An enhanced content selection mechanism for personalization of video news programmes

In this paper, we propose a content selection framework that improves the users’ experience when they are enriching or authoring pieces of news. This framework combines a variety of techniques to retrieve semantically related videos, based on a set of criteria which are specified automatically depending on the media’s constraints. The combination of different content selection mechanisms can improve the quality of the retrieved scenes, because each technique’s limitations are minimized by other techniques’ strengths. We present an evaluation based on a number of experiments, which show that the retrieved results are better when all criteria are used at time.     

A cognition-centered personalization framework for cultural-heritage content

User Modeling and User-Adapted Interaction - The heterogeneity of the audience of cultural heritage institutions introduces numerous challenges to the delivery of the content. Considering that...     

User interface features influencing overall ease of use and personalization

This study examines user perceptions of gathering information on, locating, and buying compact discs through Web-based interfaces. A group of part-time MBA students in Hong Kong performed a specific set of tasks using the Web sites of two vendors of compact discs. The tasks were selected to be general enough to apply to a variety of on-line shopping situations. As they performed tasks, they answered questions regarding their perceptions of several user interface features as well as their perceptions of overall ease of use. User interface features associated with a perception of overall ease of use were identified, as were differences based on personal characteristics of respondents. Also, differences in perceptions of interface features that might be due to personal characteristics, such as age and sex, were identified. The methodology and results help identify user interface features that a designer should emphasize and features for which personalization is likely to be beneficial.     

User interfaces for mobile content

The UI must be designed to facilitate the widespread and increased use of mobile content. Current UI approaches in the computing domain - such as windows, icons, menus, and pointing (WIMP) - are inadequate and inappropriate for mobile applications. Mobile device constraints pose several UI design challenges. UIs for emerging mobile devices consist of hardware and software combinations such as additional buttons for text input and special functional keys, touch-sensitive screens, audio/visual interfaces, and haptic interfaces to provide virtual-reality (VR) experiences. Analysts expect that third-generation mobile phones and beyond will be conduits for enormous amounts of multimedia content. Several major issues along a variety of dimensions must be addressed in the design and provision of content to these mobile devices. Because they are crucial for enhancing the user experience and enabling the effective consumption of mobile content, UIs are therefore among the major issues that need further study and work.     

面向新闻推荐的用户兴趣模型构建与更新

针对新闻推荐系统中用户兴趣模型构建与用户兴趣漂移问题,提出了一种面向新闻推荐的用户兴趣模型构建与更新方法。首先采用向量空间模型与bisecting K-means聚类算法构建了原始用户兴趣模型;然后以艾宾浩斯遗忘曲线为基础构造了遗忘函数,并以此对用户兴趣模型进行时间加权,从而达到对用户兴趣模型更新的目的。实验以基于用户的协同过滤推荐、基于物品的协同过滤推荐为baseline,实验结果表明所构建的原始用户兴趣模型推荐性能更优,在◢F◣值上提升了4%,更新后的模型与原始模型相比◢F◣值提高了1.3%。     

User comments for news recommendation in forum-based social media

News recommendation and user interaction are important features in many Web-based news services. The former helps users identify the most relevant news for further information. The latter enables collaborated information sharing among users with their comments following news postings. This research is intended to marry these two features together for an adaptive recommender system that utilizes reader comments to refine the recommendation of news in accordance with the evolving topic. This then turns the traditional “push-data” type of news recommendation to “discussion” moderator that can intelligently assist online forums. In addition, to alleviate the problem of recommending essentially identical articles, the relationship (duplicate, generalization, or specialization) between recommended news articles and the original posting is investigated. Our experiments indicate that our proposed solutions provide an improved news recommendation service in forum-based social media.     

Transportation ontology definition and application for the content personalization of user interfaces

Ontologies have been largely exploited in many domains and studies. In this paper, we present a new application of a domain ontology for generating personalized user interfaces for transportation interactive systems. The concepts, relationships and axioms of transportation ontology are exploited during the semi-automatic generation of personalized user interfaces. Personalization deals with the capacity of adaptation of a user interface, reflecting what is known about the user and the domain application. It can be performed on the interface container presentation (e.g., layout, colors, sizes) and in the content provided in their input/output (e.g., data, information, document). In this paper, the transportation ontology is used to provide the content personalization. This paper presents the ontology and how it is used for the personalization of user interfaces for developing transportation interactive systems by model-driven engineering.     

Computer anxiety and attitudes towards microcomputer use

Survey data gathered from 187 participants were used to examine the relationship between demographic variables, computer training and experience, management support and system quality and computer anxiety, and attitudes toward microcomputers. Results of hierarchical multiple regression analysis showed that the quality of the computer-based information system which represents the interface and the interaction between the participants and the system has a strong positive effect on attitudes toward microcomputers, and a significant reduction on computer anxiety. Computer training contributes strongly to decrease in computer anxiety and has an indirect effect on attitudes toward microcomputers. However, computer experience and management support were found to affect the attitudes toward microcomputers directly. Among the demographic variables, gender was the only one which correlated highly with computer anxiety. Implications for the design of information and decision support systems and future research are discussed.     

