• 한국컴퓨터정보학회논문지
• /
• 제22권10호
• /
• pp.121-128
• /
• 2017

In this paper, we propose a multi-label classification method in which multi-label classification estimation techniques are applied to resolving location prediction problem. Most of previous studies related to location prediction have focused on the use of single-label classification by using contextual information such as user's movement paths, demographic information, etc. However, in this paper, we focused on the case where users are free to visit multiple locations, forcing decision-makers to use multi-labeled dataset. By using 2373 contextual dataset which was compiled from college students, we have obtained the best results with classifiers such as bagging, random subspace, and decision tree with the multi-label classification estimation methods like binary relevance(BR), binary pairwise classification (PW).

• 한국컴퓨터정보학회논문지
• /
• 제22권11호
• /
• pp.97-104
• /
• 2017

Location prediction has been successfully utilized to provide high quality of location-based services to customers in many applications. In its usual form, the conventional type of location prediction is to predict future locations based on user's past movement history. However, as location prediction needs are expanded into much complicated cases, it becomes necessary quite frequently to make inference on the locations that target user visited in the past. Typical cases include the identification of locations that infectious disease carriers may have visited before, and crime suspects may have dropped by on a certain day at a specific time-band. Therefore, primary goal of this study is to predict locations that users visited in the past. Information used for this purpose include user's demographic information and movement histories. Data mining classifiers such as Bayesian network, neural network, support vector machine, decision tree were adopted to analyze 6868 contextual dataset and compare classifiers' performance. Results show that general Bayesian network is the most robust classifier.

• International Journal of Contents
• /
• 제18권3호
• /
• pp.11-20
• /
• 2022

Human Emotion Recognition is an exciting topic that has been attracting many researchers for a lengthy time. In recent years, there has been an increasing interest in exploiting contextual information on emotion recognition. Some previous explorations in psychology show that emotional perception is impacted by facial expressions, as well as contextual information from the scene, such as human activities, interactions, and body poses. Those explorations initialize a trend in computer vision in exploring the critical role of contexts, by considering them as modalities to infer predicted emotion along with facial expressions. However, the contextual information has not been fully exploited. The scene emotion created by the surrounding environment, can shape how people perceive emotion. Besides, additive fusion in multimodal training fashion is not practical, because the contributions of each modality are not equal to the final prediction. The purpose of this paper was to contribute to this growing area of research, by exploring the effectiveness of the emotional scene gist in the input image, to infer the emotional state of the primary target. The emotional scene gist includes emotion, emotional feelings, and actions or events that directly trigger emotional reactions in the input image. We also present an attention-based fusion network, to combine multimodal features based on their impacts on the target emotional state. We demonstrate the effectiveness of the method, through a significant improvement on the EMOTIC dataset.

• International Journal of Computer Science & Network Security
• /
• 제23권9호
• /
• pp.47-54
• /
• 2023

For Bengali music emotion classification, deep learning models, particularly CNN and RNN are frequently used. But previous researches had the flaws of low accuracy and overfitting problem. In this research, attention-based Conv1D and BiGRU model is designed for music emotion classification and comparative experimentation shows that the proposed model is classifying emotions more accurate. We have proposed a Conv1D and Bi-GRU with the attention-based model for emotion classification of our Bengali music dataset. The model integrates attention-based. Wav preprocessing makes use of MFCCs. To reduce the dimensionality of the feature space, contextual features were extracted from two Conv1D layers. In order to solve the overfitting problems, dropouts are utilized. Two bidirectional GRUs networks are used to update previous and future emotion representation of the output from the Conv1D layers. Two BiGRU layers are conntected to an attention mechanism to give various MFCC feature vectors more attention. Moreover, the attention mechanism has increased the accuracy of the proposed classification model. The vector is finally classified into four emotion classes: Angry, Happy, Relax, Sad; using a dense, fully connected layer with softmax activation. The proposed Conv1D+BiGRU+Attention model is efficient at classifying emotions in the Bengali music dataset than baseline methods. For our Bengali music dataset, the performance of our proposed model is 95%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권7호
• /
• pp.2468-2487
• /
• 2015

