• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.6
• /
• pp.235-242
• /
• 2021

In this work, we develop a Korean dependency parser based on a stack-pointer network that consists of a pointer network and an internal stack. The parser has an encoder and decoder and builds a dependency tree for an input sentence in a depth-first manner. The encoder of the parser encodes an input sentence, and the decoder selects a child for the word at the top of the stack at each step. Since the parser has the internal stack where a search path is stored, the parser can utilize information of previously derived subtrees when selecting a child node. Previous studies used only a grandparent and the most recently visited sibling without considering a subtree structure. In this paper, we introduce graph attention networks that can represent a previously derived subtree. Then we modify our parser based on the stack-pointer network to utilize subtree information produced by the graph attention networks. After training the dependency parser using Sejong and Everyone's corpus, we evaluate the parser's performance. Experimental results show that the proposed parser achieves better performance than the previous approaches at sentence-level accuracies when adopting 2-depth graph attention networks.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.151-155
• /
• 2001

Mobile agent technology has been the subject of much attention in the last few years, mainly due to the proliferation of distributed software technologies combined with the distributed AI research field. In this paper, we present a design of communication networks of agents that cooperate with each other for forwarding messages to the specific mobile agent in order to make the overall system location transparent. In order to make the material accessible to general intelligent system researchers, we present the general ideas abstractly in terms of the graph theory. In particular, a proxy network is defined as a directed acyclic graph satisfying some structural conditions. In turns out that the definition ensures some kind of reliability of the network, in the sense that as long as at most one proxy agent is abnormal, there agent exists a communication path, from every proxy agent to the target agent, without passing through the abnormal proxy. As the basis for the implementation of this scheme, an appropriate initial proxy network is specified and the dynamic nature of the network is represented by a set of graph transformation rules. It is shown that those rules are sound, in the sense that all graphs created from the initial proxy network by zero or more applications of the rules are guaranteed to be proxy networks. Finally, we will discuss some implementation issues.

• Journal of Korea Multimedia Society
• /
• v.25 no.11
• /
• pp.1643-1652
• /
• 2022

Most of the existing machine reading research has used Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) algorithms as networks. Among them, RNN was slow in training, and Question Answering Network (QANet) was announced to improve training speed. QANet is a model composed of CNN and self-attention. CNN extracts semantic and syntactic information well from the local corpus, but there is a limit to extracting the corresponding information from the global corpus. Graph Convolutional Networks (GCN) extracts semantic and syntactic information relatively well from the global corpus. In this paper, to take advantage of this strength of GCN, we propose I-QANet, which changed the CNN of QANet to GCN. The proposed model performed 1.2 times faster than the baseline in the Stanford Question Answering Dataset (SQuAD) dataset and showed 0.2% higher performance in Exact Match (EM) and 0.7% higher in F1. Furthermore, in the Korean Question Answering Dataset (KorQuAD) dataset consisting only of Korean, the learning time was 1.1 times faster than the baseline, and the EM and F1 performance were also 0.9% and 0.7% higher, respectively.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.463-465
• /
• 2021

센서 데이터는 여러가지 원인으로 인해 데이터 결측치가 발생할 수 있으며, 결측치로 인한 데이터의 처리 방식에 따라 데이터 분석 결과가 다르게 해석될 수 있다. 이는 펫 헬스 케어 서비스에서 치명적인 문제로 연결될 수 있다. 따라서 본 논문에서는 펫 웨어러블 디바이스로부터 수집되는 다양한 센서 데이터의 결측치를 처리하기 위해 GATs(Graph Attention neTworks)와 LSTM(Long Short Term Memory)을 결합하여 활용한 데이터 결측치 처리 기법을 제안한다. 펫 웨어러블 디바이스의 센서 데이터가 서로 연관성을 가지고 있다는 점을 바탕으로 인접 노드의 Attention 수치와 Feature map을 도출한다. 이후 Prediction Layer 를 통해 결측치의 Feature 를 예측한다. 예측된 Feature 를 기반으로 Decoding 과정과 함께 결측치 보간이 이루어진다. 제안된 기법은 모델의 변형을 통해 이상치 탐지에도 활용할 수 있을 것으로 기대한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.399-401
• /
• 2022

POI (Point-of-Interest) 추천은 다양한 위치 기반 서비스에서 중요한 역할을 있다. 기존 연구에서는 사용자의 모바일 선호도를 모델링하기 위해 과거의 체크인의 공간-시간적 관계를 추출한다. 그러나 사용자 궤적에 숨겨진 개인 방문 경향을 반영할 수 있는 structured feature 는 잘 활용되지 않는다. 이 논문에서는 궤적 그래프를 결합한 시공간 인식 attention 네트워크를 제안한다. 개인의 선호도가 시간이 지남에 따라 변할 수 있다는 점을 고려하면 Dynamic GCN (Graph Convolution Network) 모듈은 POI 들의 공간적 상관관계를 동적으로 집계할 수 있다. LBSN (Location-Based Social Networks) 데이터 세트에서 검증된 새 모델은 기존 모델보다 약 9.0% 성능이 뛰어나다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.4008-4023
• /
• 2022