Differential responses of loyal versus habitual consumers towards mobile site personalization on privacy management

We examine how two different underlying mechanisms of behavioral loyalty to a brand—attitudinal loyalty and habit—impact smartphone users' privacy management when they browse personalized vs. non-personalized mobile websites. The online experimental study conducted with Amazon Mechanical Turk workers (N = 73) finds different responses of attitudinal loyalty and habit towards personalization in significant three-way interactions between personalization, attitudinal loyalty, and habit on privacy disclosure and protection behaviors. When interacting with a personalized website, highly habitual consumers without high level of attitudinal loyalty disclosed the most personal information on a personalized mobile site, and displayed the least intention of protecting their privacy on their smartphones, whereas consumers with high levels of both habit and attitudinal loyalty reported the highest tendency of privacy protection behavior. However, habit and personalization do not have a significant effect on disclosure behaviors when users have high attitudinal loyalty to a brand. Theoretical and practical implications are discussed.     

When does web-based personalization really work? The distinction between actual personalization and perceived personalization

A number of prior studies have suggested that personalization is more efficacious than non-personalization. However, the existing literature is somewhat ambiguous on whether the test of personalization effects should be based on a message sender's actual personalization process or a message recipient's perception of the message. It is argued in this article that an actual personalization process does not automatically yield more favorable effects because people's perceptions of personalized messages tend to be biased. Through three experiments, it is demonstrated that testing personalization effects based on a message sender's actual personalization process can be problematic and produce misleading results. Specifically, a personalized message can be perceived as non-personalized and a non-personalized message can be perceived as personalized. The key finding is that perceived personalization, instead of actual personalization, is the underlying psychological mechanism of message effectiveness. A message will show superior effects when it is perceived to be personalized by a message recipient, regardless of whether it is actually personalized or not.     

网络用户行为可信性集值统计度量方法

对网络用户行为可信性度量研究是当前网络安全的研究热点.本文通过分析网络用户的行为数据,建立用户可信行为评估树,提出用户行为等级划分及评判方法,从而确定可信度阈值.将集值统计方法引入可信性度量,提出一种动态多维度量算法,通过该算法确定网络用户行为数据评估值,并利用线性回归对用户行为进行预测.这种方法通过客观证据的收集来确定行为数据评估值,改变了以往靠主观来评定预测值的方式,并且将证据值扩大为一个值域.实验结果表明,该方法能够在网络环境中对用户行为进行准确预测,反映网络用户行为可信性的动态变化特征.     

Students' attitudes towards control methods in computer-assisted instruction

Abstract Computer-assisted instruction (CAI) has been used in the teaching of dental students for the last decade at the Royal Dental College Aarhus, Denmark. In oral radiology a number of programs structured according to different teaching principles have been available. The aim of this study was to investigate students' attitudes to three main structures: a fully teacher-controlled structure (TC), a menu based structure with partial learner control (structural communication, SC), and a fully learner-controlled structure (LC) applied on programs for oral radiology. Fifty-seven students were interviewed. Most students (53%) preferred the LC structure, and 56% recognized that they had most influence on program progression in this structure. Male and female students' attitudes differed, however, as significantly more males than females preferred the LC structure ( p < 0.05). No statistically significant differences were found between the two sexes when asked which structure provided the greatest learning effects and learner influence, but the female students found the SC as efficient as the LC structure.     

User modelling and cognitive user support: towards structured development

Model-driven engineering approaches have turned out useful when handling different perspectives on human–computer interaction, such as user profiles and problem-domain data. Their latest flavour, Model-Driven Architecture (MDA^?), targets towards platform-independent models (PIMs) and adjacent transformation mechanisms to adapt to user needs and tasks. Although in the field of user modelling and its major application domain, namely adaptive hypermedia systems (AHS), considerable effort has been spent on adaptation towards user needs, a structured development approach could not be established so far. User-oriented application designs are highly distinctive and can hardly be compared or mapped to novel or existing developments without major re-engineering effort. This paper develops an understanding of existing capabilities of already applied user-modelling techniques from a model-based perspective. Revealing the context of user models and user modelling allows determining general concepts for representing and processing knowledge for adaptation. The obtained findings show primarily technically motivated approaches, rather than designs grounded in findings from human factors. For human-centred design, a shift is suggested towards distributed cognition as a methodological and operational frame of reference for user modelling. This could help overcome existing limitations in adaptation. The corresponding research agenda requires directions on how to map psychological constructs to user-model elements and adaptable user-interface elements, such as mapping field dependence to content annotation features, in a transparent and empirically grounded way.     

