's lives. One of the fields that welcome AI is education. AI can play a crucial role in active learning, which is a teaching method that encourages students to take an active role in their learning process. Active learning can involve a wide range of activities, such as problem-solving, group discussions, and self-reflection. This quantitative research paper aims to explore students' engagement and perceptions of patterns of embracing AI tools and their relationship to student learning outcomes. A diverse sample of 355 students from a highly reputable university in the United Arab Emirates participated in the research study (UAE). Descriptive Analysis and Correlation Coefficient were used to achieve the paper's objectives. Findings uncovered students’ high engagement level in classroom activities, and they perceived AI tools as useful and effortless, which promotes their engagement and attention inside classrooms. Furthermore, the results demonstrated the existence of a positive association between students' perceptions of embracing AI and student engagement in the learning environment. However, students’ learning outcomes have a non-remarkable association with student engagement and their perception of AI patterns. The study's findings suggest embracing and promoting AI applications in classrooms to keep students engaged. This study's use of Neutrosophic sets offers a fresh way to approach the problem of ambiguity when evaluating students' opinions and performance in relation to AI technologies. Neutrosophic sets, a mathematical framework designed to manage uncertainty, provide a sophisticated way to understand how students' complicated interactions with AI operate. This integration promises a more comprehensive and flexible educational paradigm and is a trailblazing method of negotiating the complexities of active learning in higher education.

Improvement of Translation Accuracy for the Word Sense Disambiguation System using Novel Classifier Approach

Machine Translation (MT) is a crucial application of Natural language Processing (NLP). This MT technique automatic and based on computers. One of the most modern techniques adopted in MT is Machine Learning (ML). Over the past few years, ML has grown in popularity during MT process among researchers. Ambiguity is a major challenge in MT. Word Sense Disambiguation (WSD) is a common technique for solving the ambiguity problem. ML approaches are commonly used for the WSD techniques and are used for training and testing purposes. The outcome prediction of the test data gives encouraging results. Text classification is one of the most significant techniques for resolving the WSD. In this paper, we have analyzed some common supervised ML text classification algorithms and also proposed a “hybrid model” called “AmbiF.” We have compared the results of all analyzed algorithms with the proposed model “AmbiF. The analyzed supervised algorithms are Decision Tree (DT), Bayesian network, Support Vector Machines (SVMs), K-Nearest Neighbor (KNN), Random Forest (RF), and Logistic Regression (LR). The range of accuracy for all the algorithms that were examined is between sixty-eight and eighty-four percent. To improve the accuracy of the AmbiF model, we have merged the DT, SVM, and Naïve Bayes (NB)-classifier approach. For testing the model, we have used the ten-fold cross-validation test method. The AmbiF model’s accuracy has been reported eighty-five percent. Comparing the AmbiF model to all other analyzed supervised ML classification algorithms, it has also demonstrated great precision, recall, and F-score. Waikato Environment for Knowledge Analysis (WEKA)’s ML-tool is used to analyze the algorithms and the AmbiF model.

Partial Multilabel Learning Using Noise-Tolerant Broad Learning System With Label Enhancement and Dimensionality Reduction.

Partial multilabel learning (PML) addresses the issue of noisy supervision, which contains an overcomplete set of candidate labels for each instance with only a valid subset of training data. Using label enhancement techniques, researchers have computed the probability of a label being ground truth. However, enhancing labels in the noisy label space makes it impossible for the existing partial multilabel label enhancement methods to achieve satisfactory results. Besides, few methods simultaneously involve the ambiguity problem, the feature space's redundancy, and the model's efficiency in PML. To address these issues, this article presents a novel joint partial multilabel framework using broad learning systems (namely BLS-PML) with three innovative mechanisms: 1) a trustworthy label space is reconstructed through a novel label enhancement method to avoid the bias caused by noisy labels; 2) a low-dimensional feature space is obtained by a confidence-based dimensionality reduction method to reduce the effect of redundancy in the feature space; and 3) a noise-tolerant BLS is proposed by adding a dimensionality reduction layer and a trustworthy label layer to deal with PML problem. We evaluated it on six real-world and seven synthetic datasets, using eight state-of-the-art partial multilabel algorithms as baselines and six evaluation metrics. Out of 144 experimental scenarios, our method significantly outperforms the baselines by about 80%, demonstrating its robustness and effectiveness in handling partial multilabel tasks.

