應用相關回饋提高個人相關網頁搜尋準確度

近年來隨著網路普及率的提高，以及Web 2.0已發展成熟，越來越多個人用戶參與網路上各式議題討論，而形成網路輿論。對名人而言，網路輿論不論是正面或負面的討論，都會影響到個人形象、支持度，因此名人有關心網路上與個人相關正、負網路輿論的需求。而當名人利用搜尋引擎搜尋與自己相關的最新網路輿論時，搜尋引擎無法快速地將最新相關議題排序至前，且近年來巨量資料時代來臨，當搜尋準確度越低，將造成使用者在資料篩選上極大的負擔。故本研究目的是建置適合個人使用的輿論情勢監測系統，應用使用者相關回饋，搭配本研究提出的每日關鍵字權重學習機制，作為系統學習、區別最新相關議題的基礎，參考過去三日關鍵字權重，賦予隔日搜尋結果排序權重，以提高個人相關網頁搜尋準確度。本研究結果顯示，本系統可提高個人相關網頁搜尋準確度，但每日搜尋準確度高低，與每日學習議題總數呈反比關係。

關鍵字

相關回饋；個人化搜尋；關鍵字權重

並列摘要

In recent years, because internet penetration rate is raised and the development of Web 2.0 in Taiwan have been stable, there are more and more personal users taking part in and discussing the internet issues, internet public opinion firmed in this way. For celebrity, the internet opinion will affect their image and support, no matter positive or negative. When celebrity searched the latest internet opinion about themselves by search engine, it couldn’t quickly sort the latest issues about themselves to top. In addition to facing the age of the Big data, if searched accuracy is lower, personal users will be more inefficient. Therefore, the research purpose is building a personal monitoring system of public opinion on internet. It is according to the user’s relevance feedback and daily learning mechanism of keyword weighting by the study. Both of them are the system’s basic of learning and distinguishing the latest related opinion. According to past three day’s result of keyword weighting, it can predict next day’s relevant sequence of searching result in order to improve next day’s personal web page search precision. Search results show that this personal monitoring system of public opinion on internet can improve personal web page search precision. But each day’s search precision is inversely proportional to each day’s total number of learning issues.

並列關鍵字

relevance feedback ； personalized search ； keyword weighting

參考文獻

Chen, L.C. (2011), Building a Web-snippet clustering system based on a mixed clustering method. Online Information Review, 35(4), 611-635.

Chris Buckley, Gerard Salton, James Allan. (1994). The effect of adding relevance information in a relevance feedback environment, In Proceedings of SIGIR 17, 292-300.

Ferragina, P. & Guli, A. (2008), A personalized search engine based on Web-snippet hierarchical clustering. Software: Practice and Experience, 38(2), 189-225.

Joachims, T., & Radlinski, F. (2007). Search engines that learn from implicit feedback. IEEE Computer, 40(8), 34-40.

Kelly, D., & Teevan, J. (2003). Implicit feedback for inferring user preference: a bibliography. ACM SIGIR Forum 37, 18-28

國際替代計量

應用相關回饋提高個人相關網頁搜尋準確度

未授權

主題瀏覽