In web searching, trustable and precise results are greatly affected by the inherent uncertainty in the input queries. Queries submitted to search engines are by nature ambiguous and constitute a significant proportion of the instances given to web search engines. Ambiguous queries pose real challenges for the web search engines due to versatility of information. Temporal based approaches whereas somehow reduce the uncertainty in queries but still lack to provide results according to users aspirations. Web search science has created an interest for the researchers to incorporate contextual information for resolving the uncertainty in search results. In this paper, we propose an Adaptive Disambiguation Approach (ADA) of hybrid nature that makes use of both the temporal and contextual information to improve user experience. The proposed hybrid approach presents the search results to the users based on their location and temporal information. A Java based prototype of the systems is developed and evaluated using standard dataset to determine its efficacy in terms of precision, accuracy, recall, and F1-measure. Supported by experimental results, ADA demonstrates better results along all the axes as compared to temporal based approaches.

• Journal of Information Processing Systems
• /
• 제17권4호
• /
• pp.818-833
• /
• 2021

To address the problems of low precision rate, insufficient feature extraction, and poor contextual ability in existing text sentiment analysis methods, a mixed model account of a CNN-BiLSTM-TE (convolutional neural network, bidirectional long short-term memory, and topic extraction) model was proposed. First, Chinese text data was converted into vectors through the method of transfer learning by Word2Vec. Second, local features were extracted by the CNN model. Then, contextual information was extracted by the BiLSTM neural network and the emotional tendency was obtained using softmax. Finally, topics were extracted by the term frequency-inverse document frequency and K-means. Compared with the CNN, BiLSTM, and gate recurrent unit (GRU) models, the CNN-BiLSTM-TE model's F1-score was higher than other models by 0.0147, 0.006, and 0.0052, respectively. Then compared with CNN-LSTM, LSTM-CNN, and BiLSTM-CNN models, the F1-score was higher by 0.0071, 0.0038, and 0.0049, respectively. Experimental results showed that the CNN-BiLSTM-TE model can effectively improve various indicators in application. Lastly, performed scalability verification through a takeaway dataset, which has great value in practical applications.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.1795-1811
• /
• 2019

In order to achieve rapid and accurate detection of vehicle objects in complex traffic conditions, we propose a novel vehicle detection method. Firstly, more contextual and small-object vehicle information can be obtained by our Joint Feature Network (JFN). Secondly, our Evolved Region Proposal Network (EPRN) generates initial anchor boxes by adding an improved version of the region proposal network in this network, and at the same time filters out a large number of false vehicle boxes by soft-Non Maximum Suppression (NMS). Then, our Mask Network (MaskN) generates an example that includes the vehicle occlusion, the generator and discriminator can learn from each other in order to further improve the vehicle object detection capability. Finally, these candidate vehicle detection boxes are optimized to obtain the final vehicle detection boxes by the Fine-Tuning Network(FTN). Through the evaluation experiment on the DETRAC benchmark dataset, we find that in terms of mAP, our method exceeds Faster-RCNN by 11.15%, YOLO by 11.88%, and EB by 1.64%. Besides, our algorithm also has achieved top2 comaring with MS-CNN, YOLO-v3, RefineNet, RetinaNet, Faster-rcnn, DSSD and YOLO-v2 of vehicle category in KITTI dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.3868-3888
• /
• 2022

A widely used social networking service like Twitter has the ability to disseminate information to large groups of people even during a pandemic. At the same time, it is a convenient medium to share irrelevant and unverified information online and poses a potential threat to society. In this research, conventional machine learning algorithms are analyzed to classify the data as either non-rumor data or rumor data. Machine learning techniques have limited tuning capability and make decisions based on their learning. To tackle this problem the authors propose a deep learning-based Rumor Detection Neural Network model to predict the rumor tweet in real-world events. This model comprises three layers, AttCNN layer is used to extract local and position invariant features from the data, AttBi-LSTM layer to extract important semantic or contextual information and HPOOL to combine the down sampling patches of the input feature maps from the average and maximum pooling layers. A dataset from Kaggle and ground dataset #gaja are used to train the proposed Rumor Detection Neural Network to determine the veracity of the rumor. The experimental results of the RDNN Classifier demonstrate an accuracy of 93.24% and 95.41% in identifying rumor tweets in real-time events.

