• Journal of Aquaculture
• /
• v.17 no.2
• /
• pp.139-143
• /
• 2004

In the fish oil forced oxidized at 6$0^{\circ}C$ for 10 days, changes in the levels of peroxide (POV), anisidine (AnV), total oxidation (Totox), iodine (IV), acid (AV) and fatty acids composition were measured. The levels of POV, AnV and Totox remained unchanged or decreased after reaching the maximum. The concentrations of polyunsaturated fatty acids (PUFA) such as Docosa hexaenoic acid (DHA) or Eicosa pentaenoic acid (EPA) decreased with extended oxidation of fish oil. In saturated fatty acids (SFA) like C16:0, their concentration increased with decreasing PUFA. The ratios of PUFA/SFA and DHA/C16:0 decreased with extended oxidation of fish oil. Using a single parameter of POV, AnV, Totox, AV, IV, or fatty acids for evaluation of the quality of fish oil may prove difficult. Besides other parameters, the ratios of PUFA/SFA and/or DHA/C16:0 could be a good index to evaluate the quality of fish oil.

• Journal of KIISE:Software and Applications
• /
• v.36 no.1
• /
• pp.62-69
• /
• 2009

The purpose of document summarization is to provide easy and quick understanding of documents by extracting summarized information from the documents produced by various application programs. In this paper, we propose a document summarization method that creates and analyzes a connection graph representing the similarity of keyword lists of sentences in a document taking into account the mean length(the number of keywords) of sentences of the document. We implemented a system that automatically generate a summary from a document using the proposed method. To evaluate the performance of the method, we used a set of 20 documents associated with their correct summaries and measured the precision, the recall and the F-measure. The experiment results show that the proposed method is more efficient compared with the existing methods.

• The Journal of the Korea Contents Association
• /
• v.11 no.3
• /
• pp.467-476
• /
• 2011

Compound noun phrase (CNP) is important factor for semantic information process because the meaning of the CNP is more disambiguous than that of single word. However, the CNP can be expressed in various types even though it expresses same meaning. It is called syntactic diversity. It makes information system difficult to grasp sense identity. In order to resolve the syntactic diversity in this research, we propose an indexing method for compound noun phrase. The main purpose is to make identical index term for various types of CNPs which has same meaning. To do so, the research follows next steps. For the first, we make rule template and utilize the template to extract CNPs from set of domestic research papers. In general, the CNP has a unique meaning. Considering the characteristic, we suggest synthesis rules of index terms and apply the rule to CNPs extracted in previous step. For the objective performance evaluation of the research, a test set, HANTEC 2.0, was utilized and the result was compared to baseline model. Through the experiment and the evaluation, we have confirmed that the indexing method suggested in this paper could positively affect retrieval precision and improve performance of the information retrieval.

• Journal of Digital Convergence
• /
• v.18 no.2
• /
• pp.367-376
• /
• 2020

According to recent statistics on new consumer market trends, 'alone consumption' is at the center. This study focuses on the social big data that attracts the public's opinions in that it is important for a certain social trend to comprehensively understand the various fields such as society, locality, culture, marketing, economics, and psychology that form the background for it. Therefore, we set up the linkage of 'solo consumption' and conducted research on new consumer market trends using Opinion Analisys. As a result of this trend analysis, representative keywords such as 'honbab', 'honsul' and 'honyoeng' were derived and analyzed the trend of new consumer market using this data. Alone consumption is an inevitable new consumption trend caused by demographic change after the global economic crisis. The importance as a trend reflecting this will be further strengthened. Trend analysis by social big data will help scientific and systematic business distribution strategies and planning to help make new and valuable decisions and decisions about new consumer markets.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.9
• /
• pp.451-464
• /
• 2016

Nowadays, entity search such as Google Product Search and Yahoo Pipes has been in the spotlight. The entity search engines have been used to retrieve web pages relevant with a particular entity. However, if an entity (e.g., Chinatown movie) has various meanings (e.g., Chinatown movies, Chinatown restaurants, and Incheon Chinatown), then the accuracy of the search result will be decreased significantly. To address this problem, in this article, we propose a novel method that quantifies the importance of search queries and then offers the best query for the entity search, based on Frequent Pattern (FP)-Tree, considering the correlation between the entity relevance and the frequency of web pages. According to the experimental results presented in this paper, the proposed method (59% in the average precision) improved the accuracy five times, compared to the traditional query terms (less than 10% in the average precision).