Considering Medical Professionalism from the Standpoint of Pharmacy Education

Worldwide interest in teaching medical professionalism has increased drastically over the past two decades and is recognized as an important core competency. It is also essential in pharmacy education. However, there is no single definition of medical professionalism owing to its multifaceted nature, leading to difficulty in understanding it. The foundational concept of professionalism are the social contract and accountability, which describe the relationship between the profession and the society which it serves. Profession must understand expectations from the society, which is trustworthy, assures competence, and devoted to the public good for the contract based on their mutual trust. In "teaching," three basic educational actions ("setting expectations," "providing experiences," and "evaluating outcomes") are required. There are two learning goals of professionalism education: the minimum goal of not doing unprofessional acts and the aspirational goal of pursuing a higher level of interiorized professionalism which leads to the professional identity formation. The true professionals are "reflective practitioners," who have the ability to manage ambiguous problems using their interiorized professionalism in complicated situations. Therefore, reflection is one of the central concepts of professionalism education. The Professionalism Mini-Evaluation Exercise (P-MEX), an observational tool to evaluate medical professionalism, has some favorable aspects; the Japanese version is available and is a guide to specific actions for professionalism through its items, although some cautions must be exercised when using it. Considering that teaching professionalism includes not only formal but informal and hidden curricula, all of the staff in the educational environments should consider professionalism education by understanding professionalism.

Trial application of the envelope method to the potential ambiguity problem

Trial application of the envelope method to the potential ambiguity problem

Selecting the foremost big data tool to optimize YouTube data in dynamic Fermatean fuzzy knowledge.

Big data pertains to extensive and intricate compilations of information that necessitate the implementation of proficient and cost-effective evaluation and analysis tools to derive insights and support decision making. The Fermatean fuzzy set theory possesses remarkable capability in capturing imprecision due to its capacity to accommodate complex and ambiguous problem descriptions. This paper presents the study of the concepts of dynamic ordered weighted aggregation operators in the context of Fermatean fuzzy environment. In numerous practical decision making scenarios, the term "dynamic" frequently denotes the capability of obtaining decision-relevant data at various time intervals. In this study, we introduce two novel aggregation operators: Fermatean fuzzy dynamic ordered weighted averaging and geometric operators. We investigate the attributes of these operators in detail, offering a comprehensive description of their salient features. We present a step-by-step mathematical algorithm for decision making scenarios in the context of proposed methodologies. In addition, we highlight the significance of these approaches by presenting the solution to the decision making problem and determining the most effective big data analytics platform for YouTube data analysis. Finally, we perform a thorough comparative analysis to assess the effectiveness of the suggested approaches in comparison to a variety of existing techniques.

Metric3D v2: A Versatile Monocular Geometric Foundation Model for Zero-shot Metric Depth and Surface Normal Estimation.

We introduce Metric3D v2, a geometric foundation model for zero-shot metric depth and surface normal estimation from a single image, which is crucial for metric 3D recovery. While depth and normal are geometrically related and highly complimentary, they present distinct challenges. State-of-the-art (SoTA) monocular depth methods achieve zero-shot generalization by learning affine-invariant depths, which cannot recover real-world metrics. Meanwhile, SoTA normal estimation methods have limited zero-shot performance due to the lack of large-scale labeled data. To tackle these issues, we propose solutions for both metric depth estimation and surface normal estimation. For metric depth estimation, we show that the key to a zero-shot single-view model lies in resolving the metric ambiguity from various camera models and large-scale data training. We propose a canonical camera space transformation module, which explicitly addresses the ambiguity problem and can be effortlessly plugged into existing monocular models. For surface normal estimation, we propose a joint depth-normal optimization module to distill diverse data knowledge from metric depth, enabling normal estimators to learn beyond normal labels. Equipped with these modules, our depth-normal models can be stably trained with over 16 million of images from thousands of camera models with different-type annotations, resulting in zero-shot generalization to in-the-wild images with unseen camera settings. Our method currently ranks the 1st on various zero-shot and non-zero-shot benchmarks for metric depth, affine-invariant-depth as well as surface-normal prediction, shown in Fig.1. Notably, we surpassed the ultra-recent MarigoldDepth and DepthAnything on various depth benchmarks including NYUv2 and KITTI. Our method enables the accurate recovery of metric 3D structures on randomly collected internet images, paving the way for plausible single-image metrology. The potential benefits extend to downstream tasks, which can be significantly improved by simply plugging in our model. For example, our model relieves the scale drift issues of monocular-SLAM (Fig.3), leading to high-quality metric scale dense mapping. These applications highlight the versatility of Metric3D v2 models as geometric foundation models. Our project page is at https://JUGGHM.github.io/Metric3Dv2.

