敦煌壁画数字图像语义描述方法研究

数字图像的语义描述与标注是解决图像检索中语义鸿沟问题的关键。由于缺乏面向领域的有效的数字图像描述方法规范,基于图像底层视觉特征的机器标注和基于专家知识的人为标注的标注结果都存在标注信息质量不高和结果不统一的问题。针对这一现实问题,本文基于图像元数据和信息需求理论,针对敦煌壁画数字图像这一特定文化遗产领域,提出了语义描述框架和领域主题词表相结合的数字图像内容语义描述方法,详细阐述了语义层次及其相互关系。同时,从图像语义描述粒度的角度讨论了语义粒度大小对标注成本的影响,以及该语义描述框架的可移植性问题。图4。表5。参考文献25。     

基于IIIF和语义知识图谱的印章资源整合与知识发现研究

[目的/意义] 数字人文研究的图像资源中蕴含大量信息但利用率极低，不能在异构数据库和不同的应用程序中得到有效的共享与重用，国际图像互操作框架打破了图像资源交换和共享的障碍。[方法/过程] 研究结合国际图像互操作框架和语义知识图谱（关联数据技术）进行图像资源的整合、共享与知识发现，对资源之间的关系进行揭示和知识推理，并通过CNNs算法对图像特征的提取与识别实现基于图像特征的语义检索辅助知识发现。[结果/结论] 提出一套数字人文图像资源整合与知识发现解决方案，并以印章图像资源为应用对象构建"印章知识中心"对以上解决方案的可行性和实践性进行实证检验。     

A semantic differential for speech correction concepts

The following study details the procedure by means of which a Semantic Differential for use in investigations in the area of speech correction was developed. It is the third in a series of studies by the same author to appear in SM, the preceding two having dealt with general speech and the theatre arts, respectively. The three studies are similar in scope and content, defining the problems, demonstrating the methodology, and offering suggestions concerning possible applications of the measuring instruments. The author acknowledges that the studies are exploratory and the results tentative. He suggests that the Semantic Differential offers a new approach to certain dimensions of meaning in the speech area which have been slow to yield to older methods.     

语义WikiI特性分析与应用研究

以Web2.0时代信息服务向知识服务转变为背景,结合语义Web技术,对智能语义Wiki做详细的分析研究。首先对语义Wiki特性进行分析,通过调研从标注机制、导航机制和检索机制三个方面展开深入研究,针对有代表性的语义Wiki项目进行比较和分析。     

“Cultumization?” The Impact of Cultural Priming on Customized Communication

This study examined the impact of cultural priming on the effects of customized communication. Two existing customization approaches (tailoring and targeting) were tested with 102 participants in an experiment. It was found that when participants were primed with individualistic meanings, they tended to generate a more favorable attitude toward tailored messages than targeted messages. In contrast, when participants were primed with collectivist meanings, they formed a more favorable attitude toward targeted messages than tailored ones. Based on the study findings, the conceptualization of customization needs to incorporate the cultural factor.     

基于对称性和GC矩的数字图书馆图像检索技术

针对数字图书馆中的图像检索问题,提出一种新型的基于内容的图像检索技术。这种方法的主要创新性在于在图像检索的过程中利用图像对称性特征及广义复合矩(GC矩)。一般人类的视觉功能对对称性的感知比较敏感,而GC矩具有仿射不变性,对于图像检索具有更高的鲁棒性。因此,配合使用对称性和GC矩进行检测,可以准确检出具有视觉相似性的图像。实验结果表明,对称性特征可以去除大量的冗余图像,提高检索精度和加快检索速度。     

Nonverbal cue context and episodic memory: On the availability and endurance of nonverbal behaviors as retrieval cues

Two investigations provide evidence that nonverbal cues (in this case, hand gestures) play a role as contextual cues in the retrieval of co‐occurring verbal messages. In the first study, type of hand gesture and processing strategies were investigated as possible determinants of message retrieval. Results indicated that the type of hand gesture was a major determinant of language recall. The second study investigated the durability and persistence of nonverbal cues in providing utterance retrieval. Results indicated that nonverbal cues provided retrieval of messages one week after exposure to conversational stimuli and that once again the type of gesture made a difference in the effectiveness of recall.     

