• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.01a
• /
• pp.181-182
• /
• 2015

음악 검색에 있어 소셜 태그 정보는 사용자로 하여금 음악의 내재적 의미를 빠르게 파악할 수 있도록 한다. 음악의 소셜 태그 정보는 음악 추천 시스템을 활용하는 사용자(청취자)에 의해 점진적으로 완성되기 때문에 초기에 완전한 태그 정보를 수집하는 것은 어렵다. 본 논문에서는 음악의 일부 태그가 누락되어 있는 상황에서 음악 정보 검색을 자동으로 수행할 수 있는 클래스 분류 알고리즘을 제안하고자 한다.

• Journal of Internet Computing and Services
• /
• v.14 no.2
• /
• pp.1-13
• /
• 2013

Recent advance and popularization of smart devices and web application services based on cloud computing have made end-users to directly produce and, at the same time, consume the image contents. This leads to demands of unified contents management services. Thus, this paper proposestag clustering method based on semantic similarity for effective image categorization. We calculate the cost of semantic similarity between tags and cluster tags that are closely related. If tags are in a cluster, we suppose that images with them are also in a same cluster. Furthermore, we could recommend tags for new images on the basis of initial clusters.

• KIISE Transactions on Computing Practices
• /
• v.22 no.6
• /
• pp.290-295
• /
• 2016

Recently, folksonomy-based image-sharing sites where users cooperatively make and utilize tags of image annotation have been gaining popularity. Typically, these sites retrieve images for a user request using simple text-based matching and display retrieved images in the form of photo stream. However, these tags are personal and subjective and images are not categorized, which results in poor retrieval accuracy and low user satisfaction. In this paper, we propose a categorization scheme for folksonomy images which can improve the retrieval accuracy in the tag-based image retrieval systems. Consequently, images are classified by the semantic similarity using text-information and image-information generated on the folksonomy. To evaluate the performance of our proposed scheme, we collect folksonomy images and categorize them using text features and image features. And then, we compare its retrieval accuracy with that of existing systems.

• Journal of Computing Science and Engineering
• /
• v.3 no.3
• /
• pp.165-180
• /
• 2009

The HKIB, or Hankookilbo, test collections are two archives of Korean newswire stories manually categorized with semi-hierarchical or hierarchical category taxonomies. The base newswire stories were made available by the Hankook Ilbo (The Korea Daily) for research purposes. At first, Chungnam National University and KISTI collaborated to manually tag 40,075 news stories with categories by semi-hierarchical and balanced three-level classification scheme, where each news story has only one level-3 category (single-labeling). We refer to this original data set as HKIB-40075 test collection. And then Yonsei University and KISTI collaborated to select 20,000 newswire stories from the HKIB-40075 test collection, to rearrange the classification scheme to be fully hierarchical but unbalanced, and to assign one or more categories to each news story (multi-labeling). We refer to this modified data set as HKIB-20000 test collection. We benchmark a k-NN categorization algorithm both on HKIB-20000 and on HKIB-40075, illustrating properties of the collections, providing baseline results for future studies, and suggesting new directions for further research on Korean text categorization problem.

• Science of Emotion and Sensibility
• /
• v.18 no.4
• /
• pp.75-86
• /
• 2015

As a part of integrated study for technical communication between consumer and manufacturer of functional clothing, functional clothing consumers were categorized based on their perception of the functional outdoor jacket and the differences in the source of information and use of hang-tag were examined according to the groups. Questionnaire survey was conducted on 472 males and females in their 20s to 60s, who have experience of purchasing functional outdoor jackets. Perception of functional outdoor jackets was classified into four factors ('functionality', 'price', 'experience/knowledge' and 'brand/design') and the respondents were categorized into three groups according to the factors: 'unconversant/brand & design pursuing group', 'conversant/function pursuing group' and 'high-priced products preferring group'. 'Unconversant/brand & design pursuing group' was composed of high rate of office workers and students in their 20s and 30s who use internet most frequently as source of information while 'conversant/function pursuing group' was composed of high rate of professionals in their 40s and 50s who make the most frequent use of information on hang-tag. 'High-priced products preferring group', which has high rate of the elderly in their 60s and housewives, obtain their information mainly from salespersons. 'Unconversant/brand & design pursuing group' was satisfied with functional information more than other two groups and 'conversant/function pursuing group' concerned for the hang-tag and required more explanation of terminology used in hang-tag. The results effectively explained the features of consumer groups and their different responses toward the information of functional outdoor jacket.

• Journal of KIISE:Software and Applications
• /
• v.36 no.9
• /
• pp.749-757
• /
• 2009

A Internet user can easily access to the target information by web searching using some key-words or categories in the present Internet environment. When some meta-data which represent attributes of several data structures well are used, then more accurate result which is matched with the intention of users can be provided. This study proposes a multiple dimensional vector tagging method for the small web user group who interest in maintaining and sharing the bookmark for common interesting topics. The proposed method uses vector tag method for increasing the effect of categorization, management, and retrieval of target information. The vector tag composes with two or more components of the user defined priority. The basic vector space is created time of information and reference value. The calculated vector value shows the usability of information and became the metric of ranking. The ranking accuracy of the proposed method compares with that of a simply link structure, The proposed method shows better results for corresponding the intention of users.

