• Journal of the Korean Society for Library and Information Science
• /
• v.48 no.4
• /
• pp.91-113
• /
• 2014

Although social metadata has strengths in creating amount of user-contributed resource descriptions, its function is limited because of its non-systematic characteristics. This research proposed an alternative approach to semantic organization of social metadata. It analyzed the semantics of tags created in LibraryThing in order to provide bibliographic categories for describing information resources. Social information Architecture is adopted in generating the bibliographic categories so that social metadata framework can be constructed. This framework can provide the conceptual foundations for semantically organizing social metadata and is expected to be applied to the existing approaches to automatically organize social metadata.

• Journal of Information Management
• /
• v.42 no.3
• /
• pp.147-165
• /
• 2011

This paper presents a core meta-tag of OpenURL written in Key/Encoded-Value format in the field of observation research, to distribute the scientific data, produced in many experimentations and observations, on the OpenURL service architecture. So far, the OpenURL hasn't supplied a meta-tag represented scientific data because it has focused on circulation of scholarly and technological information extracted from thesis, proceedings, journals, literatures, etc. The DataCite consortium metadata were analyzed and compared with the Dublin Core metadata, OECD metadata, and Directory Interchange Format metadata to develop a core meta-tag in observation research.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.9
• /
• pp.927-935
• /
• 2010

In the era of Web 2.0, users create a number of their own Web contents. So, multimedia search becomes much more important than ever. A tag is a simple keyword which describes the Web contents including URL, pictures, and videos. Tags perform a role of descriptors of Web contents and Web metadata properly. If the number of tagged Web data increases, users are more likely to find the desired search result because the system includes the Web contents which have richer Web metadata. However, the number of users who use tags as Web metadata is relatively small. Because of the cumbersome process of adding tags, or users do not know what to add for the better accessibility from the public. Given situation, tag recommendation, which helps the process of adding tags, has been studied to solve these problems. When a user adds some Web contents, the tag recommendation system recommends relevant tags for the Web contents to the use, and the user selects recommended tags. We analyze and categorize various tag recommendation algorithms in tagging system.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10c
• /
• pp.72-77
• /
• 2007

웹의 성장에 따른 기하급수적인 정보의 축적으로 인한 정보과다(Information Overload) 현상의 심화를 해결하기 위해 이루어져 온 많은 연구 중 하나인 추천 시스템은 사용자에게 고수준의 편의성을 제공하기 위한 시스템으로써 발전해 왔다. 그러나 과거에 고도로 집중화되어 관리, 구축되어 오던 정보와는 달리 Web2.0라는 새로운 웹 환경의 도래와 함께 태그, 블로그 등 새로운 형태와 특성을 가지는 점보들이 등장하게 되었다. 웹의 컨텐츠에 대한 메타정보를 사용자가 직접 입력한 Web2.0 기반의 태그 데이터론 활용해서 추천 시스템의 성능을 향상시킬 수 있는 기법을 연구하였다. 추천 기법 중 가장 대표적이고 기초적인 협업 필터링 기법에 태그를 활용하며 태그에 사용자에 대한 중요도를 감안한 가중치 부여 기법에 연구한다. 유사한 성향을 가진 사용자를 식별하는데 있어 태그 집합간의 유사도를 비교하는 방법을 사용하며 사용자의 성향을 반영하기 위해서 태그와 사용자의 선호도 정수와의 연관성을 분석해서 이를 태그의 가중치로 환산하는 기법을 제안한다.

• The Korean Journal of Archival Studies
• /
• no.67
• /
• pp.99-138
• /
• 2021

In the era of big data, analyzing not only structured data but also unstructured data is emerging as an important task. Official documents produced by government agencies are also subject to big data analysis as large text-based unstructured data. From the perspective of internal work efficiency, knowledge management, records management, etc, it is necessary to analyze big data of public documents to derive useful implications. However, since many of the public documents currently held by public institutions are not in open format, a pre-processing process of extracting text from a bitstream is required for big data analysis. In addition, since contextual metadata is not sufficiently stored in the document file, separate efforts to secure metadata are required for high-quality analysis. In conclusion, the current official documents have a low level of machine readability, so big data analysis becomes expensive.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.4
• /
• pp.401-411
• /
• 2008