• The Journal of the Korea Contents Association
• /
• v.19 no.7
• /
• pp.427-435
• /
• 2019

Soundscape provides important resources to understand social and cultural aspects of our society, however, it is still its infancy to study on the research framework to record, conserve, categorize, and analyze soundscapes. Topic modeling is an automatic approach to discover hidden themes that are disperse in unstructured documents, thus topic modeling is robust enough to find latent topics such as research trends behind a collection of documents. The purpose of this paper is to discover topics on current soundscape research based on topic modeling, furthermore, to discuss the possibilities to design a metadata system for sound archives and to improve Soundscape Ontology which is currently developing.

• Economic and Environmental Geology
• /
• v.55 no.3
• /
• pp.273-280
• /
• 2022

The purpose of this study is to explore the social perception on geoscience using a big data analysis and to propose a way to enhance people's perception on geoscience. For this, 5,044 media articles including geoscience produced by 54 media companies from January 1, 2010 to April 14, 2022. were analyzed. Big data analyses were applied. The results of analyses are as follows: Media articles consist of key words of research institute, some countries of America, China and Japan, City of Pohang, CEO of KIGAM. And geology, industry, development of mineral resources, environment, energy, nuclear power, and groundwater are highlighted as key words. Also, it is confirmed that topics related to geoscience such as expert, environment and research institute are not individually isolated, but interconnected and linked to topics in the center of future, industry, and global. Based on this result, ways to enhance people's perception on geoscience were discussed.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.31-38
• /
• 2023

Style transfer is one of deep learning-based image processing techniques that has been actively researched recently. These research efforts have led to significant improvements in the quality of result images. Style transfer is a technology that takes a content image and a style image as inputs and generates a transformed result image by applying the characteristics of the style image to the content image. It is becoming increasingly important in exploiting the diversity of digital content. To improve the usability of style transfer technology, ensuring stable performance is crucial. Recently, in the field of natural language processing, the concept of Transformers has been actively utilized. Attention maps, which forms the basis of Transformers, is also being actively applied and researched in the development of style transfer techniques. In this paper, we analyze the representative techniques SANet and AdaAttN and propose a novel attention map-based structure which can generate improved style transfer results. The results demonstrate that the proposed technique effectively preserves the structure of the content image while applying the characteristics of the style image.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.8
• /
• pp.403-408
• /
• 2007

In a large corpus-based speech synthesizer, a break, which is a parameter influencing the naturalness and intelligibility, is used as an important feature during a unit selection process. Japanese is a language having intonations, which ate indicated by the relative differences in pitch heights and the APs(Accentual Phrases) are placed according to the changes of the accents while a break occurs on a boundary of the APs. Although a break can be predicted by using J-ToBI(Japanese-Tones and Break Indices), which is a rule-based or statistical approach, it is very difficult to predict a break exactly due to the flexibility. Therefore, in this paper, a method is to conduct a unit search by dividing breaks into two types, such as a fixed break and a flexible break, in order to use the advantages of a large-scale corpus, which includes various types of prosodies. As a result of an experiment, the proposed unit selection method contributed itself to enhance the naturalness of synthesized speeches.

• Journal of IKEEE
• /
• v.28 no.1
• /
• pp.53-64
• /
• 2024

With the advancement of NLP(Natural Language Processing) models, conversational AI such as ChatGPT is becoming increasingly popular. To enhance processing speed and reduce power consumption, it is important to implement the Transformer algorithm, which forms the basis of the latest natural language processing models, in hardware. In particular, the multi-head attention and feed-forward network, which analyze the relationships between different words in a sentence through matrix multiplication, are the most computationally intensive core algorithms in the Transformer. In this paper, we propose a new variable systolic array based on the number of input words to enhance matrix multiplication speed. Quantization maintains Transformer accuracy, boosting memory efficiency and speed. For evaluation purposes, this paper verifies the clock cycles required in multi-head attention and feed-forward network and compares the performance with other multipliers.