Web Track 2003技术报告

文本检索会议(Trec)是信息检索领域一年一度的学术交流和系统评测活动,本届Trec的WebTrack任务由命名网页发现/主页发现子任务和主题提取子任务组成。笔者在上届Trec的基础上,根据不同的子任务的需求,使用了不同的方法。在命名网页发现子任务中,锚文本、网页标题与网页内容是最重要的资源,而Ulr的目录层数对主页发现子任务中主页的判别有重要作用。多检索系统的投票机制,能大幅提高主题提取子任务的性能。     

一种基于多叉树的HTML到XML的转换方法

当前的Web信息大多数都是HTML格式的，由于HTML文件中没有严格的结构性，故很难能用一种有效的方法来检索或提取隐藏其中的数据．针对HTML的这种缺陷，本文提出了基于多叉树的HTML到XML转换方法，把对HTML的信息检索问题转化为对XML的检索问题，以便简化下一步的检索问题．     

一种基于多叉树的HTML到XML的转换方法

当前的Web信息大多数都是HTML格式的，由于HTML文件中没有严格的结构性，故很难能用一种有效的方法来检索或提取隐藏其中的数据，针对HTML的这种缺陷，本文提出了基于多叉树的HTML到XML转换方法，把对HTML的信息检索问题转化为对XML的检索问题，以便简化下一步的检索问题。     

Dynamic task allocation for multi-robot search and retrieval tasks

Many application domains require search and retrieval, which is also known in the robotic domain as foraging. For example, in a search and rescue domain, a disaster area needs to be explored and transportation of survivors to a safe area needs to be arranged. Performing such a search and retrieval task by more than one robot increases performance if they are able to distribute their workload efficiently and evenly. In this work, we study the Multi-Robot Task Allocation (MRTA) problem in the search and retrieval domain, where a team of robots is required to cooperatively search for targets of interest in an environment and also retrieve them back to a home base. In comparison with typical foraging tasks, we look at a more general search and retrieval task in which the targets are distinguished with various types, and task allocation also requires taking into account temporal constraints on the team goal. As usual, robots have no prior knowledge about the location of targets in the environment but in addition they need to deliver targets to the home base in a specific order according to their types, which significantly increases the complexity of a foraging problem. We first use a graph-based model to analyse the search and retrieval problem and the dynamics of exploration and retrieval within a cooperative team. We then proceed to present an extended auction-based approach, as well as a prediction approach. The essential difference between these two approaches is that the task allocation and execution procedures in the auction approach are running in parallel, whereas a robot in the prediction approach only needs to choose a task to perform when it has no thing to do. The auction approach uses a winner determination mechanism to allocate tasks to each robot, whereas the robots in the prediction approach implicitly coordinate their activities by team reasoning that leads to consensuses about task allocation. We use the Blocks World for Teams (BW4T) simulator to evaluate the two approaches in our experimental study.     

基于运动轨迹的视频检索方法

基于内容的视频检索是目前多媒体检索领域中的一个重要课题,由于视频体现了图像中对象的动态特征,因此如何提取视频对象的运动特征并且用于检索成为基于内容视频检索研究的热点.利用改进的8方向链码编码对视频对象运动轨迹进行编码,重点突出了对象运动的变化特征,同时采用归一化编码技术,增强了改进编码方法对尺度和旋转变换的鲁棒性.最后采用编辑距离来度量运动轨迹的相似性测度,实验结果表明,该算法能够有效地检索出运动轨迹相似的视频片断.     

基于视觉显著图的彩色图像检索

为提高图像检索效率,提出一种基于视觉显著图的彩色图像检索方法。利用视觉显著图去除原图中与检索任务无关的背景信息,保留用户感兴趣的图像区域信息,采用小波域的BDIP-BVLC方法提取图像特征,并引入二次查询度量策略进行距离度量。实验结果表明,与基于显著性加权的检索方法相比,该方法的平均查准率较高。     

Modelling of menu design in computerized work

The objective of the research was to propose and validate a theoretically meaningful link between three constructs of hierarchical menu design: menu dimension, task complexity, and user knowledge structure. Twenty-four subjects participated in a nested factorial experiment. The subjects performed a menu retrieval task using a hierarchical menu system constructed for use in the domain of utility boiler control. The dependent variables were time to respond and accuracy. The independent variables were menu dimension, task complexity and user knowledge structure. Four hypotheses were tested. The foundation of the hypotheses was based upon the premise that when task complexity is low, the short-term memory requirements of the menu retrieval task are low. Thus, the user's knowledge structure will not affect performance because it is not required for the chunking of visual information. The objectives of this research were met and are presented in the context of an information processing model for psychomotor tasks.     