Folksonomy, which is created through the collaborative tagging from many users, is one of the driving factors of Web 2.0. Tags are said to be the web metadata describing a web document. If we are able to find the semantic subsumption relationships between tags created through the collaborative tagging, it can help users understand the metadata more intuitively. In this paper, targeting del.icio.us tag data, we propose a method named FolksoViz for deriving subsumption relationships between tags by using Wikipedia texts. For this purpose, we propose a statistical model for deriving subsumption relationships based on the frequency of each tag on the Wikipedia texts, and TSD(Tag Sense Disambiguation) method for mapping each tag to a corresponding Wikipedia text. The derived subsumption pairs are visualized effectively on the screen. The experiment shows that our proposed algorithm managed to find the correct subsumption pairs with high accuracy.

• Journal of Information Management
• /
• v.42 no.4
• /
• pp.39-58
• /
• 2011

This paper presents a meta-tag expanded from scholarly service types of OpenURL written in Key/Encoded-Value format, after analyzing new scholarly service types and DataCite metadata elements which are for research data publishing and services. So far, OpenURL Z39.88 standard, KEVFormat: Sch-Svc, supporting six scholarly service type only, the expansion of this standard is needed for a research data circulation. New eight scholarly service types were extracted, after analyzing and comparing with the Scopus, Web of Science, and NDSL services. And nine representative attributes were extracted, after analyzing intensively the DataCite's elements.

• Journal of the Korean Society for information Management
• /
• v.28 no.4
• /
• pp.161-182
• /
• 2011

Metadata and folksonomy are two main approaches in representing, organizing, and retrieving resources in the current information environment. Many researches have conducted studies to combine of metadata and folksonomy in order to utilize the strengths of both approaches. This research proposed an approach to utilize both metadata and folksonomy in representing resources by using microcontents. Microcontents in this research is a conceptual structure that reflects dynamic characteristics of folksonomy and the structure of metadata. By connecting folksonomy with metadata through this microcontents structure, both approaches can maximize their strengths and minimize their weaknesses in representing, organizing, and retrieving resources.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10b
• /
• pp.1159-1162
• /
• 2001

본 연구는 웹 컨텐츠 통합 서비스에 관한 것으로 메타 브라우저, 중계 웹 서버, 포린 페이지 저작기, 포린 페이지 저장기로 구성한다. 메타 브라우저를 통해 사용자가 웹 사이트를 탐색하면서 웹 컨텐츠를 선택하며, 포린 페이지 저작기를 통해 각 사이트의 컨텐츠들로 포린페이지를 저작한다. 중계 웹 서버에서는 포린 페이지에 사용된 컨텐츠를 주기적으로 모니터링하여 컨텐츠 변화 감지시에 해당 컨텐츠로 구성된 포린페이지도 자동으로 갱신한다. 컨텐츠 추출을 위해 뭔 문서로 태그 트리를 구성하고, 그룹 시간 관계를 정의하여 포린 페이지 재생 모델을 제시했으며, 동기화를 위해 종료 제한 시간을 예측한다. 컨텐츠 변화 탐지 및 자동 갱신을 위해 컨텐츠 태그 트리와 웹 문서의 테그 트리간 차이값을 구하여 컨텐츠 변화 감지 방법을 제시한다.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.10
• /
• pp.399-406
• /
• 2021

With the steady growth of the content industry, the need for research that automatically recommending content suitable for individual tastes is increasing. In order to improve the accuracy of automatic content recommendation, it is needed to fuse existing recommendation techniques using users' preference history for contents along with recommendation techniques using content metadata or features extracted from the content itself. In this work, we propose a new graph-based music recommendation method which learns an LSTM-based classification model to automatically extract appropriate tagging words from sound data and apply the extracted tagging words together with the users' preferred music lists and music metadata to graph-based music recommendation. Experimental results show that the proposed method outperforms existing recommendation methods in terms of the recommendation accuracy.