Voice browsing

The Web provides universal access to information of staggering quantity and diversity. The universality results from the choice of a single means for representing information (HTML) and a single means for delivering it (HTTP). The end-user value produced by this choice is unrivaled in the history of data communications. As great as it is today, the Web's value continues to increase as the number of suppliers and consumers of information increases (Metcalfe's law). No new technology is needed to sustain this growth, except at the transport level, where bandwidth must increase to meet our user demand. What important limitations does the Web have that new technology could remove? One limitation involves access. While the Web (thanks to the geographical reach of the Internet beneath it) has many points of access, it supports just one type of access: visual, from a PC. As our reliance on the Web increases, so does our need to access it. Especially valuable and natural, would be voice access from a phone, especially a mobile phone. The next five years (2000-2005) will see the widespread deployment of voice browsers: digital personae who converse with their users by phone, retrieve Web information (for example, flight schedules) at their user's request, and carry out transactions (for example, changing a travel itinerary) according to their user's instructions. While the most familiar visual browsers today are lightweight PC applications, voice browsers will run in 24/7 network operations centers with the required telephony hardware and speech recognition software     

Contextual multi-dimensional browsing

Browsing efficiency depends to a large extent on the organization of the information being browsed. An emerging method of organizing Web content, called multi-faceted categorization, adopts a flat structure and treats different facets of the browsed content as equal and independent. The advantage of this structure is that it allows users to slice and dice the information space from any of the facets they wish to browse, and in any arbitrary order, thereby facilitating so-called multi-dimensional browsing. However, because of limited human information processing capacities, too many choices and too much browsing freedom tend to disorient users and increase the difficulty of information gathering. A more effective means of organizing information is therefore needed.In the literature, it is well documented that context plays a crucial role in making relevance judgments about categorical attributes. Since browsing always occurs in relation to certain contexts, we propose taking context into account for multi-faceted categorization to help organize different facets of information on the browsing interface. Our experiment results demonstrate that the context-sensitive arrangement of categories helps users find information more efficiently. Moreover, users perceive the context-sensitive interface as easier to use than non-contextual interfaces.     

Enhanced browsing in digital libraries: three new approaches to browsing in Greenstone

Browsing is part of the information-seeking process, used when information needs are ill-defined or unspecific. Browsing and searching are often interleaved during information seeking to accommodate changing awareness of information needs. Digital libraries often do not support browsing. Described here are three browsing systems created for the Greenstone digital library software .     

Influence-based Twitter browsing with NavigTweet

Directed links in social media determine the flow of information and, hence, indicate a user's influence. This paper proposes a novel visual framework to explore Twitter's ‘Who Follows Who’ relationships, by browsing the friends’ network to identify key influencers based on the actual influence of the content they share. We have developed NavigTweet, a visualization tool for the influence-based exploration of Twitter network. NavigTweet embeds a force-directed algorithm to display the graph in a multi-clustered way. To assess the user experience with NavigTweet, we have conducted a pre-release qualitative pilot study. We also report on the study and results of post-release user feedback survey.     

Taking control of web browsing

Adherence to accessibility guidelines for Web pages does not necessarily guarantee a usable or satisfying Web experience for all persons with disabilities. The needs of many of these individuals fall outside the guidelines for accessible content that most Web authors take into consideration. Many of these users wish, for example, that they simply could ‘enlarge’ what is on a Web page. They also express the wish that pages would be ‘less confusing’. To meet these needs, Web browsers and various software applications provide for a variety of ways in which page presentations can be altered. The effects of these alterations often have unexpected consequences. Some designs accommodate these alterations better than others. This article discusses one such application that allows users to control features of Web page presentation and explores design features that facilitate such control.     

Loosely-coupled approach towards multi-modal browsing

Contemplating the concept of universal-access multi-modal browsing comes as one of the emerging killer technologies that promises broader and more flexible access to information, faster task completion, and advanced user experience. Inheriting the best from GUI and speech, based on the circumstances, hardware capabilities, and environment, multi-modalitys great advantage is to provide application developers with a scalable blend of input and output channels that may accommodate any user, device, and platform. This article describes a flexible multi-modal browser architecture, named Ferda the Ant, which reuses uni-modal browser technologies available for VoiceXML, WML, and HTML browsing. A central component, the Virtual Proxy, acts as a synchronization coordinator. This browser architecture can be implemented in either a single client configuration, or by distributing the browser components across the network. We have defined and implemented a synchronization protocol to communicate the changes occurring in the context of a component browser to the other browsers participating in the multi-modal browser framework. Browser wrappers implement the required synchronization protocol functionality at each of the component browsers. The component browsers comply with existing content authoring standards, and we have designed a set of markup-level authoring conventions that facilitate maintaining the browser synchronization .     

