Since its inception at Zhejiang University in 2005, the International Conference on the Universal Digital Library (ICUDL) has been held around the globe in Alexandria, Egypt, Carnegie Mellon University, USA, and at Allahabad in India. This annual event has been a strong driving force for exchange in digital library technologies, international scientific and cultural cooperation, as well as an influence on the development of the Universal Digital Library. Now the sixth ICUDL is being held in Hangzhou (China) again, which is bound to become a good opportunity for us to summarize the past while casting our eyes into the future. The theme of this year’s conference is ‘Data Ocean and Cloud Computing’.

We present our approach for constructing a federated searching system for Japanese humanities digital libraries using automatic metadata mapping. We discuss some achievements in the ongoing research to construct a federated searching system. The goals of this system are (1) to perform metadata mapping automatically for Japanese heterogeneous humanities digital libraries and (2) to let users access multiple humanities digital libraries by using only one query input. The proposed automatic metadata mapping method produced the average mapping precision of 94.9%. We also address the metadata-related challenges facing Japanese humanities databases.

One of the main roles I have played as a director of the Universal Digital Library has been to write grant proposals to support our work. Both for this project and for another project, Olive.org, an archive of executable content, how to sustain the final product is the most difficult challenge. This paper discusses the various models that might be adopted to sustain a large corpus of digital material, such as that of the Million Book Project. Methods discussed here include government funding, foundations and nonprofits, university homes, and joining existing projects. All individuals working with large digital projects should be concerned about how their work will be kept available to the public.

Good planning and communication are crucial to the digitization process. The originators of project proposals and donors who finance the work are often not technically sophisticated. They may have vastly different expectations of the complexity of the work, what are technically feasible and realistic time frames to achieve acceptable results than does the technical staff that carries out the work. In addition, managers may compete for resources (personnel and scheduling priority) for their projects. This can lead to stress, misunderstandings, and lowered technical staff morale. At Carnegie Mellon University Libraries, we developed a tactical planning system that mitigated these problems, improved interpersonal relationships and expedited the digitization process.

We present a novel approach for extracting noun phrases in general and named entities in particular from a digital repository of text documents. The problem of coreference resolution has been divided into two subproblems: pronoun resolution and non-pronominal resolution. A rule based-technique was used for pronoun resolution while a learning approach for non-pronominal resolution. For named entity resolution, disambiguation arises mainly due to polysemy and synonymy. The proposed approach fixes both problems with the help of WordNet and the Word Sense Disambiguation tool. The proposed approach, to our knowledge, outperforms several baseline techniques with a higher balanced F-measure, which is harmonic mean of recall and precision. The improvements in the system performance are due to the filtering of antecedents for the anaphor based on several linguistic disagreements, use of a hybrid approach, and increment in the feature vector to include more linguistic details in the learning technique.

The increasing amount of videos on the Internet and digital libraries highlights the necessity and importance of interactive video services such as automatically associating additional materials (e.g., advertising logos and relevant selling information) with the video content so as to enrich the viewing experience. Toward this end, this paper presents a novel approach for user-targeted video content association (VCA). In this approach, the salient objects are extracted automatically from the video stream using complementary saliency maps. According to these salient objects, the VCA system can push the related logo images to the users. Since the salient objects often correspond to important video content, the associated images can be considered as content-related. Our VCA system also allows users to associate images to the preferred video content through simple interactions by the mouse and an infrared pen. Moreover, by learning the preference of each user through collecting feedbacks on the pulled or pushed images, the VCA system can provide user-targeted services. Experimental results show that our approach can effectively and efficiently extract the salient objects. Moreover, subjective evaluations show that our system can provide content-related and user-targeted VCA services in a less intrusive way.

For automatic object detection tasks, large amounts of training images are usually labeled to achieve more reliable training of the object classifiers; this is cost-expensive since it requires hiring professionals to label large-scale training images. When a large number of object classes come into view, the issue of obtaining a large enough amount of the labeled training images becomes more critical. There are three potential solutions to reduce the burden for image labeling: (1) allowing people to provide the object labels loosely at the image level rather than at the object level (e.g., loosely-tagged images without identifying the exact object locations in the images); (2) harnessing large-scale collaboratively-tagged images that are available on the Internet; and, (3) developing new machine learning algorithms that can directly leverage large-scale collaboratively- or loosely-tagged images for achieving more effective training of a large number of object classifiers. Based on these observations, a multi-task multi-label multiple instance learning (MTML-MIL) algorithm is developed in this paper by leveraging both inter-object correlations and large-scale loosely-labeled images for object classifier training. By seamlessly integrating multi-task learning, multi-label learning, and multiple instance learning, our MTML-MIL algorithm can achieve more accurate training of a large number of inter-related object classifiers (where an object network is constructed for determining the inter-related learning tasks directly in the feature space rather than in the label space). Our experimental results have shown that our MTML-MIL algorithm can achieve higher detection accuracy rates for automatic object detection.