A Unified Framework for Contextual and Factoid Question Generation

Question generation (QG) aims to automatically generate fluent and relevant questions, where the two most mainstream directions are generating questions from unstructured contextual texts (CQG), such as news articles, and generating questions from structured factoid texts (FQG), such as knowledge graphs or tables. Existing methods for these two tasks mainly face challenges of limited internal structural information as well as scarce background information, while these two tasks can benefit each other for alleviating these issues. For example, when meeting the entity mention “United Kingdom” in CQG, it can be inferred that it is a country in European continent based on the structural knowledge “(Europe, countries_within, United Kingdom)” in FQG. And when meeting the entity “Houston Rockets” in FQG, more background information, such as “an American professional basketball team based in Houston since 1971”, can be found in the related passages of CQG. To this end, we propose a unified framework for the tasks of CQG and FQG, where: (i) two types of task-sharing modules are developed to learn shared contextual and structural knowledge, where the task format is unified with a pseudo passage reformulation strategy; (ii) for the CQG task, a task-specific knowledge module with a knowledge selection and aggregation mechanism is introduced, so as to incorporate more factoid knowledge from external knowledge graphs and alleviate the word ambiguity problem; and (iii) for the FQG task, a task-specific passage module with a multi-level passage fusion mechanism is designed to extract fine-grained word-level knowledge. Experimental results in both automatic and human evaluation show the effectiveness of our proposed method.

The Beginning And Trend Of Vagueness In Urdu Poetry

In this research paper, an attempt has been made to understand the beginnings of ambiguity in modern Urdu poetry and to understand it as a phenomenon. The ambiguous narrative of the modern poem has been debated in the literary world for decades. In the movement of linguistic formations that started in the sixties, an attempt was made to express a new poetic expression by completely deviating from the tradition. The intellectual and stylistic changes in poetry during this period gave rise to serious problems of ambiguity in modern poetry. The movement of linguistic formations died out within a few years after its failure, but the trend of ambiguity in poetry continued. Therefore, the trend of vagueness has been studied under the intellectual degeneration in the contemporary situation. In this regard, the impact of European literary trends and modern socio-linguistic ideas on Urdu poetry has also been examined.

A Study on Toponymic Entity Recognition Based on Pre-Trained Models Fused with Local Features for Genglubu in the South China Sea

Toponymic entity recognition is currently a critical research hotspot in knowledge graphs. Under the guidance of the national ancient book protection policy and the promotion of the wave of digital humanities research, this paper proposes a toponymic entity recognition model (ALBERT-Conv1D-BiLSTM-CRF) based on the fusion of a pre-trained language model and local features to address the problems of toponymic ambiguity and the differences in ancient and modern grammatical structures in the field of the Genglubu. This model extracts global features with the ALBERT module, fuses global and local features with the Conv1D module, performs sequence modeling with the BiLSTM module to capture deep semantics and long-distance dependency information, and finally, completes sequence annotation with the CRF module. The experiments show that while taking into account the computational resources and cost, this improved model is significantly improved compared with the benchmark model (ALBERT-BiLSTM-CRF), and the precision, recall, and F1 are increased by 0.74%, 1.28%, and 1.01% to 98.08%, 96.67%, and 97.37%, respectively. The model achieved good results in the field of Genglubu.

The Relation Between Attention, Inhibition and Word Learning in Young Children

In their natural environments, children usually see several novel objects while they hear the labels for these objects, making it difficult for them to know exactly which objects these words refer to. This referential ambiguity problem can be alleviated through selective attention and inhibitory control because a child who focuses on plausible referents while inhibiting irrelevant ones during object labelling has a higher chance of identifying the intended referent. The present study examined this hypothesis by testing 3.5- to 4.5-year-old children. In particular, we examined the links between children’s word learning (a cross-situational learning task), selective attention (flanker task), inhibitory control (day-night Stroop task) while controlling for working memory (Corsi block task). We found that children learned the novel word-object associations and completed the cognitive control tasks successfully. However, we did not find any association between word learning and cognitive control or memory span. We argue that the lack of a significant association between cognitive control and word learning may be indicative of a more exploratory style of learning in young children.

