• The Journal of the Acoustical Society of Korea
• /
• v.18 no.5
• /
• pp.44-51
• /
• 1999

This paper presents a massively parallel computational model for the efficient integration of speech and natural language understanding. The phoneme model is based on continuous Hidden Markov Model with context dependent phonemes, and the language model is based on a knowledge base approach. To construct the knowledge base, we adopt a hierarchically-structured semantic network and a memory-based parsing technique that employs parallel marker-passing as an inference mechanism. Our parallel speech recognition algorithm is implemented in a multi-Transputer system using distributed-memory MIMD multiprocessors. Experimental results show that the parallel speech recognition system performs better in recognition accuracy than a word network-based speech recognition system. The recognition accuracy is further improved by applying code-phoneme statistics. Besides, speedup experiments demonstrate the possibility of constructing a realtime parallel speech recognition system.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.5
• /
• pp.385-389
• /
• 2009

As a new community forming environment, the blog platform enables sharing of the resources in blogosphere through active information exchange. Many researches have been performed to recommend appropriate resources to users from vast amounts of blog resources. As one of the solutions OSEM defines the knowledge base in the blogosphere with ontology for effectively modeling it. In this paper, we propose a technique of converting the knowledge base into the OWL ontology for sharing it on the semantic web environment. An inference method is then applied to the OWL ontology for recommending interest blog communities. For this aim, a mapping method is offered and then SWRL inference and SPARQL query based on the ontology are employed to extract interest blog communities.

• Language and Information
• /
• v.12 no.1
• /
• pp.27-65
• /
• 2008

This paper presents a set of methodologies for aligning hypernym-hyponym noun pairs between Korean and English, based on the EuroWordNet approach. Following the methods conducted in EuroWordNet, our approach makes extensive use of WordNet in four steps of the building process: 1) Monolingual dictionaries have been used to extract proper hypernym-hyponym noun pairs, 2) bilingual dictionary has converted the extracted pairs, 3) Word Net has been used as a backbone of alignment criteria, and 4) WordNet has been used to select the most similar pair among the candidates. The importance of this study lies not only on enriching semantic links between two languages, but also on integrating lexical resources based on a language specific and dependent structure. Our approaches are aimed at building an accurate and detailed lexical resource with proper measures rather than at fast development of generic one using NLP technique.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.2
• /
• pp.243-248
• /
• 2011

According to increasing of accessing multimedia stream contents, Web services have become popular. However, these Web services are not supported with the same quality to Web clients who frequently access multimedia services. This paper proposes ontological technique to apply client level Quality of Service(QoS) that provides two different levels to serve Web service with proper quality by contribution value. And, it describes with UML(Unified Modeling Language) how to relate QoS and SLA(Service Level Agreement). Main contribution of this paper is to support client level QoS and SLA and to use Ontology for it. Therefore, this work uses an ontology-based approach to organize QoS and SLA, enabling semantic classification of all Web services based on domains and QoS and SLA attributes.

• The Research Journal of the Costume Culture
• /
• v.14 no.6
• /
• pp.1015-1026
• /
• 2006

The purpose of the study were to find out (1) the effect of motives on perceiver's image perception on ramie fabrics, and perceiver's trait, age and gender on sensory image evaluation of ramie fabrics. The research was a quasi experiment and experimental materials developed for the study were a set of material stimuli and semantic differential scales to measure sensory image of the stimuli, an aesthetic value scale. the independent design was motif design techniques(Plain Weave, burnt-out, embroidery, stripe, check). The subjects were 421 adults in Daejeon and Seachun. The results was as follows: The factor analysis of semantic differential scales for the ramie materials emerged 4 different image dimensions: attractiveness, hand, elegance, weight). The five design techniques showed significantly different image affects on some selective dimensions. The burn-out design gave the most attractive image, the embroidery design gave the softest image and plain weaved fabric presented the lightest hand image. Consumer's aesthetic values, gender and age tended to affect sensory image evaluation of ramie materials. On conclusion the result revealed that design strategy for the ramie material, design development though motives will be an essential process. and for material design pursued design image and target consumer's trait should be carefully considered.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.8
• /
• pp.2041-2049
• /
• 1998

This paper suggests a technique for program segmentation and maintenance using PSDG(Post-State Dependency Graph) that improves the quality of a software by identifying and detecting defects in already fixed source code. A program segmentation is performed by utilizing source code analysis which combines the measures of static, dynamic and semantic slicing when we need understandability of defect in programs for corrective maintanence. It provides users with a segmental principle to split a program by tracing state dependency of a source code with the graph, and clustering and highlighting, Through a modeling of the PSDG, elimination of ineffective program deadcode and generalization of related program segments arc possible, Additionally, it can be correlated with other design modeb as STD(State Transition Diagram), also be used as design documents.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.04a
• /
• pp.722-725
• /
• 2015