Tactile web browsing for blind people

Information on the World Wide Web becomes more and more important for our society. For blind people this is a chance to access more information for their everyday life. In this paper we propose novel methods to present web pages including graphical information on a tactile output device. We present a Mozilla Firefox Extension for the tactile rendering of web pages and for the handling of user interactions. This approach benefits from the Firefox built-in web page handling including parsing of HTML documents, formatting with cascading style sheets (CSS), handling of dynamic web content controlled by JavaScript code, etc. Graphical information can be explored and filtered interactively in a special mode for raster images and Scalable Vector Graphics (SVG). Mathematical expressions encoded in the mathematical markup language (MathML) are transformed directly into LaTeX or into a notation for blind people. The tactile web browser supports feedback that is provided via voice output.

On interactive browsing of large images

A new effective mechanism is proposed for the browsing of large compressed images over the Internet. The image is compressed with the JPEG 2000 into one single bitstream and put on the server. During the browsing process, the user specifies a region of interest (ROI) with certain spatial and resolution constraint. The browser only downloads the portion of the compressed bitstream that covers the current ROI, and the download is performed in a progressive fashion so that a coarse view of the ROI can be rendered very quickly and then gradually refined as more and more bitstream arrives. In the case of the switch of ROI, e.g., zooming in/out or panning around, the browser uses existing compressed bitstream in cache to quickly render a coarse view of the new ROI, and in the same time, request a new set of compressed bitstream corresponding to the updated view. The system greatly improves the experience of browsing large images over the slow networks.     

Shot clustering techniques for story browsing

Automatic video segmentation is the first and necessary step for organizing a long video file into several smaller units. The smallest basic unit is a shot. Relevant shots are typically grouped into a high-level unit called a scene. Each scene is part of a story. Browsing these scenes unfolds the entire story of a film, enabling users to locate their desired video segments quickly and efficiently. Existing scene definitions are rather broad, making it difficult to compare the performance of existing techniques and to develop a better one. This paper introduces a stricter scene definition for narrative films and presents ShotWeave, a novel technique for clustering relevant shots into a scene using the stricter definition. The crux of ShotWeave is its feature extraction and comparison. Visual features are extracted from selected regions of representative frames of shots. These regions capture essential information needed to maintain viewers' thought in the presence of shot breaks. The new feature comparison is developed based on common continuity-editing techniques used in film making. Experiments were performed on full-length films with a wide range of camera motions and a complex composition of shots. The experimental results show that ShotWeave outperforms two recent techniques utilizing global visual features in terms of segmentation accuracy and time.     

Adaptively browsing image databases with PIBE

Browsing large image collections is a complex and often tedious task, due to the semantic gap existing between the user subjective notion of similarity and the one according to which a browsing system organizes the images. In this paper we propose PIBE, an adaptive image browsing system, which provides users with a hierarchical view of images (the Browsing Tree) that can be customized according to user preferences. A key feature of PIBE is that it maintains local similarity criteria for each portion of the Browsing Tree. This makes it possible both to avoid costly global reorganization upon execution of user actions and, combined with a persistent storage of the Browsing Tree, to efficiently support multiple browsing tasks. We present the basic principles of PIBE and report experimental results showing the effectiveness of its browsing and personalization functionalities.     

Distance browsing in distributed multimedia databases

The state of the art of searching for non-text data (e.g., images) is to use extracted metadata annotations or text, which might be available as a related information. However, supporting real content-based audiovisual search, based on similarity search on features, is significantly more expensive than searching for text. Moreover, such search exhibits linear scalability with respect to the dataset size, so parallel query execution is needed.In this paper, we present a Distributed Incremental Nearest Neighbor algorithm (DINN) for finding closest objects in an incremental fashion over data distributed among computer nodes, each able to perform its local Incremental Nearest Neighbor (local-INN) algorithm. We prove that our algorithm is optimum with respect to both the number of involved nodes and the number of local-INN invocations. An implementation of our DINN algorithm, on a real P2P system called MCAN, was used for conducting an extensive experimental evaluation on a real-life dataset.The proposed algorithm is being used in two running projects: SAPIR and NeP4B.     