Agents for case-based software reuse

The effective reuse of previously engineered components has become a core activity in any object-oriented software development project. The task is, however, often problematic when it comes to actual retrieval and understanding of the class library in, e.g., Java. Newcomers, especially, in a software company or novices in Java programming will need time to obtain a good overview of available components. This article explores the use of intelligent agents in a case-based tool for software reuse. By introducing agent support to the retrieval mechanism of the tool we show how retrieval efficiency may be improved. The cooperating agents assist the user in retrieval of code for potential reuse in an automated way and in the background. This makes it possible for the developer to concentrate fully on her task. The tool aids in program understanding and adaptation. Thus, it allows an exploratory approach to program development and increases reuse efficiency.     

P2P社区的信息检索算法设计

为解决P2P社区的资源定位及信息检索问题,采用混合型P2P网络模型,将社区内的检索划分为本地检索、组内搜索和组间搜索。对于本地检索设计了新的词条权重的计算方法,解决了同构文档集内的文本检索问题。对于组内搜索和组间搜索,通过设计节点选择策略,使一部分与查询相关度高的节点执行查询任务。最后提出结果融合的方法并对特定的实验数据进行测试,实验表明设计的算法在较小的查询开销下,能取得较好的检索效果。     

基于链接的方法进行Web信息检索的TREC实验研究

本文通过TREC实验研究基于链接信息的检索对Web信息检索的影响,包括使用链接描述文本,链接结构以及将基于链接的方法和传统基于内容检索的方法合并。得到如下结论:首先,链接描述文档对网页主题的概括有高度的精确性,但是对网页内容的描述有极大的不完全性;其次,与传统检索方法相比,使用链接文本在网页定位的任务上能够使系统性能提高96% ,但是在信息查询任务上没有帮助;最后,将基于链 接信息的检索与传统的基于内容检索技术合并,在网页入口定位任务上总能将系统性能提高48%到124.8% ,而对特定信息查询任务也能在一定程度上改善检索效果。     

基于SMO的多层次文本分类法研究

在以往的自动文本分类研究中,大多比较流行的分类技术都是在一个层次上将文本分成几个类别。但随着信息检索的量越来越大,文本的种类将越来越多,仅仅通过一层对海量信息进行组织分类越来越不适合海量信息的检索工作,这种平坦式的分类组织难以进一步提高信息检索的速度。论文将SMO分类算法结合到文本分类研究中,通过构建多层支持向量机文本分类树,实现了基于SMO的多层次文本分类系统。     

知识管理中基于本体的扩展检索方法

在知识管理系统中，为有效地解决用户查询与文档之间相同概念的不同表达形式造成的失配问题，提出一种基于本体、以面向任务情景的结构化描述作为信息体内容的语义索引的双向扩展检索方法，通过相容匹配和知识联网2种机制实现了扩展检索，分别对应于自上而下的和自下而上的2种途径；并采用查询重写模板（QRT）来搜索与当前任务相关的知识．基于原始查询和本体，QRT生成大量的子查询，同时将与原始查询相关度的权重传递给子查询式．自上而下方法或知识联网机制通过组织、任务本体检索到相关知识项．自下而上方法在任务情景中搜索相似任务，并获取包含该任务描述的知识项．2种方法都应用QRT实现基于本体的知识检索．实验结果表明：文中方法提高了知识管理系统的检索效率和准确率．     

Analysis of Minimum Distances in High-Dimensional Musical Spaces

We propose an automatic method for measuring content-based music similarity, enhancing the current generation of music search engines and recommended systems. Many previous approaches to track similarity require brute-force, pair-wise processing between all audio features in a database and therefore are not practical for large collections. However, in an Internet-connected world, where users have access to millions of musical tracks, efficiency is crucial. Our approach uses features extracted from unlabeled audio data and near-neigbor retrieval using a distance threshold, determined by analysis, to solve a range of retrieval tasks. The tasks require temporal features-analogous to the technique of shingling used for text retrieval. To measure similarity, we count pairs of audio shingles, between a query and target track, that are below a distance threshold. The distribution of between-shingle distances is different for each database; therefore, we present an analysis of the distribution of minimum distances between shingles and a method for estimating a distance threshold for optimal retrieval performance. The method is compatible with locality-sensitive hashing (LSH)-allowing implementation with retrieval times several orders of magnitude faster than those using exhaustive distance computations. We evaluate the performance of our proposed method on three contrasting music similarity tasks: retrieval of mis-attributed recordings (fingerprint), retrieval of the same work performed by different artists (cover songs), and retrieval of edited and sampled versions of a query track by remix artists (remixes). Our method achieves near-perfect performance in the first two tasks and 75% precision at 70% recall in the third task. Each task was performed on a test database comprising 4.5 million audio shingles.     