以多语本体库为核心的跨语言信息检索映射技术研究进展——EuroWordNet案例分析

[目的/意义]通过对国内外多语本体领域映射技术相关研究成果的总结和EuroWordNet案例分析,为国内跨语言信息检索系统映射机制的建立提供借鉴和参考。[方法/过程]选取目前发展较为成熟的多语本体库EuroWordNet作为案例,分别从数据库设计、本体构建、概念存储和多语文化差异的映射处理4个方面对其中间语言索引机制(Inter-Lingual-Index,ILI)进行分析。[结果/结论]嵌入式的数据库结构设计、概念抽取及同义词集对应关系的界定、概念存储的细粒度化和复杂等价关系的建立是建立跨语言信息检索映射机制的关键。     

基于XML文本片段和内容归并的图像检索研究

研究利用XML文本片段和图像的内容特征（颜色）实现图像的检索。基于XML多媒体数字图书馆检索系统平台WHU-XML,对XML文本和图像构建索引,并在此基础上,采用线性归并法,实现基于XML文本片段的图像检索和基于图像内容特征（颜色）检索的结合。研究结果表明,当文本检索权重大于图像内容检索的权重时,检索效果比只采用单一检索方式时好。     

Learning to rank with (a lot of) word features

In this article we present Supervised Semantic Indexing which defines a class of nonlinear (quadratic) models that are discriminatively trained to directly map from the word content in a query-document or document-document pair to a ranking score. Like Latent Semantic Indexing (LSI), our models take account of correlations between words (synonymy, polysemy). However, unlike LSI our models are trained from a supervised signal directly on the ranking task of interest, which we argue is the reason for our superior results. As the query and target texts are modeled separately, our approach is easily generalized to different retrieval tasks, such as cross-language retrieval or online advertising placement. Dealing with models on all pairs of words features is computationally challenging. We propose several improvements to our basic model for addressing this issue, including low rank (but diagonal preserving) representations, correlated feature hashing and sparsification. We provide an empirical study of all these methods on retrieval tasks based on Wikipedia documents as well as an Internet advertisement task. We obtain state-of-the-art performance while providing realistically scalable methods.     

Culturally tailored narrative evidence about family health history: a moderated mediation analysis

This study tests a theoretical model named the Extended Elaboration Likelihood Model for Narrative Evidence Processing (EELM-NEP) to investigate associations between message processing, message outcomes, and behavioral intentions (BI) in the context of family health history communication. Furthermore, to investigate the role of cultural tailoring within these associations, this study explores how the type of culturally tailored narrative evidence and participants’ cultural identity moderate the relationships between message processing, message outcomes, and BI. A percentile bootstrap method and pairwise comparison tests in structural equation modeling (SEM) were employed. According to the results, identification significantly predicted engagement, and engagement had significant and positive effects on perceived message quality (PMQ), comprehension, perceived message effectiveness (PME), and BI. PME and resistance to counter-arguing significantly and positively affected BI. Importantly, the EELM-ENP model as well as pairwise comparison tests revealed that the messages were more effective when the participants read culturally distant messages. Findings suggest that cultural novelty might be more effective than cultural closeness in narrative persuasion.     

Text retrieval: an overview

The current state of the art of text retrieval, and indications of radical developments to come, should be of concern to all involved in text processing in the information chain from author to end‐user. This brief overview, slanted towards the interests of publishers, looks at the recent history of retrieval techniques, and present attempts to deal with information overload and the relative inability of many to conduct effective searches. As networks proliferate, the problem of ‘mapping’ between databases increases in complexity, and mention is made of attempts by the UK government to establish a system to support interdepartmental search and ‘open government’. The Semantic Web is seen as the next big development, but will require enormous effort. It is concluded that text retrieval is less often a separate facility as it is increasingly embedded, as a required tool, in larger systems.     

基于领域本体的生物医学语义检索机制研究——以GoPubMed和SEGoPubMed为例

基于领域本体对大型文本语料库的语义检索研究成为近年来信息检索领域研究的热点。GoPubMed和SEGoPubMed作为该领域研究成果的杰出代表,为PubMed存在的问题提供了良好的解决方案。对GoPubMed和SEGoPubMed语义检索机制进行深入研究后显示,GoPubMed应用基于语义网的语义分类工具——_GO和MeSH,将PubMed检索结果统计并对应到相应的分类导航目录,使用户利用该导航便可以快速找到自己所需要的文献,但其仍是基于布尔逻辑模型的字符串匹配机制;而比GoPubMed性能更优越的SEGoPubMed应用潜在语义分析（LSA）技术开创了语义敏感地浏览、连接大型文本语料库与本体的新模式,将具有更广泛的应用前景。     

New media–new voices: satirical representations of Nigeria’s socio-politics in Ogas at the top

New media are increasingly providing spaces and opportunities for media houses and activist groups engaged in socio-political reform in Africa. In Nigeria, social media are becoming platforms for communicating messages of resistance against oppressive political and exploitative economic power structures. This study analyzed Ogas at the top (OATT), an online puppetry series by Buni TV, as a way of examining new platforms and message content in Nigeria’s rapidly changing media sphere. Relying on semiotics and critical discourse analysis perspectives, the study analyzed select episodes of the series, to gauge how producers constructed powerful visual and linguistic messages to boldly satirize social injustices perpetrated by Nigeria’s political elites.     

