1,349 Hits in 4.1 sec

Neural Graph Embedding Methods for Natural Language Processing [article]

Shikhar Vashishth
2020 arXiv   pre-print
Recently, Graph Convolutional Networks (GCNs) have been proposed to address this shortcoming and have been successfully applied for several problems.  ...  In the second part of the thesis, we utilize GCNs for Document Timestamping problem and for learning word embeddings using dependency context of a word instead of sequential context.  ...  Convolution based framework for incorporating diverse semantic information in learned embeddings.  ... 
arXiv:1911.03042v3 fatcat:fruw547yxnev5pmnlij76wovcy

Visualising Argumentation Graphs with Graph Embeddings and t-SNE [article]

Lars Malmqvist, Tommy Yuan, Suresh Manandhar
2021 arXiv   pre-print
It shows that such a visualisation approach can work for argumentation and show interesting structural properties of argumentation graphs, opening up paths for further research in the area.  ...  This paper applies t-SNE, a visualisation technique familiar from Deep Neural Network research to argumentation graphs by applying it to the output of graph embeddings generated using several different  ...  After the model had been trained for 4 hours, the training was stopped and the output of the last convolutional layer was used as a graph embedding for visualisation by extracting the raw features of the  ... 
arXiv:2107.00528v1 fatcat:wrq7ftvefbdofmfqg6eyzmdgrm

Semi-Supervised Node Classification by Graph Convolutional Networks and Extracted Side Information [article]

Mohammad Esmaeili, Aria Nosratinia
2020 arXiv   pre-print
First, this paper provides a method for extracting side information from a graph realization.  ...  Indeed, the extracted side information can be replaced by a sequence of side information that is independent of the graph structure.  ...  Extracting Side Information For extracting side information that is as much as possible independent from the output of the GCN block, the side information is extracted either from the given feature matrix  ... 
arXiv:2009.13734v2 fatcat:tqk3kcb2abhc3igulyavifxuvi

HybridTabNet: Towards Better Table Detection in Scanned Document Images

Danish Nazir, Khurram Azeem Hashmi, Alain Pagani, Marcus Liwicki, Didier Stricker, Muhammad Zeshan Afzal
2021 Applied Sciences  
Tables in document images are an important entity since they contain crucial information. Therefore, accurate table detection can significantly improve the information extraction from documents.  ...  Our two-stage table detector uses the ResNeXt-101 backbone for feature extraction and Hybrid Task Cascade (HTC) to localize the tables in scanned document images.  ...  We used ResNeXt-101 as a backbone for feature extraction with Cardinality = 64 and bottleneck width = 4d.  ... 
doi:10.3390/app11188396 fatcat:uwzdjpyfmnbnxbxzamgbqkvrji

EX-Action: Automatically Extracting Threat Actions from Cyber Threat Intelligence Report Based on Multimodal Learning

Huixia Zhang, Guowei Shen, Chun Guo, Yunhe Cui, Chaohui Jiang, Liguo Zhang
2021 Security and Communication Networks  
To address this problem, we propose EX-Action, a framework for extracting threat actions from CTI reports.  ...  At the same time, a metric is used to evaluate the information completeness of the extracted action obtained by EX-Action.  ...  Action Feature Extraction. In this module, EX-Action extracts five types of features for each action. e extraction framework of action's features is shown in Figure 3 .  ... 
doi:10.1155/2021/5586335 fatcat:f56ms2kidfcrbmxio3lvyw7oie

Predictive Modeling Applied to Structured Clinical Data Extracted from Electronic Health Records: An Architectural Hypothesis and A First Experiment

Alessandra Pieroni, Alessandro Cabroni, Francesca Fallucchi, Noemi Scarpato
2021 Journal of Computer Science  
We try the architecture by slightly increasing both cardinalities of datasets and extracted features.  ...  Second module produces predictions and it implements alternatively one from Graph Convolutional Network, Simplified Graph Transduction Game, Nearest Nodes and Classes Graph.  ...  Then, second module (based on Graph Convolutional Network (GCN), Simplified Graph Transduction Game (SGTG) or Nearest Nodes and Classes Graph (NNCG)) makes a prediction on a particular patient.  ... 
doi:10.3844/jcssp.2021.762.775 fatcat:da2fhk54wjcfxaynwr6fnadqtm

Semi-supervised learning of hierarchical representations of molecules using neural message passing [article]

Hai Nguyen, Shin-ichi Maeda, Kenta Oono
2017 arXiv   pre-print
In this paper, we propose an unsupervised hierarchical feature extraction algorithm for molecules (or more generally, graph-structured objects with fixed number of types of nodes and edges), which is applicable  ...  With the rapid increase of compound databases available in medicinal and material science, there is a growing need for learning representations of molecules in a semi-supervised manner.  ...  Graph convolutions, which are extensions of convolution operation from multi-dimensional arrays like images or texts to arbitrary graphs, are attracting much attention.  ... 
arXiv:1711.10168v2 fatcat:g2nk7v7lgjcpppjpr2as5yzhti

A Survey of Deep Learning Approaches for OCR and Document Understanding [article]

Nishant Subramani and Alexandre Matton and Malcolm Greaves and Adrian Lam
2021 arXiv   pre-print
In this survey paper, we review different techniques for document understanding for documents written in English and consolidate methodologies present in literature to act as a jumping-off point for researchers  ...  Documents are a core part of many businesses in many fields such as law, finance, and technology among others.  ...  ., 2011) Information Extraction The goal of information extraction for document understanding is to take documents that may have diverse layouts and extract information into a structured format.  ... 
arXiv:2011.13534v2 fatcat:4d3htgyqyjbh7loxrovy7fe5w4