• Parasites, Hosts and Diseases
• /
• v.49 no.3
• /
• pp.221-228
• /
• 2011

Rodent malaria parasites, such as Plasmodium berghei, are practical and useful model organisms for human malaria research because of their analogies to the human malaria in terms of structure, physiology, and life cycle. Exploiting the available genetic sequence information, we constructed a cDNA library from the erythrocytic stages of P. berghei and analyzed the expressed sequence tag (EST). A total of 10,040 ESTs were generated and assembled into 2,462 clusters. These EST clusters were compared against public protein databases and 48 putative new transcripts, most of which were hypothetical proteins with unknown function, were identified. Genes encoding ribosomal or membrane proteins and purine nucleotide phosphorylases were highly abundant clusters in P. berghei. Protein domain analyses and the Gene Ontology functional categorization revealed translation/protein folding, metabolism, protein degradation, and multiple family of variant antigens to be mainly prevalent. The presently-collected ESTs and its bioinformatic analysis will be useful resources to identify for drug target and vaccine candidates and validate gene predictions of P. berghei.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06c
• /
• pp.427-429
• /
• 2012

개인이 사용할 수 있는 스마트 기기가 다양해지면서 여러 기기로 생산된 사진 콘텐츠가 어떤 기준이나 규칙 없이 분산되어 있어 콘텐츠를 관리하고 원하는 콘텐츠를 검색하는 것이 어려워졌다. 따라서 본 논문에서는 개인 사진 콘텐츠를 효과적으로 분류하기 위하여 의미적 유사도를 기반으로 한 태그 클러스터링 기법을 제안한다. 태그들 사이의 유사도를 계산하여 서로 관련이 있다고 판단되는 태그들을 클러스터링 하는데, 태그가 같은 클러스터에 포함되어 있으면 그 태그를 가진 사진들도 유사성을 가진다고 볼 수 있으므로 개인 사진들을 의미에 따라 분류하는데 이용할 수 있다.

• Proceedings of the Korean Aquaculture Society Conference
• /
• 2003.10a
• /
• pp.44-44
• /
• 2003

Gene expression in five tissues of the abalone (Haliotis discus hannai) was investigated using an expressed sequence tag (EST) analysis. Randomly selected clones were obtained from cDNA libraries constructed with gill (GI), digestive diverticula(DD), hepatopancreas (HP), foot/mucus (FM) and rectangular muscle (RM). Of 1,235 clonesanalyzed (288 clones for GI, DD, HP each,166 for FM, and 205 for RM), 741 (60.0%) clones in total turned out to share significant similarity with the sequences from NCBI GenBank (less than 10/sup -3/ of e-values), 423 sequences showed poor similarity (> 10/sup -3/), and 71 sequences didn't match with any sequences in GenBank. The percent unique sequence (singleton) was ranged from 56.1% (RM) to 74.7% (FM) among libraries. On the other hand, overall percent singleton was 55.3% when all the ESTs from five libraries were assembled into contigs. Analysis of the organisms represented by the best hit for each EST (e-values < 10/sup -3/) showed that 23.8% matched with mammalian entries, 24.0% with mollusks, 14.4% with insects, 11.6% with fish and 26.2% with others. The expressed patterns differed among the tissues when judged by the categorization of the sequences from each library into 10 broad functional classes. In all the libraries, the class I (no hit o. poor similarity) was the largest category with an average of 40.1%. This largest class was followed by class V (general metabolisms) in DD (21.9%), GI (14.6%) and HP (16.7%), while the 'cell structure and motility'(class VI) was the second largest class in remaining two libraries (31.2% for RM and 9.6% for FM). The class IX (cell division and proliferation) was the smallest class in all the libraries (less than 3%). This report provides the first tissue-specific lists of expressed abalone genes, which could be a fundamental basis for genomics program of abalone species.

• Korean Security Journal
• /
• no.40
• /
• pp.57-86
• /
• 2014

As the revision of Building Code including applying crime prevention design to buildings passed recently and target hardening ought to be evidence-based, we studied the Modus Operandi (MO) and intrusion tools of domestic burglary to earn basic data for improvement of crime prevention hardware in the future. To be specific, we reviewed related academic literature and police official statistics of domestic burglary critically and interviewed detectives in charge of burglary to specify and categorize MO and tools. We can derive some implications from research findings, including improvement of the statistical system for the MO of burglary, active sharing of the MO of burglary among the criminal justice agencies and related industries and experts. Also, crime prevention advice and education for the local residents focused on MO of burglary can be recommended. Based on this research, to enhance the level of community safety significantly, performance tests of crime prevention hardware such as security doors and windows etc. and the study on related certification system should be vitalized.