Evaluation of alternative waveforms for animated mimic displays

Animated mimic displays can be used to present system information regarding physical form, function, and causality. However, a potential limitation in current designs has been identified: the presence of ambiguous apparent motion. Two theoretical explanations of ambiguous apparent motion are discussed (Fourier and correspondence hypotheses). Two alternative designs (stair-step and approximate sinusoid luminance waveforms) were evaluated. The velocity matches obtained in Experiment 1 indicate that the sinusoidal waveform produced significantly better performance for both accuracy and latency than the stair-step wave-form. The velocity estimates obtained in Experiment 2 indicate that ambiguous apparent motion was not visible with the sinusoidal waveform, but was with the stair-step waveform. One of the two hypotheses (correspondence) provides a reasonable fit with the obtained velocity estimates. A fundamental goal in the design of animated mimic displays is to provide unambiguous mappings between perceived velocity and actual flow rates. Critical factors in design (e.g., waveform, chromatic/luminance contrast, spatial/temporal frequency) are discussed. Actual or potential applications of this research include the design of more effective animated mimic displays.     

A cognitive engineering approach to improving signalized left turn intersections.

This research evaluated the effect of providing advanced left turn information to individuals faced with deciding whether or not it is safe to turn at a signalized intersection. Younger (mean age 37 years) and older (mean age 71 years) drivers were tested in simulations of approaching an intersection with and without advanced cueing. Experiment 1 used a featureless background; Experiment 2 used an animated presentation of traffic control displays. In both experiments the subjects had to determine whether or not they had right-of-way to make a left turn. Exaggerated error rates and increased decision latencies for displays conveying unprotected turn status occurred in both studies. Cueing drivers with advanced notice of the decision rule through a redundant upstream posting of sign elements improved both accuracy and latency of younger and older drivers' decisions.     

基于多重信任的协同过滤推荐算法

针对评分数据稀疏性和用户冷启动所导致的协同过滤推荐系统的准确度与覆盖率较低的问题,文中融合显性信任和隐性信任因素,提出了一种基于多重信任的协同过滤推荐算法。首先,依据用户间推荐评分的准确性与可依赖度因子,提出一种改进的均方差(Mean Squared Difference,MSD)信任度量方法,并在此基础上提出基于隐性信任信息的评分模型;其次,以最大信任传播距离为约束,提出一种显性信任信息的关系模型;最后,依据评分相似性与显性信任关系,利用0-1背包组合优化策略选择出目标用户的最优近邻集合,从而进行评分预测。在Epinions数据集上与多种主流算法的对比仿真实验结果表明,该算法通过引入有效评分和显性信任关系,极大地缓解了数据稀疏性和冷启动问题,并且在不牺牲覆盖率的条件下显著提升了推荐准确度。     

对抛物方程使用新显格式的区域分解算法

我们提出了两个具有改进稳定性限制条件的新显格式．与经典显格式相比,稳定性限制条 件分别对两维抛物问题放宽了4倍,对一维问题放宽了2倍,同时它的精度与经典全隐格式 的相同．然后,我们通过在内边界点使用大步长的这种新显格式,在内点使用全隐格式,设计 了一个有限差分区域分解算法,稳定性限制条件分别对一维抛物问题放宽了2m2倍,对二维 问题放宽了4m2倍．从而我们能使用一个大的时间步长,这使我们在并行求解抛物问题时能 节省大量的计算量．     

Mitigating bias blind spot via a serious video game

We employed a serious video game to train participants on bias blind spot (BBS), capturing training effects on BBS mitigation and knowledge at three points in time. Experiment 1 (N = 703) compared the effects of hybrid training (a combination of implicit and explicit training) to implicit training; Experiment 2 (N = 620) tested the effects of just-in-time versus delayed feedback; and Experiment 3 (N = 626) examined the effects of singleplayer versus multiplayer learning environments. We also tested differences in game duration (30 vs. 60 min play) and repetition (single vs. repeated play). Overall, the video game decreased BBS linearly over time and increased BBS knowledge at posttest, but knowledge decayed at 8-week posttest. These and other results are discussed, along with the implications, limitations, and future research directions.     

