Incorporating accountability mechanisms in online services requires effective trust management and immutable, traceable source of truth for transaction evidence. The emergence the blockchain technology brings high hopes fulfilling most those requirements. However, a major challenge is to find proper consensus protocol that applicable crowdsourcing particular general. Building upon idea using as underlying enable tracing transactions service contracts dispute arbitration, this paper proposes...

The semantic Web can make e-commerce interactions more flexible and automated by standardizing ontologies, message content, protocols. This paper investigates how Services technologies be used to support service advertisement discovery in e-commerce. In particular, it describes the design implementation of a matchmaking prototype that uses DAML -S based ontology description logic reasoner compare ontology-based descriptions. By representing semantics descriptions, matchmaker enables behavior...

Naive Bayes algorithm is one of the most effective methods in field text classification, but only large training sample set can it get a more accurate result. The requirement number samples not brings heavy work for previous manual also puts forward higher request storage and computing resources during computer post-processing. This paper mainly studies Naïve classification based on Poisson distribution model, experimental results show that this method keeps high accuracy even small set.

Traditional neural network based short text classification algorithms for sentiment is easy to find the errors. In order solve this problem, Word Vector Model (Word2vec), Bidirectional Long-term and Short-term Memory networks (BiLSTM) convolutional (CNN) are combined. The experiment shows that accuracy of CNN-BiLSTM model associated with Word2vec word embedding achieved 91.48%. This proves hybrid performs better than single structure in text.

Zehui Lin, Liwei Wu, Mingxuan Wang, Lei Li. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

Robust segmentation in adverse weather conditions is crucial for autonomous driving. However, these scenes struggle with recognition and make annotations expensive, resulting poor performance. As a result, the Segment Anything Model (SAM) was recently proposed to finely segment spatial structure of provide powerful prior information, thus showing great promise resolving problems. SAM cannot be applied directly different geographic scales non-semantic outputs. To address issues, we propose...

Introduction Traditional Graph Pattern Matching (GPM) research mainly focuses on improving the accuracy and efficiency of complex network analysis fast subgraph retrieval. Despite their ability to return subgraphs quickly accurately, these methods are limited applications without medical data research. Methods In order overcome this limitation, based existing GPM with lung cancer knowledge graph, paper introduces Monte Carlo method proposes an edge-level multi-constraint graph pattern...

Introduction Traditional Graph Pattern Matching (GPM) research mainly focuses on improving the accuracy and efficiency of complex network analysis fast subgraph retrieval. Despite their ability to return subgraphs quickly accurately, these methods are limited applications without medical data research. Methods In order overcome this limitation, based existing GPM with lung cancer knowledge graph, paper introduces Monte Carlo method proposes an edge-level multi-constraint graph pattern...

The advancement of the underlying technology and hardware devices Internet Things (IoT) has led to emergence several new applications that are influencing progress human society in era Artificial Intelligence IoT (AIoT). application AIoT not only revolutionized efficiency social life but also brought about moral ethical hazards as well legal issues. At a technological level, algorithmic mechanism AI driving is more complex than typical machine learning requires regulation because its...

This observational study examines the feasibility and safety of a new preloaded robot-assisted thrombectomy system specifically designed for mechanical thrombectomy.

We present a new, embarrassingly simple approach to instance segmentation in images. Compared many other dense prediction tasks, e.g., semantic segmentation, it is the arbitrary number of instances that have made much more challenging. In order predict mask for each instance, mainstream approaches either follow 'detect-thensegment' strategy as used by Mask R-CNN, or category masks first then use clustering techniques group pixels into individual instances. view task from completely new...

Text-based question answering (TBQA) has been studied extensively in recent years. Most existing approaches focus on finding the answer to a within single paragraph. However, many difficult questions require multiple supporting evidence from scattered text among two or more documents. In this paper, we propose Dynamically Fused Graph Network(DFGN), novel method those requiring and reasoning over them. Inspired by human's step-by-step behavior, DFGN includes dynamic fusion layer that starts...