Cloud computing platforms support the Internet of Vehicles, but the main bottlenecks are high latency and massive data transmission in cloud-based processing. Vehicular fog computing has emerged as a promising paradigm to accommodate the increasing computational needs of vehicles. It provides low latency network services that are most important for latency-sensitive tasks. The dynamic nature of VFC, having vehicles with heterogeneous computing resources, vehicle mobility, and diverse tasks with different priorities are the main challenges in vehicular fog networks. In VFC, vehicles can share their idle compute resources with other task-generating vehicles. So, scheduling the tasks on the idle resources of resource-limited vehicles is very important. Existing solutions use a heuristic approach to solve this issue but lack generalizability and adaptability. In this paper, we describe a PPO-based intelligent, priority and deadline-aware online and distributed resource allocation and task scheduling algorithm, called IRATS, in vehicular fog networks. IRATS formulates the resource allocation problem as a Markov decision process to minimize the waiting time and delay of tasks. For vehicles sharing their idle resources, we design a task scheduler for the orderly execution of received tasks according to their priorities using multi-level queues. We conducted extensive simulations using SUMO, OMNeT++, Veins, and veins-gym to validate the effectiveness of the presented algorithm. The simulation results confirm that the proposed algorithm improves the percentage of in-time completed tasks and decreases the packet loss, waiting time, and end-to-end delay as compared to random, A2C, and DQN algorithms considering the task priority and link duration of vehicles.

The recent success of neural networks in NLP applications has provided a strong impetus to develop supervised models for semantic role labeling (SRL) that forego the requirement for extensive feature engineering. Recent state-of-the-art approaches require high-quality annotated datasets that are costly to obtain and almost unavailable for low-resource languages. We present a semi-supervised approach that utilizes both labeled and unlabeled data to provide performance improvement over a mere supervised SRL model. We show that our proposed semi-supervised SRL model provides larger improvement over a supervised model in the scenario where labeled training data size is small. Our SRL system leverages unlabeled data under the language modeling paradigm. We demonstrate that the incorporation of a self pre-trained bidirectional language model (S-PrLM) into a SRL system can help in SRL performance improvement by learning composition functions from the unlabeled data. Previous researches have concluded that syntax information is very useful for high-performing SRL systems, so we incorporate syntax information by employing an unsupervised approach to leverage dependency path information to connect argument candidates in vector space, which helps in distinguishing arguments with similar contexts but different syntactic functions. The basic idea is to connect predicate ( w p ) with argument candidate ( w a ) with the dependency path ( r ) between them in the embedding space. Experiments on the CoNLL-2008 and CoNLL-2009 datasets confirm that our full SRL model outperforms previous best models in terms of F 1 score.

Implicit discourse relation recognition is a challenging task due to the absence of the necessary informative clues from explicit connectives. An implicit discourse relation recognizer has to carefully tackle the semantic similarity of sentence pairs and the severe data sparsity issue. In this article, we learn token embeddings to encode the structure of a sentence from a dependency point of view in their representations and use them to initialize a baseline model to make it really strong. Then, we propose a novel memory component to tackle the data sparsity issue by allowing the model to master the entire training set, which helps in achieving further performance improvement. The memory mechanism adequately memorizes information by pairing representations and discourse relations of all training instances, thus filling the slot of the data-hungry issue in the current implicit discourse relation recognizer. The proposed memory component, if attached with any suitable baseline, can help in performance enhancement. The experiments show that our full model with memorizing the entire training data provides excellent results on PDTB and CDTB datasets, outperforming the baselines by a fair margin.

Camera-based multiclass (1-D and 2-D) barcode detectors that can help in decoding barcodes in different complex scenes have huge potential applications in situations where Internet of Things (IoT) is combined with artificial intelligence (AI) and augmented reality (AR). The decoding rate in such applications under real-life complex scenes is greatly affected by two major factors: first, we cannot accurately localize the barcodes, and second, we cannot decode the blur samples. In this article, we first propose a barcode localization algorithm that is capable of regressing four vertices of barcodes accurately. Our localization method comprises an anchor-free approach that outputs multiscale output prediction maps. These segmentation-like maps of each scale are then further divided into three types of maps (classification, centerness, and 8-D regression). Eight-dimensional localization result of barcodes along with classification result is then obtained after postprocessing. Second, we propose a conditional generative adversarial network-based model for deblurring blur QR codes. Extensive decoding experiments on a challenging complex scene data set show that our localization and deblurring methods can contribute to improving the decoding rate of existing barcode decoders.

The task of semantic role labeling ( SRL ) is dedicated to finding the predicate-argument structure. Previous works on SRL are mostly supervised and do not consider the difficulty in labeling each example which can be very expensive and time-consuming. In this article, we present the first neural unsupervised model for SRL. To decompose the task as two argument related subtasks, identification and clustering, we propose a pipeline that correspondingly consists of two neural modules. First, we train a neural model on two syntax-aware statistically developed rules. The neural model gets the relevance signal for each token in a sentence, to feed into a BiLSTM, and then an adversarial layer for noise-adding and classifying simultaneously, thus enabling the model to learn the semantic structure of a sentence. Then we propose another neural model for argument role clustering, which is done through clustering the learned argument embeddings biased toward their dependency relations. Experiments on the CoNLL-2009 English dataset demonstrate that our model outperforms the previous state-of-the-art baseline in terms of non-neural models for argument identification and classification.