Since machine learning was introduced into cross-site scripting (XSS) attack detection, many researchers have conducted related studies and achieved significant results, such as saving time and labor costs by not maintaining a rule database, which is required by traditional XSS attack detection methods. However, this topic came across some problems, such as poor generalization ability, significant false negative rate (FNR) and false positive rate (FPR). Moreover, the automatic clustering property of graph convolutional networks (GCN) has attracted the attention of researchers. In the field of natural language process (NLP), the results of graph embedding based on GCN are automatically clustered in space without any training, which means that text data can be classified just by the embedding process based on GCN. Previously, other methods required training with the help of labeled data after embedding to complete data classification. With the help of the GCN auto-clustering feature and labeled data, this research proposes an approach to detect XSS attacks (called GCNXSS) to mine the dependencies between the units that constitute an XSS payload. First, GCNXSS transforms a URL into a word homogeneous graph based on word co-occurrence relationships. Then, GCNXSS inputs the graph into the GCN model for graph embedding and gets the classification results. Experimental results show that GCNXSS achieved successful results with accuracy, precision, recall, F1-score, FNR, FPR, and predicted time scores of 99.97%, 99.75%, 99.97%, 99.86%, 0.03%, 0.03%, and 0.0461ms. Compared with existing methods, GCNXSS has a lower FNR and FPR with stronger generalization ability.

• ETRI Journal
• /
• v.38 no.4
• /
• pp.714-723
• /
• 2016

Considerable attention has been given to processing graph data in recent years. An efficient method for computing the node proximity is one of the most challenging problems for many applications such as recommendation systems and social networks. Regarding large-scale, mutable datasets and user queries, top-k query processing has gained significant interest. This paper presents a novel method to find top-k answers in a node proximity search based on the well-known measure, Personalized PageRank (PPR). First, we introduce a distribution state transition graph (DSTG) to depict iterative steps for solving the PPR equation. Second, we propose a weight distribution model of a DSTG to capture the states of intermediate PPR scores and their distribution. Using a DSTG, we can selectively follow and compare multiple random paths with different lengths to find the most promising nodes. Moreover, we prove that the results of our method are equivalent to the PPR results. Comparative performance studies using two real datasets clearly show that our method is practical and accurate.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.14 no.4
• /
• pp.314-322
• /
• 2021

Skeleton-based action recognition has attracted considerable attention in human action recognition. Recent methods for skeleton-based action recognition employ spatiotemporal graph convolutional networks (GCNs) and have remarkable performance. However, most of them have heavy computational complexity for robust action recognition. To solve this problem, we propose a shuffle graph convolutional network (SGCN) which is a lightweight graph convolutional network using pointwise group convolution rather than pointwise convolution to reduce computational cost. Our SGCN is composed of spatial and temporal GCN. The spatial shuffle GCN contains pointwise group convolution and part shuffle module which enhances local and global information between correlated joints. In addition, the temporal shuffle GCN contains depthwise convolution to maintain a large receptive field. Our model achieves comparable performance with lowest computational cost and exceeds the performance of baseline at 0.3% and 1.2% on NTU RGB+D and NTU RGB+D 120 datasets, respectively.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.471-476
• /
• 2024

With the spread of various smart devices and computing devices, big data generation is occurring widely. Machine learning is an algorithm that performs reasoning by learning data patterns. Among the various machine learning algorithms, the algorithm that attracts attention is deep learning based on neural networks. Deep learning is achieving rapid performance improvement with the release of various applications. Recently, among deep learning algorithms, attempts to analyze data using graph structures are increasing. In this study, we present a graph generation method for transferring to a deep learning network. This paper proposes a method of generalizing node properties and edge weights in the graph generation process and converting them into a structure for deep learning input by presenting a matricization We present a method of applying a linear transformation matrix that can preserve attribute and weight information in the graph generation process. Finally, we present a deep learning input structure of a general graph and present an approach for performance analysis.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.20-23
• /
• 2020

개체 연결이란 문서에서 등장한 멘션(Mention)들을 지식 기반(Knowledge Base)상의 하나의 개체에 연결하는 문제를 말한다. 개체 연결은 개체를 찾는 멘션 탐지(mention detection)과정과 인식된 멘션에 대해 중의성을 해결하여 하나의 개체를 찾는 개체 중의성 해결(Entity disambiguation)과정으로 구성된다. 본 논문에서는 개체 정보를 강화하기 위해 wikipedia2vec정보를 결합하여 Entity 정보를 강화하고 문장 내에 모든 개체 정보를 활용하기 위해 집합적 개체를 정의하고 그래프 구조를 표현하기 위해 GNN을 활용하여 기존보다 높은 성능을 이끌어내었다.