모바일 환경에서 사용자의 GPS 궤적은 위치기반서비스(Location Based Service)에서 새로운 자원으로써 활용되고 있다. 위치기반서비스의 확장을 위해 단순히 사용자의 위치를 지도에 표시하는 것뿐만 아니라 사용자들이 위치했던 장소들이 내포하고 있는 의미를 발견해 내는 것이 필요하다. 이를 위해 최근 사용자의 위치정보에 관심지점(POI: Point of Interest)의 정보를 결합하여 시맨틱 궤적(Semantic Trajectory)을 생성하고 분석하는 연구들이 진행되고 있다. 이러한 기존연구의 경우 시맨틱 궤적을 생성하기 위해, 사용자의 GPS 궤적과 POI의 면적 정보(polygon)가 겹칠 경우를 찾아내서 이를 시맨틱 궤적으로 생성하였다. 하지만 대부분 공개된 POI 정보는 실제 장소들의 면적 정보를 제공하지 않고 좌표(point) 값 만을 제공하기 때문에 기존의 방법으로는 시맨틱 궤적을 생성하지 못하는 문제가 있다. 본 논문에서는 사용자의 GPS 궤적과 POI의 좌표 값을 이용하여 사용자가 실제 방문했을 것으로 예상되는 POI 를 추정하고 이를 시맨틱 궤적으로 생성해 내는 방법을 제안한다. 제안하는 기법은 GPS 궤적의 속력 정보를 사용하여 사용자가 정지했었던 구간을 판별하고, 정지 구간 주변의 POI 밀도에 따라 정지 구간을 영역으로 확장한다. 그리고 영역에 포함된 POI 중 정지 구간과의 거리가 가장 가깝고, 가장 오랜 시간 포함되었던 POI를 사용자가 방문했던 POI로 판단한다. 이 방법은 POI의 면적정보가 없는 제한적인 상황에서도 시맨틱 궤적을 생성할 수 있다는 장점을 가진다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4706-4724
• /
• 2020

With various structured data, such as the company size, loan balance, and savings accounts, the voice of customer (VOC), which is text data containing contact history and counseling details was analyzed in this study. To analyze unstructured data, the term frequency-inverse document frequency (TF-IDF) analysis, semantic network analysis, sentiment analysis, and a convolutional neural network (CNN) were implemented. A performance comparison of the models revealed that the predictive model using the CNN provided the best performance with regard to predictive power, followed by the model using the TF-IDF, and then the model using semantic network analysis. In particular, a character-level CNN and a word-level CNN were developed separately, and the character-level CNN exhibited better performance, according to an analysis for the Korean language. Moreover, a systematic selection model for optimal text mining techniques was proposed, suggesting which analytical technique is appropriate for analyzing text data depending on the context. This study also provides evidence that the results of previous studies, indicating that individual customers leave when their loyalty and switching cost are low, are also applicable to corporate customers and suggests that VOC data indicating customers' needs are very effective for predicting their behavior.

• Journal of Information Science Theory and Practice
• /
• v.8 no.4
• /
• pp.67-86
• /
• 2020

Information need has been one of the main motivations for a person using a search engine. Queries can represent very different information needs. Ironically, a query can be a poor representation of the information need because the user can find it difficult to express the information need. Query Expansion (QE) is being popularly used to address this limitation. While QE can be considered as a language-independent technique, recent findings have shown that in certain cases, language plays an important role. Arabic is a language with a particularly large vocabulary rich in words with synonymous shades of meaning and has high morphological complexity. This paper, therefore, provides a review on QE for Arabic information retrieval, the intention being to identify the recent state-of-the-art of this burgeoning area. In this review, we primarily discuss statistical QE approaches that include document analysis, search, browse log analyses, and web knowledge analyses, in addition to the semantic QE approaches, which use semantic knowledge structures to extract meaningful word relationships. Finally, our conclusion is that QE regarding the Arabic language is subjected to additional investigation and research due to the intricate nature of this language.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.441-443
• /
• 2021

Sentence or text similarity is a measure of the degree of similarity between two sentences. Techniques for measuring text similarity include Jacquard similarity, cosine similarity, Euclidean similarity, and Manhattan similarity. Currently, the cosine similarity technique is most often used, but since this is an analysis according to the occurrence or frequency of a word in a sentence, the analysis on the semantic relationship is insufficient. Therefore, we try to improve the efficiency of analysis on the similarity of sentences by giving relations between words using ontology and including semantic similarity when extracting words that are commonly included in two sentences.