Identifying Web browsing trends and patterns

Since it entered the popular culture in 1994, the World Wide Web has grown from approximately two million servers to more than 110 million in 2001, according to the Internet Software Consortium. Jupiter Media Metrix, an Internet research company, estimates that during this same period, the number of US home Web users has likewise increased from 3 million to more than 89 million. Has this phenomenal growth resulted in fundamental changes in the way users browse the Web? To address this question, we analyzed a sample of more than 20,000 Internet users who accessed the Web from July 1997 through December 1999. This nationally representative data set collected by Jupiter Media Metrix is unique because it was gathered from users' computers, not servers, thus it is immune to caching problems present in other Web usage studies     

Individual differences in hypertext browsing strategies

The purpose of this study was to investigate differences in web browsing strategies between older and younger participants and also between individuals displaying verbaliser and imager cognitive styles. Typically, verbalisers process information in word form, whereas imagers process information in terms of images. Ten minutes each were allowed to 58 participants to read information in either a hierarchical or relational hypertext with the expectation of answering questions on this information. Browsing strategies were assessed firstly by using three indices of browsing patterns (number of pages visited, proportion of pages visited and pages revisited) and secondly by analysing assessments of browsing strategies given by independent judges. Differences were found between imagers and verbalisers for the indices of browsing patterns, with verbalisers visiting more pages in the hierarchical architecture and imagers visiting more pages in the relational architecture. Similarly, differences were noted between younger and older participants, with the younger groups visiting more pages in the hierarchical condition and the older groups visiting more pages in the relational condition. No differences were observed for the judge's ratings of browsing strategy. The findings have possible implications for the design of web-based instructional systems.     

Individual differences in hypertext browsing strategies

The purpose of this study was to investigate differences in web browsing strategies between older and younger participants and also between individuals displaying verbaliser and imager cognitive styles. Typically, verbalisers process information in word form, whereas imagers process information in terms of images. Ten minutes each were allowed to 58 participants to read information in either a hierarchical or relational hypertext with the expectation of answering questions on this information. Browsing strategies were assessed firstly by using three indices of browsing patterns (number of pages visited, proportion of pages visited and pages revisited) and secondly by analysing assessments of browsing strategies given by independent judges. Differences were found between imagers and verbalisers for the indices of browsing patterns, with verbalisers visiting more pages in the hierarchical architecture and imagers visiting more pages in the relational architecture. Similarly, differences were noted between younger and older participants, with the younger groups visiting more pages in the hierarchical condition and the older groups visiting more pages in the relational condition. No differences were observed for the judge's ratings of browsing strategy. The findings have possible implications for the design of web-based instructional systems.     

Spatial metaphors and disorientation in hypertext browsing

Abstract

The spatial metaphor can be used as a framework for explaining and designing tools that alleviate disorientation problems in hypertext systems. The approach based on this metaphor would involve developing tools analogous to navigational aids in physical environments and applying analogous concepts from research on human spatial processing and navigation in physical spaces. Research on hypertext browsing with respect to the spatial metaphor is reviewed and contrasted with the larger task context where users are trying to explore, learn, analyse, and summarize the contents of the hypertext space.     

Video parsing and browsing using compressed data

Parsing video content is an important first step in the video indexing process. This paper presents algorithms to automate the video parsing task, including partitioning a source video into clips and classifying those clips according to camera operations, using compressed video data. We have developed two algorithms and a hybrid approach to partitioning video data compressed according to the JPEG and MPEG standards. The algorithms utilize both the video content encoded in DCT (Discrete Cosine Transform) coefficients and the motion vectors between frames. The hybrid approach integrates the two algorithms and incorporates multi-pass strategies and motion analyses to improve both accuracy and processing speed. Also, we present content-based video browsing tools which utilize the information, particularly about the shot boundaries and key frames, obtained from parsing.     

一种高效的用户浏览行为采集方法

针对在多用户、多Web站点的网络访问环境下存在的数据采集问题,提出了一种新的数据采集及清理方法。引入网络嗅探的方式进行浏览行为数据的采集;在结合传统数据清理方法的基础上,提出一种利用HTTP请求间存在的引用和时间关系来清除HTTP请求产生的大量的非用户显式点击的附带请求。实验结果表明,该方法可以有效地采集浏览行为数据并清除大量的附带请求,抽取出能够反映用户显式点击的页面基请求,为用户浏览行为建模提供准确的数据源。     

基于用户浏览日志的上下文相关新闻推荐

个性化新闻推荐的难点在于用户在浏览新闻时目的性不强,容易受各种环境因素的影响,导致其浏览行为难以预测。以往的研究仅仅强调推荐内容相关的或者和用户长期兴趣相匹配的新闻,忽视了环境因素的影响。为此,需要研究上下文相关的新闻推荐算法。具体做法是从用户的浏览日志中提取上下文相关特征,然后训练一个Logistic回归模型来预测用户接下来最可能阅读的新闻。真实数据上的实验结果表明,上下文相关新闻推荐方法效果明显好于传统方法,也验证了上下文信息对用户浏览行为的重要影响。