Csswin-unet: a Swin-unet network for semantic segmentation of remote sensing images by aggregating contextual information and extracting spatial information

ABSTRACTImage interpretation algorithms based on deep learning are becoming increasingly important in land cover information acquisition. We propose CSSwin-unet, a network designed for the semantic segmentation of remote sensing images. CSSwin-unet is based on Swin-unet, follows the U-shaped codec structure of U-Net, but utilizes Swin transformer blocks with superior global modelling capabilities to constitute the codec. In addition, we design a parallel branch in the encoder with a context aggregation module (CAM) to enhance contextual information extraction and alleviate the semantic ambiguity problem resulting from occlusion. To address the problem of semantic information mismatch between codecs and improve the model’s ability to extract spatial information, we constructed a space extraction module (SEM) in the skip connections, which replaces the direct copying of encoder features in Swin-unet. To reduce information loss during the downsampling process and strengthen the segmentation capacity of the network, we designed a feature shrinkage module (FSM) in the downsampling session. We conducted comprehensive ablation experiments on a dataset we produced ourselves and compared the results with other advanced methods. The test results showed significant improvement, with mIoU, mF1, and OA values improving by 2.83%, 2.47%, and 2.05%, respectively, compared to the second best performing model, Swin-unet. The above results prove the excellent performance of CSSwin-unet.

SOZIL: Self-Optimal Zero-Shot Imitation Learning

Zero-shot imitation learning has demonstrated its superiority to learn complex robotic tasks with less human participation. Recent studies show convincing performance under the condition that the robot follows the demonstration strictly by the learned inverse model. However, these methods are difficult to achieve satisfactory performance in imitation when the demonstration is suboptimal, and the learning of the learned inverse models is vulnerable to label ambiguity issues. In this paper, we propose Self-Optimal Zero-shot Imitation Learning (SOZIL) to tackle these problems. The contribution of SOZIL is twofold. First, Goal Consistency Loss (GCL) is designed to learn the multi-step goal-conditioned policy from exploration data. By directly using the goal state as supervision, GCL solves the label ambiguity problem caused by trajectory and action diversity. Second, Estimation-based Keyframe Extraction (EKE) is developed to optimize demonstrations. We formulate the keyframe extraction process as a path optimization problem under suboptimal control. By predicting the performance of the learned policy in executing transitions of any two states, EKE creates a directed graph containing all candidate paths and extracts keyframes by solving the graph’s shortest path problem. Furthermore, the proposed method is evaluated with various simulated and real-world robotic manipulating experiments such as cable harness assembly, rope manipulation, and block moving. Experimental results show that SOZIL achieves a superior success rate and manipulation efficiency than baselines.

QSCC: A Quaternion Semantic Cell Convolution Graph Neural Network for MBD Product Model Classification

Model-based Definition (MBD) 3D engineering has gained wide popularity which enables smart manufacturing. Its tip orientation to the effective reuse of MBD-based product models for creating efficient digital twins. The deep learning-based 3D model classification is an important way in the most recent years, but the existing methods only concern geometric information with very little or no manufacturing information. In this paper, we proposed quaternion semantic cell convolution (QSCC), the quaternion semantic cell convolution graph neural network for MBD product model classification. It is mainly composed of the quaternion product semantic cell(QPSC) and the quaternion simple graph convolution(QSGC). The QPSC is established to fully express the digital twins' semantic information, which includes the multi-dimensional PMI information of the MBD model. The QSGC is established by embedding QPSC into the quaternion vector space, which serializes the multi-granularity semantic features to learn the context semantics through GRU. Compared with existing methods, our approach can solve the problem of model classification with similar geometry but different functions. Specifically, our approach can solve the ambiguity problem caused by blend features such as chamfer. We give quantitative and qualitative evaluations to demonstrate the accuracy and efficiency.