融合显隐式反馈的协同过滤推荐模型

融合显式和隐式反馈已被应用于提升推荐模型的性能,但是,现有的此类推荐模型未能保留显式反馈中反映用户偏好程度的信息,且现有研究认为拥有显式反馈的数据和仅拥有隐式反馈的数据对于模型具有同等影响,未能充分发挥显式反馈的优势.针对这些问题,提出一种新的融合显式和隐式反馈的协同过滤推荐模型(CEICF).首先,所提出模型提取显式反馈中的特征得到用户/物品的全局偏好向量;然后,从隐式反馈中提取用户/物品的潜在向量,进而将两种向量进行融合得到用户/物品的偏好向量;最后,使用神经网络预测用户与物品交互的可能性.在训练模型时,定义一种加权的二进制交叉熵损失函数,加强显式反馈对模型的影响来增强模型捕获用户偏好的能力.为了验证所提出模型的有效性,在覆盖不同领域的现实数据集上进行实验,实验结果表明,CEICF可有效地融合显式和隐式反馈,且推荐效果相对于基线模型有显著提升.     

EARS: Emotion-aware recommender system based on hybrid information fusion

Recommender systems suggest items that users might like according to their explicit and implicit feedback information, such as ratings, reviews, and clicks. However, most recommender systems focus mainly on the relationships between items and the user’s final purchasing behavior while ignoring the user’s emotional changes, which play an essential role in consumption activity. To address the challenge of improving the quality of recommender services, this paper proposes an emotion-aware recommender system based on hybrid information fusion in which three representative types of information are fused to comprehensively analyze the user’s features: user rating data as explicit information, user social network data as implicit information and sentiment from user reviews as emotional information. The experimental results verify that the proposed approach provides a higher prediction rating and significantly increases the recommendation accuracy.     

一种无指导的隐式篇章关系推理方法研究

该文提出一种基于信息检索的无指导方法,用于推理隐式篇章片段之间的语义连接关系,如因果关系、转折关系等。该文基于Google搜索引擎,抽取在句子结构以及语义层面上均与原隐式片段相似的显式片段,通过分析和识别相关显式关系来间接推理隐式关系。主要包括以下三个模块 构建高质量查询关键词并抽取候选显式关系;结合三种隐式关系推理模型(相似度、置信度、关联度),综合考察查询关键词以及候选关系的质量;基于排序学习的方法,统计高质量候选关系中的类别分布以实现最终隐式关系的推理。该文采用Penn Discourse TreeBank 2.0篇章语料库,最终方法精确率达到54.3%,与有指导的方法相比,提高了约14.3%。     

面向任务口语对话系统中不含槽信息话语的端到端对话控制

端到端(end-to-end)模型因其能有效避免传统管道式设计存在的错误传递与累积问题,成为了近年来口语对话系统(spoken dialogue system, SDS)的研究热点。在面向任务SDS的end-to-end对话控制中,处理携带任务领域语义信息(槽信息)的话语可以结合命名实体识别、数据库查询结果等语义特征,而不含槽信息的话语,由于缺乏领域语义信息以及表达多样,其有效对话控制仍然是一个挑战。该文提出一种融合“显式”话语特征和“隐式”上下文信息的end-to-end混合编码网络用于处理不含槽信息话语。具体地,在应用卷积神经网络(convolutional neural network, CNN)对“显式”话语序列提取得到的特征表达的基础上,通过构造和捕获对话序列中“隐式”的系统后台上下文信息,进一步丰富了系统动作分类模型的特征表达。在限定领域面向中文任务SDS中的评估结果表明,与传统的管道式SDS和经典的end-to-end SDS相比,该文的方案在不含槽信息话语的单回合处理以及对话段整体性能上都得到了显著提升。     

Interacting with proactive public displays

In this paper we describe the design and architecture of an adaptive proactive environment in which information, which reflects the communal interests of current inhabitants, is proactively displayed on large-scale public displays. Adaptation is achieved through implicit communication between the environment and personal sensor devices worn by users. These devices, called Pendle, serve two purposes: they store and make available to the environment user preferences, and they allow users to override the environment's proactive behavior by means of simple gestures. The result is a smooth integration of environment-controlled interaction (experienced by the user as implicit interaction, triggered by their presence) and user-controlled explicit interaction. Initial results show that user-controlled adaptation leads to an engaging user experience that is unobtrusive and not distracting.     

