| 摘要: | 生活腳步加快人們生活也充滿了壓力與緊張，多數人為了舒緩壓力乃藉由觀看生活上色 彩以調整自我情緒，因色彩所呈現之觀感具有影響心理情緒之特性，且情緒變化容易影 響人們生活或工作上行為。由於網路之便利性，人們亦習慣於網路上閱讀各類文章，因 此網頁之色彩變化所給予需求者情緒轉變係目前值得探討議題之一。雖然目前使用者可 透過網頁提供者獲得符合情緒之網頁，然目前分類技術尚缺針對網頁色彩為主情緒分類 機制，於此情緒網頁提供者僅能依靠自我觀點判定網頁所蘊含之情緒，進而推薦給予使 用者瀏覽，但此方式可能因提供者之主觀意識造成使用者觀看此份網頁感受不同。因此 本研究針對網頁文件為基礎發展一套「以顏色為基之網頁文件意象風格判定模式」，判 定網頁文件所蘊含之情緒，並協助情緒網頁提供者獲知網頁之情緒類別。 本研究所提出「以顏色為基之網頁文件意象風格判定」模式乃包含「使用者重點區 塊判定」模組與「網頁文件情緒類別判定」模組。首先，於「使用者重點區塊判定」模 組中，本研究乃考量使用者閱讀網頁之習慣（如網頁瀏覽動線），結合眼動追蹤以及網 頁區塊分割技術，解析網頁文件中區塊位置，以得知網頁重要區塊。接著，於「網頁文 件情緒類別判定」模組中，本研究乃擷取網頁文件標籤區域當中顏色代碼，以得知網頁 色彩分佈，進而從網頁色彩分佈狀況判定網頁文件之主要顏色，獲知目標網頁文件情緒 類別。本研究除發展模式與方法論外，並依此方法論建構一套「網頁文件意象風格判定」 系統以進行案例驗證，從而確認方法論與技術之可行性。 整體而言，本研究提出一套以顏色為基之網頁文件意象風格判定模式方法論，期望 藉由此方法可讓情緒網頁提供者能更客觀判定網頁文件情緒，並可讓網頁需求者直接獲 得符合情緒狀態之網頁文件。
With accelerating pace of life, people’s daily lives are full of stress and tension. Most people relieve stress by viewing colors in life to adjust emotions as perception of colors has psychological and emotional impact, while emotional changes are likely to affect people’s behaviors in life or work. Due to the convenience of Internet, people are accustomed to reading various articles on the Web. Therefore, emotional change arising from webpage color change has become one of the issues worth exploring. Although demanders can access webpage documents that are consistent with their emotions from the webpage providers, so far, there is no webpage color-based emotional classification mechanism. The webpage providers can only recommend webpage containing emotions upon subjective choices to demanders for browsing. Hence, due to the subjective consciousness of the webpage provider, the demanders may have different perceptions of the webpage. Therefore, this paper develops a “Color-based Webpage Image Style Judgment Model” for webpage to determine the inherent emotions and help webpage providers derive the emotional category of the webpage in advance. The proposed Color-based Webpage Image Style Judgment Model is composed of the Webpage Key Block Determination (WKBD) module and Webpage Emotional Category Determination (WECD) module. First, in the WKBD module, this paper considers the webpage browsing habits of users (e.g., the webpage browsing path), and combined eye movement tracking technology as well as the webpage block segmentation technology to analyze the block position in the webpage and explore the critical blocks of the webpage. Next, in the WECD module, this paper captures color codes contained in the webpage tags to explore the webpage color distribution, by which the major colors of the webpage are determined to obtain the emotional category of the target webpage. In addition to the proposed model and methodology, this paper also develops a Web-based system for case verification to confirm the feasibility of the methodology and the technology. Overall, this paper proposes a methodology of the “Color-based Webpage Image Style Judgment Model”, to enable emotional webpage providers to more objectively judge the emotions of the webpage documents, and the webpage demanders to directly find webpage that match their emotional state. |
