• ETRI Journal
• /
• v.45 no.1
• /
• pp.93-104
• /
• 2023

Coreference resolution is a task in discourse analysis that links several headwords used in any document object. We suggest pointer networks-based coreference resolution for Korean using multi-task learning (MTL) with an attention mechanism for a hierarchical structure. As Korean is a head-final language, the head can easily be found. Our model learns the distribution by referring to the same entity position and utilizes a pointer network to conduct coreference resolution depending on the input headword. As the input is a document, the input sequence is very long. Thus, the core idea is to learn the word- and sentence-level distributions in parallel with MTL, while using a shared representation to address the long sequence problem. The suggested technique is used to generate word representations for Korean based on contextual information using pre-trained language models for Korean. In the same experimental conditions, our model performed roughly 1.8% better on CoNLL F1 than previous research without hierarchical structure.

• Proceedings of the Korea Society of Design Studies Conference
• /
• 2000.11a
• /
• pp.46-47
• /
• 2000

전시공간에는 공간성과 시간성이 교차한다. 구조화된 공간은 시간적 인식 매커니즘으로 개별 시퀀스의 맥락적 합으로 인식된다. 그것은 마치 한편의 영화를 감상할 때나 전통 중국음식을 음미할 때에 잔상, 잔미의 연속적 롤 플레잉의 과정과 유사하다. (중략)

• KIISE Transactions on Computing Practices
• /
• v.24 no.3
• /
• pp.138-144
• /
• 2018

A video is a vivid medium similar to human's visual-linguistic experiences, since it can inculcate a sequence of situations, actions or dialogues that can be told as a story. In this study, we propose story learning/regeneration frameworks from videos with successive event order supervision for contextual coherence. The supervision induces each episode to have a form of trajectory in the latent space, which constructs a composite representation of ordering and semantics. In this study, we incorporated the use of kids videos as a training data. Some of the advantages associated with the kids videos include omnibus style, simple/explicit storyline in short, chronological narrative order, and relatively limited number of characters and spatial environments. We build the encoder-decoder structure with successive event order embedding, and train bi-directional LSTMs as sequence models considering multi-step sequence prediction. Using a series of approximately 200 episodes of kids videos named 'Pororo the Little Penguin', we give empirical results for story regeneration tasks and SEOE. In addition, each episode shows a trajectory-like shape on the latent space of the model, which gives the geometric information for the sequence models.

• ETRI Journal
• /
• v.43 no.2
• /
• pp.299-312
• /
• 2021

It is necessary to achieve high performance in the task of zero anaphora resolution (ZAR) for completely understanding the texts in Korean, Japanese, Chinese, and various other languages. Deep-learning-based models are being employed for building ZAR systems, owing to the success of deep learning in the recent years. However, the objective of building a high-quality ZAR system is far from being achieved even using these models. To enhance the current ZAR techniques, we fine-tuned a pretrained bidirectional encoder representations from transformers (BERT). Notably, BERT is a general language representation model that enables systems to utilize deep bidirectional contextual information in a natural language text. It extensively exploits the attention mechanism based upon the sequence-transduction model Transformer. In our model, classification is simultaneously performed for all the words in the input word sequence to decide whether each word can be an antecedent. We seek end-to-end learning by disallowing any use of hand-crafted or dependency-parsing features. Experimental results show that compared with other models, our approach can significantly improve the performance of ZAR.

• International Commerce and Information Review
• /
• v.10 no.4
• /
• pp.81-96
• /
• 2008

uEFL is an open source solution to integrate general business/learning activities and processes. uEFL is originally developed to adopt LD (Learning Design) specification, which represents learning as various combination of learning activities with learning conditions and outcomes. Learning activities are described with participant's role, learning environment, and contextual sequence. This viewpoint resembles BPM (Business Process Modeling). uEFL can convert LD to BPM description. uEFL engine can run converted LD activity with other business activities. This paper presents 4 templates and 2 sample models for uEFL. The templates and models will show how learning activities can be integrated with business activities efficiently.