L 1-medial skeleton-based 3D point cloud model retrieval

Multimedia Tools and Applications - Facing the task of model-based retrieval, this paper proposes an L 1-medial skeleton-based 3D point cloud model retrieval method. First, L 1-medial skeleton is...     

Document ranking and the vector-space model

Efficient and effective text retrieval techniques are critical in managing the increasing amount of textual information available in electronic form. Yet text retrieval is a daunting task because it is difficult to extract the semantics of natural language texts. Many problems must be resolved before natural language processing techniques can be effectively applied to a large collection of texts. Most existing text retrieval techniques rely on indexing keywords. Unfortunately, keywords or index terms alone cannot adequately capture the document contents, resulting in poor retrieval performance. Yet keyword indexing is widely used in commercial systems because it is still the most viable way by far to process large amounts of text. Using several simplifications of the vector-space model for text retrieval queries, the authors seek the optimal balance between processing efficiency and retrieval effectiveness as expressed in relevant document rankings     

基于改进TextTiling方法的用户新兴趣发现的研究

个性化信息检索可以根据用户的检索兴趣返回个性化的检索结果.提出了用户新兴趣发现子任务,根据用户检索对象的变化识别包含新检索兴趣的查询.同时,引入TextTiling方法并对其进行改进,使系统可以自动选择合适的动态阈值并准确发现用户检索兴趣的转移.在构建的标准评测集上的实验结果表明,改进的TextTiling方法使得用户新兴趣发现系统性能提高了16.4%,而且此子任务使得最终的个性化检索系统的性能提高了3.8%.     

跨域图像检索综述

受成像载体、成像光谱和成像条件等的影响,跨域图像在不同领域的应用日益增多,跨域图像检索已成为了许多领域研究的热点和前言。然而图像的跨域检索面临着图像视觉偏差的问题,通过传统同域图像检索的方法无法有效地得到结果。通过文献调研,系统梳理了近年来跨域图像检索领域的代表性方法。对跨域图像检索任务作出了简要说明并指出了关键问题;根据图像域的不同转换阶段,将跨域图像检索方法分为两类：基于特征空间迁移和基于图像域迁移的跨域图像检索方法,并对两类方法进行了系统总结和分析;整理了跨域图像检索在不同领域的数据集,对比了各类方法的性能;总结了现有跨域检索方法并对未来的研究方向进行了展望。     

基于Jena规则推理数字图书馆信息检索系统研究

数字图书馆的核心任务之一就是提供良好的信息检索系统,而传统的信息检索技术以关键字匹配为主,缺乏语义推理能力,对用户的查询请求没有提供语义指导,因此造成信息的误检、漏检。将Jena用于数字图书馆信息检索,首先分析了数字图书馆的特点和需求,接着提出了基于Jena数字图书馆信息检索模型,深入研究了关键技术,最后对研究进行了验证。     

Designing an information search interface for younger and older adults

OBJECTIVE: The present study examined Web-based information retrieval as a function of age for two information organization schemes: hierarchical organization and one organized around tags or keywords. BACKGROUND: Older adults' performance in information retrieval tasks has traditionally been lower compared with younger adults'. The current study examined the degree to which information organization moderated age-related performance differences on an information retrieval task. The theory of fluid and crystallized intelligence may provide insight into different kinds of information architectures that may reduce age-related differences in computer-based information retrieval performance. METHOD: Fifty younger (18-23 years of age) and 50 older (55-76 years of age) participants browsed a Web site for answers to specific questions. Half of the participants browsed the hierarchically organized system (taxonomy), which maintained a one-to-one relationship between menu link and page, whereas the other half browsed the tag-based interface, with a many-to-one relationship between menu and page. This difference was expected to interact with age-related differences in fluid and crystallized intelligence. RESULTS: Age-related differences in information retrieval performance persisted; however, a tag-based retrieval interface reduced age-related differences, as compared with a taxonomical interface. CONCLUSION: Cognitive aging theory can lead to interface interventions that reduce age-related differences in performance with technology. In an information retrieval paradigm, older adults may be able to leverage their increased crystallized intelligence to offset fluid intelligence declines in a computer-based information search task. APPLICATION: More research is necessary, but the results suggest that information retrieval interfaces organized around keywords may reduce age-related differences in performance.