Hybrid-MVS: Robust Multi-View Reconstruction With Hybrid Optimization of Visual and Depth Cues

Consumer-level RGB-D cameras have been widely used for dense 3D reconstruction of scenes. Especially for textureless or non-lambertian surfaces, consumer RGB-D cameras can ensure completeness of the reconstructed models at a low cost. However, the reconstruction quality relies heavily on the accuracy of the depth sensors. Digital cameras are also used popularly for capturing high-resolution pictures to achieve high-quality dense reconstruction of the scenes, but cannot handle textureless or non-lambertian regions well due to the visual ambiguity problem. To ensure both completeness and accuracy of the reconstructed 3D models, we propose a hybrid multi-view reconstruction pipeline named Hybrid-MVS, which combines the high-resolution images taken by a digital camera and the low-resolution RGB-D frames captured by a consumer RGB-D camera for robust reconstruction of complicated scenes with challenging textureless and non-lambertian surfaces. Unlike most existing multi-sensor systems which require explicit hardware calibration and synchronization of various sensors, the calibration and synchronization problems between the digital camera and RGB-D camera are implicitly solved for compositing reliable depth prior of the digital images in our pipeline. Especially, we propose a hybrid MVS framework for robust PatchMatch stereo and Delaunay meshing, which tightly couples both visual cues given by the digital images and depth cues from the RGB-D frames to maximize the complementary advantages. The experiments with quantitative and qualitative evaluations demonstrate the effectiveness of the proposed Hybrid-MVS framework, which can successfully achieve high-quality 3D reconstruction of complicated natural scenes with robustness to weakly textured and non-lambertian areas.

Distance-based Weighted Transformer Network for image completion

The challenge of image generation has been effectively modeled as a problem of structure priors or transformation. However, existing models have unsatisfactory performance in understanding the global input image structures because of particular inherent features (for example, local inductive prior). Recent studies have shown that self-attention is an efficient modeling technique for image completion problems. In this paper, we propose a new architecture that relies on Distance-based Weighted Transformer (DWT) to better understand the relationships between an image’s components. In our model, we leverage the strengths of both Convolutional Neural Networks (CNNs) and DWT blocks to enhance the image completion process. Specifically, CNNs are used to augment the local texture information of coarse priors and DWT blocks are used to recover certain coarse textures and coherent visual structures. Unlike current approaches that generally use CNNs to create feature maps, we use the DWT to encode global dependencies and compute distance-based weighted feature maps, which substantially minimizes the problem of visual ambiguities. Meanwhile, to better produce repeated textures, we introduce Residual Fast Fourier Convolution (Res-FFC) blocks to combine the encoder’s skip features with the coarse features provided by our generator. Furthermore, a simple yet effective technique is proposed to normalize the non-zero values of convolutions, and fine-tune the network layers for regularization of the gradient norms to provide an efficient training stabilizer. Extensive quantitative and qualitative experiments on three challenging datasets demonstrate the superiority of our proposed model compared to existing approaches.

Α tool for requirements engineering using ontologies and boilerplates

The most popular technique for specification requirements is natural language. The disadvantage of natural language is ambiguity. Boilerplates are syntactic patterns which limit the ambiguity problem associated with using natural language to specify system/software requirements. Also, using boilerplates is considered a useful tool for inexperienced engineers to define requirements. Using linguistic boilerplates, constrains the natural language syntactically. Furthermore, a domain-specific ontology is used to constrain requirements semantically, as well. In requirements specification, using ontologies helps to restrict the vocabulary to entities, properties, and property relationships which are semantically related. The above results in avoiding or making fewer mistakes. This work makes use of the combination of boilerplate and ontology. Usually, the attributes of boilerplates are completed with the help of the ontology. The contribution of this paper is that the whole boilerplates is stored in the ontology and attributes and fixed elements are part of the ontology. This combination helps to correct semantically and syntactically requirement construction. This paper proposes a tool based on a domain-specific ontology and a set of predefined generic linguistic boilerplates for requirements engineering. We create a domain-specific ontology and a minimal set of boilerplates for an ATM (Automated Teller Machine). We carried out an experiment in order to obtain evidence for the effectiveness and efficiency of our method. The experiment took the form of a case study for the ATM domain and our proposed method was evaluated by users. The contribution and novelty of our methodology is that we created a tool for defining requirements that integrates boilerplate templates and an ontology. We exploit the boilerplate language syntax, mapping them to Resource Description Framework triples which have also a linguistic nature.