• Journal of Korean Society of Transportation
• /
• v.25 no.6
• /
• pp.171-183
• /
• 2007

Urban transport demand is derived from activity participation. A variety of individual daily activities based on the decisions on activity participation result in collective spatial behavior. The travel derived from the effort to overcome the spatially distributed locations of adjacent activities represents the detailed structural relationships among activities. An activity-based approach provides an important framework of analyzing contemporary urban daily life in the sense that it studies the interaction between individuals' daily decision making and social practice in time and space, on the one hand, and socio-spatial environment on the other. The current study identifies representative patterns of urban daily activity implementations and analyzes the correlation between representative patterns and individuals' characteristics and contextual characteristics. The study shows that urban daily activity patterns can be grouped in a limited number of representative patterns, which are systematically correlated with socio-spatial characteristics. The results provide related transportation policy implications.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.40 no.8
• /
• pp.782-791
• /
• 1991

In the fields of pattern recognition such as speech recognition, several new techniques using Artifical Neural network Models have been proposed and implemented. In particular, the Multilayer Perception Model has been shown to be effective in static speech pattern recognition. But speech has dynamic or temporal characteristics and the most important point in implementing speech recognition systems using Artificial Neural Network Models for continuous speech is the learning of dynamic characteristics and the distributed cues and contextual effects that result from temporal characteristics. But Recurrent Multilayer Perceptron Model is known to be able to learn sequence of pattern. In this paper, the results of applying the Recurrent Model which has possibilities of learning tedmporal characteristics of speech to phoneme recognition is presented. The test data consist of 144 Vowel+ Consonant + Vowel speech chains made up of 4 Korean monothongs and 9 Korean plosive consonants. The input parameters of Artificial Neural Network model used are the FFT coefficients, residual error and zero crossing rates. The Baseline model showed a recognition rate of 91% for volwels and 71% for plosive consonants of one male speaker. We obtained better recognition rates from various other experiments compared to the existing multilayer perceptron model, thus showed the recurrent model to be better suited to speech recognition. And the possibility of using Recurrent Models for speech recognition was experimented by changing the configuration of this baseline model.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.579-582
• /
• 2009

With the aid of the increasing interests of Context-aware Augmented Reality Agent (AR Agent), various researches of AR Agent have been performed to explore the possibility of the agent as novel interface and the entity responding autonomously by user's input. However, in previous works, AR Agents are lack of specific method for using various contextual information. To revolve around those problems, we propose the Behavior Generation System for Context-aware AR Agent using layered architecture. Based on Belief-Desire-Intention (BDI) model and Hierarchical Task Network (HTN) searching, the sequence of agent behavior has been selected in behavior planning layer. Then, the agent evaluates appropriateness of behaviors using previous behavior and the type of input before activation. This behavior generation system can be applied for edutainment, game, and assistant agent, which need intuitive and effective behaviors to convey information. Through this research, we expect that the Context-aware AR Agent could support for not only information delivery, but also the capability of effective communication for user.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.2
• /
• pp.308-313
• /
• 2021

In this study, tweets about the 2019 Koreas-US DMZ Summit were collected along with a time sequence and analyzed by a sequential topic modeling method, Dynamic Topic Modeling(DTM). In microblogging services such as Twitter, unstructured data that mixes news and an opinion about a single event occurs at the same time on a large scale, and information and reactions are produced in the same message format. Therefore, to grasp a topic trend, the contextual meaning can be found only by performing pattern analysis reflecting the characteristics of sequential data. As a result of calculating the DTM after obtaining the topic coherence score and evaluating the Latent Dirichlet Allocation(LDA), 30 topics related to news reports and opinions were derived, and the probability of occurrence of each topic and keywords were dynamically evolving. In conclusion, the study found that DTM is a suitable model for analyzing the trend of integrated topics in a specific event over time.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.253-266
• /
• 2018

This paper proposes a methodology applying sequence tagging methodology to improve the performance of NER(Named Entity Recognition) used in QA system. In order to retrieve the correct answers stored in the database, it is necessary to switch the user's query into a language of the database such as SQL(Structured Query Language). Then, the computer can recognize the language of the user. This is the process of identifying the class or data name contained in the database. The method of retrieving the words contained in the query in the existing database and recognizing the object does not identify the homophone and the word phrases because it does not consider the context of the user's query. If there are multiple search results, all of them are returned as a result, so there can be many interpretations on the query and the time complexity for the calculation becomes large. To overcome these, this study aims to solve this problem by reflecting the contextual meaning of the query using Bidirectional LSTM-CRF. Also we tried to solve the disadvantages of the neural network model which can't identify the untrained words by using ontology knowledge based feature. Experiments were conducted on the ontology knowledge base of music domain and the performance was evaluated. In order to accurately evaluate the performance of the L-Bidirectional LSTM-CRF proposed in this study, we experimented with converting the words included in the learned query into untrained words in order to test whether the words were included in the database but correctly identified the untrained words. As a result, it was possible to recognize objects considering the context and can recognize the untrained words without re-training the L-Bidirectional LSTM-CRF mode, and it is confirmed that the performance of the object recognition as a whole is improved.