Dynamic programming for re-mapping noisy fixations in translation tasks
Dynamic programming for re-mapping noisy fixations in translation tasks
复制标题
用于重新映射翻译任务中的噪声注视点的动态编程
DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
M. Carl
中科院分区：
文献类型：
--
作者：
M. Carl
眼球追踪器允许自由的头部运动，在很多情况下不精确到阅读模式严重扭曲的程度。这些凝视模式的糟糕可用性和可解释性被一种“幼稚的”注视到符号的映射所证实，这种映射经常错误地将观察到的注视的可能漂移的中心映射到它正下方的符号上。在本文中，我通过介绍翻译任务的背景知识来扩展这种朴素的注视-符号映射。在第一步中，将注视到符号的映射序列扩展到几个可能的注视符号的晶格中，包括那些在原始注视映射的上下线上的符号。第二步，动态规划算法根据连续注视点与符号位置之间的字符距离、单词距离和像素距离，运用多种启发式算法寻找通过点阵的最佳路径，从而根据背景注视模型平滑注视路径。定性和定量评价表明，该算法提高了重新映射符号序列的准确性
Eyetrackers which allow for free head movements are in many cases imprecise to the extent that reading patterns become heavily distorted. The poor usability and interpretability of these gaze patterns is corroborated by a "naive" fixation-to-symbol mapping, which often wrongly maps the possibly drifted center of the observed fixation onto the symbol directly below it. In this paper I extend this naive fixation-to-symbol mapping by introducing background knowledge about the translation task. In a first step, the sequence of fixation-tosymbol mappings is extended into a lattice of several possible fixated symbols, including those on the line above and below the naive fixation mapping. In a second step a dynamic programming algorithm applies a number of heuristics to find the best path through the lattice, based on the probable distance in characters, in words and in pixels between successive fixations and the symbol locations, so as to smooth the gazing path according to the background gazing model. A qualitative and quantitative evaluation shows that the algorithm increases the accuracy of the re-mapped symbol sequence