Semantic role labeling (SRL) aims at elaborating the meaning of a sentence by forming a predicate-argument structure. Recent researches depicted that the effective use of syntax can improve SRL performance. However, syntax is a complicated linguistic clue and is hard to be effectively applied in a downstream task like SRL. This work effectively encodes syntax using adaptive convolution which endows strong flexibility to existing convolutional networks. The existing CNNs may help in encoding a complicated structure like syntax for SRL, but it still has shortcomings. Contrary to traditional convolutional networks that use same filters for different inputs, adaptive convolution uses adaptively generated filters conditioned on syntactically-informed inputs. We achieve this with the integration of a filter generation network which generates the input specific filters. This helps the model to focus on important syntactic features present inside the input, thus enlarging the gap between syntax-aware and syntax-agnostic SRL systems. We further study a hashing technique to compress the size of the filter generation network for SRL in terms of trainable parameters. Experiments on CoNLL-2009 dataset confirm that the proposed model substantially outperforms most previous SRL systems for both English and Chinese languages.

Implicit discourse relation classification (IDRC) is considered the most difficult component of shallow discourse parsing as the relation prediction in the absence of necessary clues requires a deep understanding of the context information of the sentences. Convolutional neural networks (CNNs) have emerged as an important encoding block for sentences in natural language processing (NLP). CNNs use a specific set of filters for the inputs which may lead to the partial coverage of contextual clues. Furthermore, conventional CNNs may not allow the initial communication between the sentences which is a crucial step for IDRC. We present an adaptive convolution approach for IDRC that utilizes context aware filters for the convolution operation. The goal is to abstract the context of sentences in the filters and let them interact with sentence representations, i.e. learning the representations through learned filters. Our model acts as a cross questioning agent by generating filters from one argument and convolving them with the other for the IDRC task. This process is analogous to the attention mechanism because both methods aim at abstracting contextual information. Different from the attention mechanism, our approach directly encodes the contextual representations in the form of filters and allows the initial communication between arguments during encoding. Furthermore, the adaptive convolution can also work alongside the attention mechanism to enhance the representational ability of the adaptive CNN encoder. Experiments on PDTB 2.0 and CDTB datasets show that our approach outperforms all the baselines by a fair margin and achieves excellent results.

Summary Due to an ever‐increasing number of Internet of Everything (IoE) devices, massive amounts of data are produced daily. Cloud computing offers storage, processing, and analysis services for handling of such large quantities of data. The increased latency and bandwidth consumption is not acceptable to real‐time applications like online gaming, smart health, video surveillance, etc. Fog computing has emerged to overcome the increase in latency and bandwidth consumption in Cloud computing. Fog Computing provides storage, processing, networking, and analytical services at the edge of a network. As Fog Computing is still in its infancy, its significant challenges include resource‐allocation and job‐scheduling. The Fog devices at the edge of the network are resource‐constrained. Therefore, it is important to decide the assignment and scheduling of a job on a Fog node. An efficient job scheduling algorithm can reduce energy consumption and response time of an application request. In this paper, we propose a novel Fog computing scheduler that supports service‐provisioning for Internet of Everything, which optimizes delay and network usage. We present a case study to optimally schedule the requests of Internet of Everything devices on Fog devices and efficiently address their demands on available resources on every Fog device. We consider delay and energy consumption as performance metrics and evaluate the proposed scheduling algorithm using iFogSim in comparison with existing approaches. The results show that the delay and network usage of the proposed scheduler improve by 32% and 16%, respectively, in comparison with FCFS approach.

SUMMARY Grid applications move large amounts of data between distributed resources, and the efficiency of a Grid depends on their timely delivery within given bounds (deadlines). In most cases, the data volume and deadline are known in advance, allowing for both network planning and connection admission control (textrmCAC). We formally define the problem and, based on this formalization, describe the operation of a feasible procedure for network reservations of deadline‐constrained bulk data transfer requests. The procedure guarantees a minimum bandwidth to meet the deadlines and allows for opportunistic utilization of residual network capacity. We propose a novel analytical model based on the solution of anM/M(nc)/1/k(s)−RPSqueue. The analytical model is validated againstns−2 simulations taking into account network level details (IP and TCP protocols), showing remarkably good coherence even under heavy loads. The model is orders of magnitude faster than simulation, which enables its application to plan the capacity of Grid networks, and to enforce CAC under the hypothesis of a dominating bottleneck on the transfer route. Copyright © 2011 John Wiley & Sons, Ltd.

Grid applications need to move large amounts of data between distributed resources within deterministic timeframes. In most cases it is possible to specify the volume and the deadline in advance. We formally define and analyze a mechanism of network reservations of bulk data transfer requests having opportunistic utilization of residual network capacity and analyze it using an M/M/1/N-RPS queue. We compare the analytical results of our mechanism with the analytical results obtained from an M/M/N/N queue, for the case in which there is no opportunistic sharing of residual capacity. We validate the analytical results through simulations in C++. The analytical model is also validated against ns-2 simulations taking into account network level details (IP and TCP protocols), showing remarkably good coherence even under heavy loads. The model is orders of magnitude faster than simulation, which enables its application for capacity planning of Grid networks (also known as network dimensioning), and to enforce connection admission control (CAC) under the practical hypothesis of a dominating bottleneck on the transfer route.