网络图像检索的提问式特征研究

为了解中文网络检索情境下图像检索需求表达方面的行为规律,设计用户图像搜索实验来采集网络图像检索过程中的提问式进行小规模实证研究,一方面获得图像检索提问式的构造和语言语法方面的一般特征;另一方面通过对高效图像检索过程中提问式的专门分析,揭示高效图像提问式的个性特征。最后结合研究结果讨论提出图像检索需求表达规律和图像检索策略。     

Exploring the applicability of the Semantic Web for discovering and navigating Australian Indigenous knowledge resources

ABSTRACT

Semantic Web ontology files can be flexibly programmed to delineate metadata relationships in machine-readable formats to create relational pathways for discovering resources both on and off the Internet. There is a global community of Semantic Web developers and users across a broad multi-disciplinary range of interests who create and share extensible open-source ontologies. In this article, the author will explore the functionality of Semantic Web techniques for representing the ontologies of relatedness through kinship that typically underpin Australian Indigenous knowledge systems, and investigate their potentials for meeting persistent demands among leading Australian Indigenous collections creators and users to be able to search and discover their hereditary knowledge resources in ways that reflect and reinforce their enduring cultural values, ways of knowing and rights-management concerns.     

基于潜在语义索引的特征选择与权重改进若干关键问题的研究与实现

分析潜在语义索引的基本原理及其特点。针对LSI的3个因素，从特征词的选取，维数约简，特征词权重3个方面进行约定和改进。并以计算机类的科技文献作为测试文档，对改进的权重算法和改进前后LSI系统的检索结果进行分析。结果显示，特征词的选择结果及检索效果都得到较大的提高，性能稳定。     

国内语义Web研究综述

语义Web并非一个独立的Web ,而是当前Web的延伸 ,其目标是让计算机能够“理解和处理”现在的Web仅仅能显示的数据 ,并为人们提供各种智能服务。目前在世界范围内对语义Web的研究尚处于起步阶段 ,与国外相比 ,我国对语义Web的研究相对较晚。本文试图对国内关于语义Web的研究做一简要的综述 ,在把握当前研究现状的基础上 ,明确今后主要的研究方向与重点问题。     

基于语义网技术构造知识服务系统

语义网的兴起,为知识服务系统的发展提供了巨大的动力。而知识管理的思想运用于语义网建设,又将为用户提供基于知识的Web服务。因此,语义网技术作为下一代Web的核心技术,知识服务作为信息服务的发展趋势,两者的相互结合将是可行而必要的,肯定会为知识服务的研究与发展开辟新的道路。本文阐述了语义网技术在知识服务中的角色,探讨了语义网与知识服务系统结合的应用前景,并论述了基于语义网技术构造知识服务系统的主要实现模式     

Multimodal biomedical image indexing and retrieval using descriptive text and global feature mapping

The images found within biomedical articles are sources of essential information useful for a variety of tasks. Due to the rapid growth of biomedical knowledge, image retrieval systems are increasingly becoming necessary tools for quickly accessing the most relevant images from the literature for a given information need. Unfortunately, article text can be a poor substitute for image content, limiting the effectiveness of existing text-based retrieval methods. Additionally, the use of visual similarity by content-based retrieval methods as the sole indicator of image relevance is problematic since the importance of an image can depend on its context rather than its appearance. For biomedical image retrieval, multimodal approaches are often desirable. We describe in this work a practical multimodal solution for indexing and retrieving the images contained in biomedical articles. Recognizing the importance of text in determining image relevance, our method combines a predominately text-based image representation with a limited amount of visual information, in the form of quantized content-based visual features, through a process called global feature mapping. The resulting multimodal image surrogates are easily indexed and searched using existing text-based retrieval systems. Our experimental results demonstrate that our multimodal strategy significantly improves upon the retrieval accuracy of existing approaches. In addition, unlike many retrieval methods that utilize content-based visual features, the response time of our approach is negligible, making it suitable for use with large collections.