Twitter (now X) has been gaining popularity with each passing day since its inception in 2006. People have been using Twitter as an instant repository to collect data and gain insight into folks’ minds on trending is...
Twitter (now X) has been gaining popularity with each passing day since its inception in 2006. People have been using Twitter as an instant repository to collect data and gain insight into folks’ minds on trending issues. Although Twitter allows access to its data through streaming and rest APIs, extracting the required data is difficult. The data (tweet) returned by Twitter is in .json format, having at least 26 fields. Each field is bundled in the dictionary form data structure. tweet metadata such as tweet “likes” and “retweets” increase the volume and complexity of a tweet, making data extraction in a cleaner format more difficult. This work aims to develop an effective two-step tweet extractor (Xtractor: an online-offline keyword-based Twitter data extractor). In the first step, Xtractor collects publicly available topical tweets using keywords, hashtags, or their list on the fly, parses all the tweet fields, and filters them to retain potential field contents without additional payload. The second step of the Xtractor evicts partially matching tweets using the regular expression method to acquire the targeted domain-specific tweets. Using the proposed two-step Twitter data extraction method, datasets concerning the people of Pakistan have been collected, which can be leveraged to get insight for decision-making, specially in the context of sentiment analysis. We found this method efficient, capacitive, and productive.
Throughput analysis for successive interference cancellation-based two-device slotted ALOHA with feedback is studied over Nakagami-m fading channels. Explicit expressions for the state transition probabilities are der...
详细信息
The accelerometer signal is capable of providing crucial information regarding object motion, posture, and vibration. Therefore, it is of great significance to investigate noise suppression techniques for acceleromete...
详细信息
Ontologies are an essential component of semantic integration approaches for information systems . In a decentralized environment, each specification of the domain reflects an Ontological view. However, the semantics ...
详细信息
We reported on the progress of fast hybrid method (FHM) for full- wave simulations of propagation of L-band microwaves in forested environment. For L band, previously we performed full wave simulations of realistic tr...
详细信息
ISBN:
(数字)9798350360325
ISBN:
(纸本)9798350360332
We reported on the progress of fast hybrid method (FHM) for full- wave simulations of propagation of L-band microwaves in forested environment. For L band, previously we performed full wave simulations of realistic trees initially at 8 meters [1], followed by 13 meters [2]. The progress in this work is at comparisons of the electromagnetic model simulations with SMAPVEX19-22 data: 1) The height of the trees have been extended to 17 meters with the multiple scattering effects of 91 trees in the spatial domain with simulated transmissivity at 0.57 2) the spatial patterns of electric field distribution are simulated with electric field as high as 1.6 that of the incident wave corresponding to 2.56 times the Poynying of the incident waves, and the patterns exhibits gaps and shadows 3) the effects of clustering of trees with gaps show different results from that of uniformly positioned trees and 4) tree structures are varied with examples of trees with two trunks branching out from the main trunk, and the case of tapering trunk radius.
Positron Emission Tomography (PET) and Structural Magnetic Resonance Imaging (sMRI) are widely used for early Alzheimer's disease (AD) diagnosis, providing anatomical and metabolic insights. However, current fusio...
详细信息
ISBN:
(数字)9798331520526
ISBN:
(纸本)9798331520533
Positron Emission Tomography (PET) and Structural Magnetic Resonance Imaging (sMRI) are widely used for early Alzheimer's disease (AD) diagnosis, providing anatomical and metabolic insights. However, current fusion methods, like simple addition or concatenation, fail to leverage their complementary information fully. To address it, we proposed a 3D multi-modal feature interaction fusion network for AD diagnosis. Specifically, we leverage a robust 3D ResNet backbone network as the architecture for feature extraction. To enable the model to dynamically focus on the areas of interest, we added a local attention module at each layer of the network. We further introduced a collaborative attention module to facilitate the interaction and fusion of features between PET and sMRI modalities, thereby enhancing the overall capability of the model to utilize complementary information from both modalities. Finally, an enhanced feature fusion module based on Transformer is integrated into the network to further strengthen the representation of the fused features. The experiment shows that our model performs better than other representative models and validates the effectiveness of the proposed method on the ADNI dataset.
In this paper, a 50-kW 1-kV/6.25-kV medium voltage dual-active-bridge (DAB) transformer has been designed and tested. The shell-type transformer structure with stacked small cores is selected to integrate the leakage ...
详细信息
Nowadays, people developed various convolutional neural network (CNN) based models for computer vision. Some famous models, such as GoogLeNet, Residual Network (ResNet), Visual Geometry Group (VGG), and You Only Look ...
详细信息
Graph Contrastive Learning (GCL) has emerged as a highly effective self-supervised approach in graph representation learning. However, prevailing GCL methods confront two primary challenges: 1) They predominantly oper...
Graph Contrastive Learning (GCL) has emerged as a highly effective self-supervised approach in graph representation learning. However, prevailing GCL methods confront two primary challenges: 1) They predominantly operate under homophily assumptions, focusing on low-frequency signals in node features while neglecting heterophilic edges that connect nodes with dissimilar features. 2) Their reliance on neighborhood aggregation for inference leads to scalability challenges and hinders deployment in real-time applications. In this paper, we introduce S3GCL, an innovative framework designed to tackle these challenges. Inspired by spectral GNNs, we initially demonstrate the correlation between frequency and homophily levels. Then, we propose a novel cosine-parameterized Chebyshev polynomial as low/high-pass filters to generate biased graph views. To resolve the inference dilemma, we incorporate an MLP encoder and enhance its awareness of graph context by introducing structurally and semantically neighboring nodes as positive pairs in the spatial domain. Finally, we formulate a cross-pass GCL objective between fullpass MLP and biased-pass GNN filtered features, eliminating the need for augmentation. Extensive experiments on real-world tasks validate S3GCL proficiency in generalization to diverse homophily levels and its superior inference efficiency.
暂无评论