Proximity compatibility and information display: effects of color, space, and object display on information integration

We report two experiments in which different features of the display of multiple channels of information are varied in their proximity to one another. The display presents three indicators of aircraft stall danger (airspeed, flaps, and bank). On some trials the stall danger is estimated, requiring information integration. On other trials the value of individual indicators is estimated, requiring focused attention. In Experiment 1 the spatial proximity between indicators and their distinctiveness in color were manipulated. Spatial proximity had little effect on either focused attention or integration, whereas a distinct color code improved focused attention and disrupted integration. In Experiment 2 the three indicators were presented as a bar graph or were combined as features of an object display, either with or without color coding of the separate dimension. Relative to the bar graph display, the object display improved information integration but disrupted focused attention. The presence of color borders restored the focused attention accuracy, with a slight cost to response time.     

一种新的JPEG图像认证半脆弱水印算法

提出了一种用于JPEG图像认证的半脆弱数字水印算法,它根据JPEG图像压缩过程中的DCT系数的不变性原理进行水印生成和嵌入调制,并结合一种有效的算法隐性地考虑到所有非零DCT系数的符号、大小关系,在未增加水印信息嵌入强度的情况下,实现了图像应对变换域攻击的更有效保护。同时,在检测算法中加入对虚警区的判断,增加了对窜改区域定位的准确性。实验表明,该算法实现的水印有很好的视觉透明度,有效的承受JPEG压缩,同时对其他恶意篡改敏感并能够准确定位篡改位置。     

一种结合显式特征和隐式特征的开发者混合推荐算法

现有开发者推荐算法通过对任务和开发者的显式信息进行挖掘, 抽取任务和开发者的显式特征, 完成针对任务的开发者推荐. 然而, 由于显式信息中的描述信息是主观的, 往往是不精确的, 现有基于显式特征的开发者推荐算法性能不够理想. 众包软件开发平台除包含大量不精确的描述信息外, 还包含客观的、较准确的"任务—开发者"成绩信息...     

改进的Hough变换检测圆方法

计算机视觉目前已经得到广泛的应用, 通过图像检测圆在工程应用中具有重要的意义. 为了在实时系统中自动检测图像中的圆并且精确得到圆心和半径, 提出了一种改进的Hough变换检测圆的方法. 先对图像边界进行过滤, 得到连续边界, 然后对每条连续边界进行排序并等分, 选择三点进行圆心和半径计算, 将传统Hough变换检测圆的三维参数降到一维参数来统计, 在降低了计算时间和内存资源占用的同时, 提高了检测圆的准确性. 实验表明, 该方法能够满足实际检测的需要. 同时, 在检测时间上, 可以满足实时性的要求.     

Establishing tradeoffs that leverage attention for utility: empirically evaluating information display in notification systems

Designing and evaluating notification systems represents an emerging challenge in the study of human–computer interaction. Users rely on notification systems to present potentially interruptive information in an efficient and effective manner to enable appropriate reaction and comprehension. Little is known about the effects of these systems on ongoing computer tasks. As the research community strives to understand information design suitable for opposing usage goals, few existing efforts lend themselves to extensibility.However, three often conflicting design objectives are interruption to primary tasks, reaction to specific notifications, and comprehension of information over time. Based on these competing parameters, we propose a unifying research theme for the field that defines success in notification systems design as achieving the desirable balance between attention and utility. This paradigm distinguishes notification systems research from traditional HCI by centering on the limitations of the human attention system.In a series of experiments that demonstrate this research approach and investigate use of animated text in secondary displays, we describe two empirical investigations focused on the three critical parameters during a browsing task. The first experiment compares tickering, blasting, and fading text, finding that tickering text is best for supporting deeper comprehension, fading best facilitates reaction, and, compared to the control condition, none of the animated displays are interruptive to the browsing task. The second experiment investigates fading and tickering animation in greater detail with similar tasks—at two different speeds and sizes. Here, we found smaller displays allowed better reaction but were more interruptive, while slower displays provides increased comprehension. Overall, the slow fade appears to be the best secondary display animation type tested. Focusing research and user studies within this field on critical parameters such as interruption, reaction, and comprehension will increase cohesion among design and evaluation efforts for notification systems.     