• 지능정보연구
• /
• 제22권1호
• /
• pp.83-105
• /
• 2016

의료IT 서비스의 유망 분야인 정신건강 증진을 위한 주관적 웰빙 서비스(subjective well-being service) 구현의 핵심은 개인의 주관적 웰빙 상태를 정확하고 무구속적이며 비용 효율적으로 측정하는 것인데 이를 위해 보편적으로 사용되는 설문지에 의한 자기보고나 신체부착형 센서 기반의 측정 방법론은 정확성은 뛰어나나 비용효율성과 무구속성에 취약하다. 비용효율성과 무구속성을 보강하기 위한 온라인 텍스트 기반의 측정 방법은 사전에 준비된 감정어 어휘만을 사용함으로써 상황에 따라 감정어로 볼 수 있는 이른바 상황적 긍부정성(contextual polarity)을 고려하지 못하여 측정 정확도가 낮다. 한편 기존의 상황적 긍부정성을 활용한 감성분석으로는 주관적 웰빙 상태인 맥락에서의 감성분석을 할 수 있는 감정어휘사전이나 온톨로지가 구축되어 있지 않다. 더구나 온톨로지 구축도 매우 노력이 소요되는 작업이다. 따라서 본 연구의 목적은 온라인상에 사용자의 의견이 표출된 비정형 텍스트로부터 주관적 웰빙과 관련한 상황감정어를 추출하고, 이를 근거로 상황적 긍부정성 파악의 정확도를 개선하는 방법을 제안하는 것이다. 기본 절차는 다음과 같다. 먼저 일반 감정어휘사전을 준비한다. 본 연구에서는 가장 대표적인 디지털 감정어휘사전인 SentiWordNet을 사용하였다. 둘째, 정신건강지수를 동적으로 추정하는데 필요한 비정형 자료인 Corpora를 온라인 서베이로 확보하였다. 셋째, Corpora로부터 세 가지 종류의 자원을 확보하였다. 넷째, 자원을 입력변수로 하고 특정 정신건강 상태의 지수값을 종속변수로 하는 추론 모형을 구축하고 추론 규칙을 추출하였다. 마지막으로, 추론 규칙으로 정신건강 상태를 추론하였다. 본 연구는 감정을 분석함에 있어, 기존의 연구들과 달리 상황적 감정어를 적용하여 특정 도메인에 따라 다양한 감정 어휘를 파악할 수 있다는 점에서 독창성이 있다.

• 한국벤처창업학회:학술대회논문집
• /
• 한국벤처창업학회 2017년도 추계학술대회
• /
• pp.149-153
• /
• 2017

Intentions influence behaviors and, consequently, individual and organizational outcomes. The ability to understand intentions becomes a central issue. The objective of this study was to present and test an Entrepreneurial Intentions (EI) model. Drawing on a generally utilized paradigm, the theory of planned behavior and Shapero's model of the Entrepreneurial Event (SEE), we show the impact of individual and contextual factors on the intention development. Relying on the Global Entrepreneurship Monitor data(GEM), we test a EI conceptual model. The EI conceptual model is tested using the dataset of GEM over 30 countries and 3 subgroups. All the variables of interest indicate positive and significant effect on EI. Our results indicate that EI is influenced by Perceived Opportunity(PO), Perceived Capability(PC) and Government Support & Policy(GSP).