Nowadays, many people can participate in intelligence production through various processes, such as publications, distributions, and services. In order to recognize, interpret, and process opinions and sentiments over the web, sentiment analysis is emerging as an important issue in human-centric information technology. Paradigms of sentiment analysis such as machineunderstandable web and human-centric web offer a promising and potential solution to mining and analyzing the very large and varied web data. To improve computers’ understanding of human-centric information, relationship analysis between persons with shared common sense information known to everyone is necessary. The seven papers in this Special Issue address some of the challenges inherent in this. The first paper “Human Likeness: Cognitive and Affective Factors Affecting Adoption of Robot-Assisted Learning Systems” by Hosun Yoo, Ohbyung Kwon, and Namyeon Lee proposes an integrated theoretical model, which explains the effect of human likeness of robots on user’s propensity to adopt robot-assisted learning systems. The human likeness is conceptualized as a combination of media richness, multimodal interaction capabilities, and parasocial relationships. To validate the proposed model, a robot-assisted learning prototype was utilized and survey data were collected from general users. The resulting data were empirically tested, and the test results were provided and analyzed. The second paper entitled “Predicting Personality Traits Related to Consumer Behavior using SNS Analysis” by Baik et al. proposes a method for predicting the four personality traits – (1) Extroversion, (2) Public Self-Consciousness, (3) Desire for Uniqueness, and (4) Self-Esteem – that correlate with buying behaviors in the recent consumer behavior discipline. They also propose another method to analyze user behaviors in a social network service by using user behavior matrix, friendship analysis, and route analysis. In the third paper entitled “An Efficient Scheme for Automatic Web Pages Categorization using Support Vector Machine”, Vinod Kumar Bhalla and Neeraj Kumar proposes a support vector machine (SVM)-based web page categorization, which is based on identification of specific and relevant features of the web pages. They also developed a feature extraction tool based on the HTML-DOM of web page. They evaluated the proposed scheme using SVM kernel as a classification tool in combination with feature extraction and statistical analysis. The fourth paper entitled “Sentiment Classification Technology Based on Markov Logic Networks” by Hui He, Zhigang Li, Chongchong Yao, and Weizhe Zhang presents a crossdomain multi-task text sentiment classification method based on Markov Logic Networks. Through many-to-one knowledge transfer, labeled text sentiment classification knowledge was successfully transferred into other domains, and the precision of the sentiment classification analysis in the text tendency domain was improved.

中文翻译：

---

以人为本的信息技术与应用迈向 Web 3.0

如今，许多人可以通过发布、分发和服务等各种流程参与情报生产。为了识别、解释和处理网络上的意见和情绪，情绪分析正在成为以人为中心的信息技术中的一个重要问题。情感分析范式，例如机器可理解的网络和以人为中心的网络，为挖掘和分析非常庞大和多样化的网络数据提供了一种有前途和潜在的解决方案。为了提高计算机对以人为中心的信息的理解，需要对共享常识信息的人之间的关系进行分析。本期特刊中的七篇论文解决了其中固有的一些挑战。第一篇论文“人类的相似性：Cosun Yoo、Ohbyung Kwon 和 Namyeon Lee 的 Cognitive and Affective Factors Affecting of Robot-Assisted Learning Systems” 提出了一个综合理论模型，该模型解释了机器人的人类相似性对用户采用机器人辅助学习系统的倾向的影响。人类的相似性被概念化为媒体丰富性、多模式交互能力和准社会关系的组合。为了验证所提出的模型，使用了机器人辅助学习原型，并从一般用户那里收集了调查数据。对所得数据进行了经验测试，并提供和分析了测试结果。Baik 等人的第二篇论文题为“使用 SNS 分析预测与消费者行为相关的个性特征”。提出了一种预测四种人格特质的方法——（1）外向性，(2) 公众自我意识，(3) 对独特性的渴望，以及 (4) 自尊——与最近消费者行为学科中的购买行为相关。他们还提出了另一种方法，通过使用用户行为矩阵、友谊分析和路线分析来分析社交网络服务中的用户行为。在题为“An Efficient Scheme for Automatic Web Pages Categorization using Support Vector Machine”的第三篇论文中，Vinod Kumar Bhalla 和 Neeraj Kumar 提出了一种基于支持向量机 (SVM) 的网页分类，它基于特定和相关特征的识别的网页。他们还开发了基于网页的 HTML-DOM 的特征提取工具。他们使用 SVM 核作为分类工具，结合特征提取和统计分析来评估所提出的方案。何慧、李志刚、姚崇崇、张伟哲在题为“基于马尔可夫逻辑网络的情感分类技术”的第四篇论文中，提出了一种基于马尔可夫逻辑网络的跨域多任务文本情感分类方法。通过多对一的知识迁移，将标记文本情感分类知识成功迁移到其他领域，提高了文本趋势域情感分类分析的精度。