• Journal of Korea Multimedia Society
• /
• v.18 no.1
• /
• pp.25-34
• /
• 2015

Recently, in distribution market, demand for electronic shelf label system is increasing gradually to provide the accurate price immediately and detailed product information to consumers and reduce operation costs. Most of electronic shelf label system companies develop the full-graphic display device to display a wide variety of product information as well as the exact price. Our system had introduced Go-Back-N retransmission method in the early. However, we encountered performance problems that it delayed updating of the electronic shelf label system and exhausted the battery life time. Proposed adaptive image retransmission technique based on the selective scheme is that tags of electronic shelf label system recognize idle time of transmission cycle and require partial image retransmission to sever by itself. As a result, it can acquire much more opportunities of partial image retransmission within the same period and increase reception rate of full image for each tags. The experimental result shows that adaptive image retransmission technique's reception rate of full image for each tags is approximately 4% higher than existing previous works. And total battery life time increases 30 hours because tag reduce wake-up time as it receive only lost data instead of whole data.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.1
• /
• pp.41-50
• /
• 2023

Recently, as the Coronavirus disease 2019 (COVID-19) prolongs along with the development of artificial intelligence technology, a non-contact society has become commonplace. Many companies are promoting digital transformation and the activation of artificial intelligence introduction to respond to this which leads to dramatic increase of demand for Chatbot. In addition, a Chatbot has reached the point of processing business transactions from responding simple inquiries. However, it is necessary to develop an API to interface with the legacy system and there are many difficulties in connecting. To solve this, it is becoming important to establish a hybrid Chatbot environment through RPA interface. Recently, the combination of RPA and Chatbot is considered an effective tool for handling many business processes. But, there are many difficulties due to the lack of interface cases and the difficulty in finding a method to development them. This study suggests a method for building a hybrid Chatbot which is an interface Chatbot(Conversational UX) and RPA(Task Automation) from the perspective of hyper-automation based on actual development cases and review of literature review is presented, so that the interface method can be understood and develop more easily. Therefore, there are implications for actively using AI Chatbot for digital transformation.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.41-49
• /
• 2022

Documents often contain relative time expressions, and it is important to define a schema of the relative time information and develop a system that extracts such information from corpus. In this study, to deal with the relative time expressions, we propose seven additional attributes of timex3: year, month, day, week, hour, minute, and second. We propose a way to represent normalized values of the relative time expressions such as before, after, and count, and also design a set of rules to extract the relative time information from texts. With a new corpus constructed using the new attributes that consists of dialog, news, and history documents, we observed that our rule-set generally achieved 70% accuracy on the 1,041 documents. Especially, with the most frequently appeared attributes such as year, day, and week, we got higher accuracies compared to other attributes. The results of this study, our proposed timex3 attributes and the rule-set, will be useful in the development of services such as question-answer systems and chatbots.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.1
• /
• pp.1-8
• /
• 2009

Learning user's preference is a key issue in intelligent system such as personalized service. The study on user preference model has adapted simple user preference model, which determines a set of preferred keywords or topic, and weights to each target. In this paper, we recommend multi-perspective user preference model that factors sentiment information in the model. Based on the topicality and sentimental information processed using natural language processing techniques, it learns a user's preference. To handle timc-variant nature of user preference, user preference is calculated by session, short-term and long term. User evaluation is used to validate the effect of user preference teaming and it shows 86.52%, 86.28%, 87.22% of accuracy for topic interest, keyword interest, and keyword favorableness.

