著者

村上 晴美 浦 芳伸 片岡 祐輔

出版者

一般社団法人 システム制御情報学会

雑誌

システム制御情報学会論文誌 (ISSN:13425668)

巻号頁・発行日

vol.29, no.2, pp.51-64, 2016-02-15 (Released:2016-05-15)

参考文献数

20

被引用文献数

1

---

To help users select and understand people while searching for them, we present a method of assigning Nippon Decimal Classification (NDC), which is a system of library classification numbers, to people on the web. By assigning NDC numbers to people, we can assign not only labels to people but also build a NDC-based people-search directory. We use a relative index in NDC, which lists the related index terms attached to NDC. We count the number of relative index terms contained in the titles of web pages (HTML files) and assign the top five NDC9 numbers to people. We developed a prototype of a people-search directory by assigning NDC9 numbers to HTML files that were manually classified from web people-search results. We evaluated the usefulness of our approach by comparing four methods and six documents and found that our method (extracting relative index terms) from the titles of web pages outperformed other methods and documents.

著者

村上 晴美 浦 芳伸 片岡 祐輔

出版者

一般社団法人 システム制御情報学会

雑誌

システム制御情報学会論文誌 (ISSN:13425668)

巻号頁・発行日

vol.29, no.2, pp.51-64, 2016

被引用文献数

1

---

To help users select and understand people while searching for them, we present a method of assigning Nippon Decimal Classification (NDC), which is a system of library classification numbers, to people on the web. By assigning NDC numbers to people, we can assign not only labels to people but also build a NDC-based people-search directory. We use a relative index in NDC, which lists the related index terms attached to NDC. We count the number of relative index terms contained in the titles of web pages (HTML files) and assign the top five NDC9 numbers to people. We developed a prototype of a people-search directory by assigning NDC9 numbers to HTML files that were manually classified from web people-search results. We evaluated the usefulness of our approach by comparing four methods and six documents and found that our method (extracting relative index terms) from the titles of web pages outperformed other methods and documents.