Graph-Based Siamese Network for Authorship Verification

Daniel Embarcadero-Ruiz, Helena Gómez-Adorno, Alberto Embarcadero-Ruiz, Gerardo Sierra
2022 Mathematics  
We model the documents in a graph representation and then a graph neural network extracts relevant features from these graph representations.  ...  We propose a Siamese Network architecture composed of graph convolutional networks along with pooling and classification layers.  ...  Acknowledgments: The authors thank CONACYT for the computer resources provided through the INAOE Supercomputing Laboratory's Deep Learning Platform for Language Technologies.  ... 
doi:10.3390/math10020277 fatcat:b45k7hhb6fgjpioku74d6tnrai

Confidence-based Graph Convolutional Networks for Semi-Supervised Learning [article]

Shikhar Vashishth, Prateek Yadav, Manik Bhandari, Partha Talukdar
2019 arXiv   pre-print
Recently, Graph Convolutional Networks (GCNs) have achieved impressive performance on the graph-based SSL task.  ...  Graph-based Semi-Supervised Learning (SSL) methods aim to address this problem by labeling a small subset of the nodes as seeds and then utilizing the graph structure to predict label scores for the rest  ...  al., 2015) is a generalization of RNN framework which can be used for graph-structured data. • GPNN (Liao et al., 2018 ) is a graph partition based algorithm which propagates information after partitioning  ... 
arXiv:1901.08255v2 fatcat:mhtsuthbybhwzpllgtajvdjkcq

Robust Spatial Filtering With Graph Convolutional Neural Networks

Felipe Petroski Such, Shagan Sah, Miguel Alexander Dominguez, Suhas Pillai, Chao Zhang, Andrew Michael, Nathan D. Cahill, Raymond Ptucha
2017 IEEE Journal on Selected Topics in Signal Processing  
As opposed to spectral methods, our framework, which we term Graph-CNNs, defines filters as polynomials of functions of the graph adjacency matrix.  ...  The simplicity and elegance of the convolutional filtering process makes them perfect for structured problems such as image, video, or voice, where vertices are homogeneous in the sense of number, location  ...  DeepWalk [26] learns latent representations of graph vertices by using random walks to extract local information which encodes structural regularities in social networks.  ... 
doi:10.1109/jstsp.2017.2726981 fatcat:sj2bq77u2faateg3nnnl75sewu

Convolutional Neural Networks Via Node-Varying Graph Filters [article]

Fernando Gama, Geert Leus, Antonio G. Marques, Alejandro Ribeiro
2018 arXiv   pre-print
The proposed design replaces the classical convolution not with a node-invariant graph filter (GF), which is the natural generalization of convolution to graph domains, but with a node-varying GF.  ...  This filter extracts different local features without increasing the output dimension of each layer and, as a result, bypasses the need for a pooling stage while involving only local operations.  ...  This is the length for which, when using two layers, the information corresponding to the whole graph can be aggregated.  ... 
arXiv:1710.10355v2 fatcat:jkk3h2zap5apnlus4vsiv6rbyu

Towards Complex Product Line Variability Modelling: Mining Relationships from Non-Boolean Descriptions

Jessie Carbonnel, Marianne Huchard, Clémentine Nebut
2019 Journal of Systems and Software  
This approach is based on Formal Concept Analysis and Pattern Structures, two mathematical frameworks for knowledge discovery that bring theoretical foundations to complex variability extraction algorithms  ...  In this paper, we propose an approach to extract complex variability information, i.e., involving features as well as multi-valued attributes and cardinalities, in the form of logical relationships.  ...  Our approach offers a mathematical framework for complex variability information extraction based on a unique and canonical structure.  ... 
doi:10.1016/j.jss.2019.06.002 fatcat:m272yunvpja6vot32oy5u73hoq

2020 Index IEEE Transactions on Pattern Analysis and Machine Intelligence Vol. 42

2021 IEEE Transactions on Pattern Analysis and Machine Intelligence  
Zhang, R., +, TPAMI Feb. 2020 291-303 Document image processing Baselines Extraction from Curved Document Images via Slope Fields Recovery.  ...  ., +, TPAMI June 2020 1289-1302 Baselines Extraction from Curved Document Images via Slope Fields Recovery.  ...  Matsukawa, T., +, 2179 -2194 Learning with Privileged Information via Adversarial Discriminative Modality Distillation.  ... 
doi:10.1109/tpami.2020.3036557 fatcat:3j6s2l53x5eqxnlsptsgbjeebe

Linked Document Embedding for Classification

Suhang Wang, Jiliang Tang, Charu Aggarwal, Huan Liu
2016 Proceedings of the 25th ACM International on Conference on Information and Knowledge Management - CIKM '16  
In this paper, we study the problem of linked document embedding for classification and propose a linked document embedding framework LDE, which combines link and label information with content information  ...  to learn document representations for classification.  ...  and label information into a probabilistic framework for classification. • CNN [11] : convolution neural network for classification.  ... 
doi:10.1145/2983323.2983755 dblp:conf/cikm/WangTAL16 fatcat:bfgay4jan5ealg5czt5hpkxdxi
« Previous Showing results 1 — 15 out of 1,349 results