GUI objects with impenetrable borders: Instruction (not practice) makes perfect

Walker and Smelcer (Proceedings of the CHI 90, ACM, New York, 1990, pp. 221–225) found that menus could be selected faster if they were placed against the edge of the screen. Doing so creates an impenetrable border between the menu and the edge of the screen that the mouse cursor cannot penetrate. This changes how users move the mouse, so that selection times quicken compared to menus with a penetrable border. Experiment 1 investigated the effect that practice has on the acquisition of objects with and without impenetrable borders. The findings suggest that excessive practice was not necessary to demonstrate border type differences; thus, the advantage of having impenetrable borders seems to be relatively instantaneous. However, it was not readily apparent whether or not participants would realize the advantages of impenetrable borders without instruction. Thus, the primary purpose of Experiment 2 was to assess whether or not users would discover the benefits of impenetrable borders spontaneously. Participants were assigned to either the penetrable or impenetrable border condition. Additionally, participants received either full instruction concerning the benefits of the target placement, or limited instruction that simply informed the participant about the nature of the task. The results demonstrated that participants receiving limited instruction selected targets with impenetrable borders faster than participants who selected objects with a penetrable border. However, an exploratory comparison suggests that only 50% of participants who received limited instruction actually detected the impenetrable border. An additional comparison suggests that with practice the participants who were successful at detecting the impenetrable border selected the targets as quickly as participants who received full instruction concerning the benefits of impenetrable borders. The findings suggest that with full instruction, all users will perform reduced selection times. Given that not all participants discovered the impenetrable border it suggests that, whenever possible, users should receive instruction pertaining to the benefits of the impenetrable borders.     

On the Evolution of Vector Distance Functions of Closed Curves

Inspired by the work by Gomes et al., we describe and analyze a vector distance function approach for the implicit evolution of closed curves of codimension larger than one. The approach is set up in complete generality, and then applied to the evolution of dynamic geometric active contours in (codimension three case). In order to carry this out one needs an explicit expression for the zero level set for which we propose a discrete connectivity method. This leads us to make connections with the new theory of cubical homology. We provide some explicit simulation results in order to illustrate the methodology. Electronic supplementary material Electronic supplementary material is available for this article at and accessible for authorised users. First online version published in October, 2005     

Effects of Information Organization and Presentation on Human Performance in Simulated Main Control Room Procedure Tasks

The organization and presentation of information are critical in designing a display for industrial safety‐critical systems. How they affect human performance is not fully studied yet. Functional displays are newly proposed for nuclear power plant (NPP) main control rooms. The effect of functional organization and the choice of information presentation styles need to be examined. This study aims to explore the effects of information organization (process display vs. functional display), parameter presentation (text vs. bar chart), and component presentation (mimic vs. simplified graph) for better understanding the pros and cons of functional displays designed following function based task analysis (FBTA). Performance data were collected from simulated procedure tasks using eight (2 × 2 × 2) displays. The results indicate that the functional display based on FBTA was more effective than the process display, and this could be mainly attributed to its functional information organization rather than to its presentation styles of parameters and components.     

An evaluation of a "time tunnel" display format for the presentation of temporal information

The time tunnel display design technique combines the benefits of configural displays (salient visual properties corresponding to critical domain semantics) with the benefits of temporal information (i.e., the value of variables and properties over time). In Experiment 1 a baseline configural display and a time tunnel display were evaluated using real-time measures of system control, fault detection, and state estimation in a simulated process control task. The results provided little evidence in support of the time tunnel format. In Experiment 2 access to the temporal context was limited: Participants performed the detection and estimation tasks with static "snapshots" of system states that had been generated in Experiment 1. The overall pattern of results indicates that the time tunnel display was more effective for state estimation tasks than was the baseline configural display and or a trend display. Issues in the design of temporal displays are discussed, including representational formats and the choice of temporal time frames. Issues in the evaluation of temporal displays are also discussed, including the role of temporal information and the critical nature of participants' access to this information. Actual or potential applications of this research include design techniques for improving graphical displays and methodological insights to guide future evaluations.     

Top-down and bottom-up influences on learning from animations

To evaluate how top-down and bottom-up processes contribute to learning from animated displays, we conducted four experiments that varied either in the design of animations or the prior knowledge of the learners. Experiments 1–3 examined whether adding interactivity and signaling to an animation benefits learners in developing a mental model of a mechanical system. Although learners utilized interactive controls and signaling devices, their comprehension of the system was no better than that of learners who saw animations without these design features. Furthermore, the majority of participants developed a mental model of the system that was incorrect and inconsistent with information displayed in the animation. Experiment 4 tested effects of domain knowledge and found, surprisingly, that even some learners with high domain knowledge initially constructed the incorrect mental model. After multiple exposures to the materials, the high knowledge learners revised their mental models to the correct one, while the low-knowledge learners maintained their erroneous models. These results suggest that learning from animations involves a complex interplay between top-down and bottom-up processes and that more emphasis should be placed on how prior knowledge is applied to interpreting animations.