• Journal of the Korea Computer Graphics Society
• /
• v.29 no.3
• /
• pp.21-31
• /
• 2023

In this paper, we introduce a case study of cinematography using digital human in virtual production. This case study deals with the system overview of virtual production using LEDs and an efficient filming pipeline using digital human. Unlike virtual production using LEDs, which mainly project the background on LEDs, in this case, we use digital human as a virtual actor to film scenes communicating with a real actor. In addition, to film the dialogue scene between the real actor and the digital human using a real-time engine, we automatically generated speech animation of the digital human in advance by applying our Korean lip-sync technology based on audio and text. We verified this filming case by using a real-time engine to produce short drama content using real actor and digital human in an LED-based virtual production environment.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.34 no.3
• /
• pp.451-461
• /
• 2024

Recently, various AI chatbots based on large language models have been released. Chatbots have the advantage of providing users with quick and easy information through interactive prompts, making them useful in various fields such as question answering, writing, and programming. However, a vulnerability in chatbots called "prompt injection attacks" has been proposed. This attack involves injecting instructions into the chatbot to violate predefined guidelines. Such attacks can be critical as they may lead to the leakage of confidential information within large language models or trigger other malicious activities. However, the vulnerability of Korean prompts has not been adequately validated. Therefore, in this paper, we aim to generate malicious Korean prompts and perform attacks on the popular chatbot to analyze their feasibility. To achieve this, we propose a system that automatically generates malicious Korean prompts by analyzing existing prompt injection attacks. Specifically, we focus on generating malicious prompts that induce harmful expressions from large language models and validate their effectiveness in practice.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.3
• /
• pp.87-94
• /
• 2024

We investigated the feasibility of utilizing Large Language Models (LLMs) to perform text-based games without training on game data in advance. We adopted ChatGPT-3.5 and its state-of-the-art, ChatGPT-4, as the systems that implemented LLM. In addition, we added the persistent memory feature proposed in this paper to ChatGPT-4 to create three game player agents. We used Zork, one of the most famous text-based games, to see if the agents could navigate through complex locations, gather information, and solve puzzles. The results showed that the agent with persistent memory had the widest range of exploration and the best score among the three agents. However, all three agents were limited in solving puzzles, indicating that LLM is vulnerable to problems that require multi-level reasoning. Nevertheless, the proposed agent was still able to visit 37.3% of the total locations and collect all the items in the locations it visited, demonstrating the potential of LLM.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.1
• /
• pp.127-143
• /
• 1996

Advanced countries are very active in deploying the National Information I infrastructure which provides universal service to promise fruitful quality of futuree life. Even in the distributed environment, we can closely converse, work together, and share information in a very convenient way. This is actually enabled with the help of groupware technology, which are currently focused and researched in a larger popularity. The aim of this study is to design a portable pack for group communications management to support the development of groupware application. In the paper we begin with technical survey, continue to build our own model for group communications man agement, and design its architecture and procedure. We also suggest group addressing mechanism under Internet environment such as how to create IP multicast address and IP port number dynamically and as a globally unique value for the communication session, with the help of the hierarchical and distributed address managers. We also indicate the reliable data transmission services to remedy the unreliable feature of the UDP multicast services, and finally the architecture/ applied to support the practical applications is briefly discussed for verification of the designed concept.

• Journal of KIISE:Software and Applications
• /
• v.30 no.3_4
• /
• pp.278-284
• /
• 2003

Speech act is the speaker's intentions indicated through utterances. It is important for understanding natural language dialogues and generating responses. This paper proposes the method of two stage that increases the performance of the korean speech act decision. The first stage is to select features from the part of speech results in sentence and from the context that uses previous speech acts. We use x$^2$ statistics(CHI) for selecting features that have showed high performance in text categorization. The second stage is to determine speech act with selected features and Neural Network. The proposed method shows the possibility of automatic speech act decision using only POS results, makes good performance by using the higher informative features and speed up by decreasing the number of features. We tested the system using our proposed method in Korean dialogue corpus transcribed from recording in real fields, and this corpus consists of 10,285 utterances and 17 speech acts. We trained it with 8,349 utterances and have test it with 1,936 utterances, obtained the correct speech act for 1,709 utterances(88.3%). This result is about 8% higher accuracy than without selecting features.