教師著作

Browse

Recent Submissions

Now showing 1 - 5 of 560
  • Publication
    以鏈結資料發展電影資訊系統之研究
    (國家圖書館, 2015-06) 鄭, 智遠 ; 柯, 皓仁 ; 柯, 皓仁
    電影是人們普遍的休閒活動之一,電影資訊也是人們經常透過網際網路查找的資訊之一,然而網際網路上龐大的電影資訊卻也有著資訊分散、重複性和不一致的問題。鏈結資料是實踐語意網的最佳實務,其基於資料的機制有助於不同系統間的資料交換與整合。本研究運用內容管理系統Drupal建構一套融入鏈結資料精神,以IMDb的電影、導演和演員資料為基礎,整合DBpedia、LinkedMDB、DBtune,以及大英圖書館等鏈結資料來源的電影資訊系統,使電影資訊內容更加豐富且提升鏈結資料的閱讀性。最後並藉由系統使用性量表來評估該系統的使用性,了解使用者對本系統的接受度,評估結果顯示本系統的使用性是可接受的。Movie watching is a common leisure activity, and thus people seek for information about movies from the Internet very often. However, for such a large data, there are prob-lems derived like information scattering, duplication and inconsistency. Linked Data is a best practice for Semantic Web, and the data-driven technique helps the exchange and in-tegration of data among systems. This study creates a linked data-based movie information system by the content management system Drupal. This system exploits the information on movies, directors, and actors from IMDb, and integrates the information from several linked-data producers, including DBpedia, LinkedMDB, DBtune, and British Library. In this manner, the system provides rich movie information and enhance readability of linked data. Finally, this study evaluates this linked data-based movie information system by the System Usability Scale (SUS) to understand user usability and acceptance of the system, and the SUS evaluations show that this system’s usability is acceptable.
  • Item
    A study on mental models of taggers and experts for article indexing based on analysis of keyword usage.
    ( 2014-08-01) 柯皓仁 ; Ya-Ning Chen ; Hao-Ren Ke
    This article explores the mental models of article indexing of taggers and experts in keyword usage. Better understanding of the mental models of taggers and experts and their usage gap may inspire better selection of appropriate keywords for organizing information resources. Using a data set of 3,972 tags from CiteULike and 6,708 descriptors from Library and Information Science Abstracts (LISA) from 1,489 scholarly articles of 13 library and information science journals, social network analysis and frequent-pattern tree methods were used to capture and build up the mental models of article indexing of taggers and experts when using keywords, and to generalize their structures and patterns. When measured with respect to the terms used, a power-law distribution, a comparison of terms used as tags and descriptors, social network analysis (including centrality, overall structure, and role equivalent) and frequent-pattern tree analysis, little similarity was found between the mental models of taggers and experts. Twenty-five patterns of path-based rules and 12 identical rules of frequent-pattern trees were shared by taggers and experts. Title- and topic-related keyword categories were the most popular keyword categories used in path-based rules of frequent-pattern trees, and also the most popular members of 25 patterns and the starting point of the 12 identical rules.
  • Publication
    淺談公共圖書館績效評估
    (臺北市立圖書館, 2014-12-01) 柯皓仁 ; 柯, 皓仁
    圖書館績效評估是圖書館整體品質管理的一環,藉由任務與目標的制定,研擬資源分配與行動方案,再運用績效評估衡量行動方案的成果是否契合任務與目標。近年來,由於政經社會環境的改變,民眾、社區與政府皆要求公共圖書館展現其營運成效,是以公共圖書館績效評估愈形重要。本文闡述圖書館績效評估的定義,再介紹德國、美國、以及我國實施之公共圖書館績效評估相關計畫;最後,則說明從投入、產出層面的績效評估進展到成效評估。
  • Item
    臺灣地區大學圖書館特色館藏發展之研究
    (國立臺灣大學圖書館, 2013-09-01) 賴雅柔 ; 柯皓仁
  • Item
    FRBRoo-based approach to heterogeneous metadata integration.
    ( 2013-01-01) 柯皓仁 ; Ya-Ning Chen ; Hao-Ren Ke
    This paper seeks to adopt FRBRoo as an ontological approach to integrate heterogeneous metadata, and transform human-understandable format into machine-understandable format for semantic query. Design/methodology/approach – Two cases of use with museum artefacts and literary works were exploited to illustrate how FRBRoo can be used to re-contextualize the semantics of elements and the semantic relationships embedded in those elements. The shared ontology was then RDFized and examples were explored to examine the feasibility of the proposed approach. Findings – FRBRoo can play a role as inter lingua aligning museum and library metadata to achieve heterogeneous metadata integration and semantic query without changing either of the original approaches to fit the other. Research limitations/implications – Exploration of more diverse use cases is required to further align the different approaches of museums and libraries using FRBRoo and make revisions. Practical implications – Solid evidence is provided for the use of FRBRoo in heterogeneous metadata integration and semantic query. Originality/value – This is the first study to elaborate how FRBRoo can play a role as a shared ontology to integrate the heterogeneous metadata generated by museums and libraries. This paper also shows how the proposed approach is distinct from the Dublin Core format crosswalk in re-contextualizing semantic meanings and their relationships, and further provides four new sub-types for mapping description language.