Medical students' attitudes towards the use of virtual patients

An increasing number of virtual patients (VPs) are being used in the classroom, which raises questions about how to implement VPs to improve students' satisfaction and enhance their learning. This study developed and validated a scale that measures acceptability and attitudes of medical students towards the use of the VP education tool in the classroom. This mixed method study first explored attitudes with 11 students in two focus groups. Later on, eight experts performed item reduction and transformation through three rounds of the Delphi‐study method, and an initial version of the scale [virtual patient integration rating scale (VPIRS)] was developed. The scale was administered among 138 medical students to determine its reliability. A total of 88 medical students responded to the final version of the VPIRS. Principal component analysis was performed in order to determine questionnaire domains. The final, validated scale contains 25 items in four domains, with a reliability of 0.864. The identified domains are as follows: (1) acquiring and maintaining knowledge, (2) facilitation of learning, (3) inauthentic learning and (4) disadvantages of learning. To the best of our knowledge, this is the first time a VP rating scale has been developed. VPIRS allows for targeted engagement with students regarding learning and evaluation with VPs, thereby providing opportunities for student‐centred teaching.     

Product architecting for personalization

Personalization is an emerging manufacturing paradigm whereby customers can tailor products to their individual needs while maintaining high production efficiency. This paradigm necessitates “personalized product architecting” for determination of customizable/personalizable product modules and cost-effective manufacturing methods. This paper presents an initial effort in developing a method for identifying appropriate product architectures and manufacturing resolutions to achieve personalization considering functional utility and manufacturing cost. Ergonomic experiments and conjoint analysis are implemented to build functions relating manufacturability, price, and utility. Using these functions, a case study based on shoe products is conducted and the common integer programming welfare problem is expanded to a mixed-integer programming optimization problem for determination of a product family incorporating both personalized and customized offerings.     

User interface patterns in recommendation-empowered content intensive multimedia applications

Design Patterns (DPs) are acknowledged as powerful conceptual tools to improve design quality and to reduce time and cost of the development process by effect of the reuse of “good” design solutions. In many fields (e.g., software engineering, web engineering, interface design) patterns are widely used by practitioners and are also investigated from a research perspective. Still, they have been seldom explored in the arena of Recommender Systems (RSs). RSs provide suggestions (“recommendations”) for items that are likely to be appropriate for the user profile, and are increasingly adopted in content-intensive multimedia applications to complement traditional forms of search in large information spaces. This paper explores RSs through the lens of User Interface (UI) Design Patterns. We have performed a systematic analysis of 54 recommendation-empowered content-intensive multimedia applications, in order to: (i) discover the occurrences of existing domain independent UI patterns; (ii) identify frequently adopted UI solutions that are not modelled by existing patterns, and define a set of new UI patterns, some of which are specific of the interfaces for recommendation features while others can be useful also in a broader context. The results of our inspection have been discussed with and evaluated by a team of experts, leading to a consolidated set of 14 new patterns that are reported in the paper. Reusing pattern-based design solutions instead of building new solutions from scratch enables novice and expert designers to build good UIs for Recommendation-empowered content intensive multimedia applications more effectively, and ultimately can improve the UX experience in this class of systems. From a broader perspective, our work can stimulate future research bridging Recommender Systems, Web Engineering and Interface Design by means of Design Patterns, and highlights new research directions also discussed in the paper.

     

基于特征聚合的假新闻内容检测模型

针对假新闻内容检测中分类算法模型的检测性能与泛化性能无法兼顾的问题,提出了一种基于特征聚合的假新闻检测模型CCNN。首先,通过双向长短时循环神经网络提取文本的全局时序特征,并采用卷积神经网络（CNN）提取窗口范围内的词语或词组特征;然后,在卷积神经网络池化层之后,采用基于双中心损失训练的特征聚合层;最后,将双向长短时记忆网络（Bi-LSTM）和CNN的特征数据按深度方向拼接成一个向量之后提供给全连接层,采用均匀损失函数uniform-sigmoid训练模型后输出最终的分类结果。实验结果表明,该模型的F1值为80.5%,在训练集和验证集上的差值为1.3个百分点;与传统的支持向量机（SVM）、朴素贝叶斯（NB）和随机森林（RF）模型相比,所提模型的F1值提升了9~14个百分点;与长短时记忆网络（LSTM）、快速文本分类（FastText）等神经网络模型相比,所提模型的泛化性能提升了1.3~2.5个百分点。由此可见,所提模型能够在提高分类性能的同时保证一定的泛化能力,提升整体性能。     

面向新闻视频内容分析的音频分层分类算法*

提出了一种规则和隐马尔可夫模型相结合的音频分层分类算法,首先利用规则将新闻节目中的音频分为静音、语音和音乐三类,然后采用隐马尔可夫模型进一步将语音和音乐细分为男主持人语音、女主持人语音、交替报道、独白语音、现场语音和音乐六类。实验结果表明,男主持人语音、女主持人语音以及音乐的分类效果最好,查准率和查全率均可达90％以上;交替报道的分类性能最差,查准率为57.5％,查全率为79.3％;其他类别的分类性能居中,在70%～90％左右。与同类算法相比,该算法分类性能较高。