人力資源網站知識管理與e面談之探討

Knowledge Management and e-Interview for a Human Resource Management Web Site

張志隆(Chih-Lung Chang)

指導教授 ： 夏延德

中原大學/工學院/資訊工程研究所/碩士(2001年)

https://doi.org/10.6840/cycu200100470

摘要

網路科技的快速發展，人們開始透過網路在人力資源網站求職求才。但一般的人力銀行網站，對於專長的描述皆以文字字串的方式，卻無法真正針對專長比對。只對一般的條件如職業類別、學歷、工作地點等等來查詢，所搜尋到的資訊卻往往不見得都是我們所需要的工作或人才。

本論文最主要的目的在於改善目前人力資源網站對於專長比對的作法。將文字字串的專長敘述轉換成選項式的方式表達，以便進行有效的條件比對，達成排序之需求。原本我們系統想利用Ontology架構來決定出兩兩「專長選項」間的相關性，藉由專長的相關性來加以排序。但由於某些專長選項在分類上有所爭議，以致於在「相關性」的表達上無法達到「一般化」，進而無法達到專長精確比對﹙Exact Match﹚與不精確比對﹙Inexact Match﹚且排序的目的。為了達到目的，本論文提出一套解決方式—「專長相關表」。

目前的人力銀行網站對於求才者與求職者兩方，大部份只做到居間的仲介，而忽略兩方之間的互動以及面試所花費的成本。所以系統也提供—「線上面談」的機制，目的是為了讓求職者與求才者在線上能達成初步的了解，藉由系統的輔助增進彼此的互動與溝通。在往後真正的面試時，彼此之間已有所認識。

關鍵字

ontology，專長相關表，不精確比對，線上面談

並列摘要

As a result of the fast development of Internet technology, people begin to search for jobs ( and companies begin to search for employees ) on the human management resource web site through the Internet. In general, a human management resource web site uses word string to describe job specialty requirements, but a string can not be used to match with specialty. We can only use career classification or work place to search, and this is unsatisfactory.

The main purpose of this thesis is to improve the ways of making specialty comparisons for a human management resource web site. Our method changes word string into items, and it can process condition comparisons and even do sorting. Originally, our system use an ontology framework to decide about the correlation between two「specialty items」, and we sort information by specialty correlation. Because some specialty items can raise some disputes in their classifications, the expression of specialty correlation cannot attain the goal of「generalization」,「exact match of specialty」,「inexact match of specialty」, and information. Sorting. To reach these goals, this thesis suggests another solution—the case of「specialty correlation table」.

Presently, most human management resource web sites impersonate an agent for both an employer and a job seeker, and they ignore interactions of between an employer and a job seeker. Our system provides a solution for this—「online interview」. The goal is to reach initial acquaintance between both an employer and a job seeker online. This way, when the real interview occurs, both the employer and the job seeker already know each other to some extent.

並列關鍵字

ontology，specialty correlation table，inexact m

參考文獻

1.鄭清文， “電子商務環境下職業選擇輔助系統建置之研究” ，元智大學資訊社會學研究所碩士論文，1999。

6.黃嘉能， “Ontology-based Slicing Hypermedia Electronic Book” ，元智大學資訊工程研究所 ，1999。

8.Chia-Lin Hsiao(蕭嘉琳)、 Gwo-Jen Hwang (黃國禎)、Judy C.R. Tseng(曾秋蓉)， ”互動式概念關係建立輔助系統在學習診斷之應用 ”，2001，全球華人學習科技論文集。

12.Sanjiv K. Bhatia and Jitender S. Deogun， ” Conceptual Clustering in Information Retrieval ”，IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART B: CYBERNETICS, VOL. 28, NO. 3, JUNE 1998，p427-p435。

13.Y-S. Chen, T.-H. Chu, “A neural network classification tree，” IEEE International Conference on Neural Networks, vol. 1, 1995, pp. 409-413.