It is now widely recognized that appropriate measures are required for digital preservation to ensure that digital data can be accessed and used currently and in the future. Among all the risks of digital preservation, format obsolescence is one of the most important. There have been several projects or initiatives dealing with the measurement method of format obsolescence risk, but there has been no mechanism to quantify the preservation risk or durability of digital formats based on a self-improving assessment model, executed with the aid of computers. This paper deals with a methodology for measuring the preservation durability of digital formats, especially for their risk assessment. This method is based on a quantitative assessment model for format risk, and can shift the non-quantifiable knowledge or experiences of field experts to a machine identifiable and processible form, or ‘risk scores’. Results can be recognized and communicated by computers automatically and formally, which can assist in the automatic/semi-automatic risk management for digital preservation, sharing this quantified knowledge among communities. Because technologies are changing quickly, the quantitative assessment model for risks will not be a status quo situation. Thus, also presented is a method to fine tune the quantitative assessment model for risk of formats through a self-learning and self-improving style.

Organizing unstructured information from books into a well-defined structure is a significant challenge in digital libraries. Most digital libraries can provide only search services at the granularity of books and few libraries allow books to be accessed at the granularity of chapters, as manually constructing directory information for books is time-consuming. Extracting structured data from scanned books thus remains an urgent and important work. In this paper, we propose a novel structured data organization framework called CMSOF to organize scanned data automatically, and apply it to a Chinese medicine digital library. In the framework, image blocks and text blocks on the scanned page of books are separated based on the gray histogram projection method or a hybrid method of region growth and the Ada-Boosting classifier at first, and then the text structure is obtained from text blocks by text size and font type recognition. Finally, image blocks and structured OCRed text are correlated at the semantic level. By integrating the structured data into a Chinese medicine information system (CMIS), we can organize the Chinese medicine books well and users can access the books with flexibility, which indicates that CMSOF is an efficient framework to organize books mixed with images and text.

We describe a user interface and interaction technique, named ‘Javelin’, designed for large display environments. It provides quick access to random screen regions and manipulation methods for screen widgets which are difficult or impossible to reach. It consists of a dynamic global thumbnail, a touchpad widget that drives the screen cursor, and a teleport widget in which interactions are transferred to its target screen region. Javelin can be easily integrated into many programs to optimize their interaction performance in large screens. The experiment and user study show that Javelin can extend user access field and enhance widget manipulation in large displays.

Recently, we designed a new experimental system MSearch, which is a cross-media meta-search system built on the database of the WikipediaMM task of ImageCLEF 2008. For a meta-search engine, the kernel problem is how to merge the results from multiple member search engines and provide a more effective rank list. This paper deals with a novel fusion model employing supervised learning. Our fusion model employs ranking SVM in training the fusion weight for each member search engine. We assume the fusion weight of each member search engine as a feature of a result document returned by the meta-search engine. For a returned result document, we first build a feature vector to represent the document, and set the value of each feature as the document’s score returned by the corresponding member search engine. Then we construct a training set from the documents returned from the meta-search engine to learn the fusion parameter. Finally, we use the linear fusion model based on the overlap set to merge the results set. Experimental results show that our approach significantly improves the performance of the cross-media meta-search (MSearch) and outperforms many of the existing fusion methods.

Component-based software reuse and development is considered to be the best way to improve the efficiency and quality of development. To meet the demands of aircraft instrument exhibition in an aviation virtual library, we present a modeling and simulation tool for cockpit instruments on component methods, achieving display functionality and providing a visual interaction platform in the aviation virtual library. For the rapid model construction and reuse of virtual library resources, we classify entities in the cockpit into several categories by their behavior, and design a template for each category to describe its attributes. This efficiently reduces the modeling cycle, and thus largely helps users to take advantage of the library resource.

The proliferation of digital materials has changed not only how information is presented but also how people expect information to be available. People want access to all forms of information, from simple text to complex multimedia. Whether or not the items in question were created digitally, they can be made to behave digitally through scanning and conversion. This improves access, but makes preservation more difficult because of the rapid rate of obsolescence of formats, hardware and software systems. In the early days of the digital age, the gap between librarians and the people working in information technology was vast. The past 20 years, however, have seen thousands of digitization projects that include scanning paper analog resources and making them available in digital format. Such work has coupled the worlds of tech and libraries. This cross pollination has resulted in rich, robust online resources worthy of preservation. Libraries have a role in preservation of such resources, much as they have had a role in preservation of the hard-copy word. The Carnegie Mellon University Libraries (CMULs) have been at the forefront where digitization of rare and primary sources is concerned. Our pioneering efforts in digitization have also involved a creation of preservation strategies for digital content.