手持裝置中圖形化介面之順暢度評比

Abstract

隨著智慧型手持裝置的市占率增加，手持裝置的順暢度也逐漸成為使用者選購手機的因素之一。目前尚無一套標準的評比方法來評測智慧型手持裝置的圖形化介面之順暢度，一般常用來評估順暢度的指標為畫面更新率與反應時間，但單憑此兩個指標不足以代表使用者互動的順暢度。因此，我們分析使用者在智慧型手持裝置之使用行為的服務品質與經驗品質之關係來建立智慧型手持裝置之順暢度經驗品質模型。我們觀察的服務品質包含畫面改變的時間間隔之變異性、時間間隔之平均值、最大時間間隔、無反應次數、超過最大時間的次數、畫面改變量。經實驗證明各指標在對數關係時，除了最大時間間隔與畫面改變量，其他的指標的相關性高達71.5%以上，易造成共線性的問題並增加模型的誤差。由此實驗結果，我們推論最長的等待時間與畫面的不連續為造成不順暢感官的原因，分別可由最大時間間隔與畫面改變量兩指標來代表。因此，最大時間間隔與畫面改變量較適合為我們服務品質的指標。另一方面，我們設計一個是非題的問卷客觀地評估經驗品質，根據問卷結果，我們可以發現使用者對於不同的使用者行為有不同程度的順暢度感官。相較於三隻手機（HTC hero、Huawei U8860和Nexus S）的問卷結果，我們模型的錯誤率皆低於9%。

With the rapid growth of smartphones in market, the smoothness of smartphones becomes a crucial factor considered by consumers in making their buying decisions. However, there is no standard benchmark for comparison. In addition, the commonly used smoothness indexes, such as the frame rates and response time, cannot cover all aspects of smoothness of smartphones. In order to fairly evaluate the smoothness of smartphone, we developed a handheld smoothness evaluation over regression (HSER) model to benchmark the smoothness of smartphones. We first recorded a video and extracted several key indexes to represent behavior-based smoothness quality of services (BQoS), including the mean of frame intervals (MFI), variance of frame intervals (VFI), maximal frame interval (MaxFI), frame no response (FNR) and times of maximal frame interval (TMaxFI) and number of frame intervals (NFI). The correlation of MFI, VFI, FNR and TMaxFI is higher than 71.5% in logarithmic relationship. To avoid the collinearity problem which may lead to extra error, MaxFI and NFI are used to be the indexes for our HSER model. We next built up a relationship between BQoS and behavior-based smoothness quality of experience (BQoE). Finally, we converted BQoE to handheld smoothness quality of experience (HQoE). In our experiment, MaxFI and NFI are also good indexes for the “non-smooth” situations which have the long waiting time and the fragmentary frames. In addition, we tested three different smartphones, HTC hero, Huawei U8860 and Nexus S, to evaluate the applicability of HSER. Our results show that the proposed HSER model is able to fairly evaluate the smoothness of smartphones because the error rate of the HSER model is lower than 9%.