News is an essential resource for understanding politics, economics, local communities, etc. Readers of news articles might browse the text quickly or read thoroughly. In this study, we analyze the contents of news articles and propose a method for encouraging the readers to read the text efficiently. We construct classifiers to infer the reading time required to complete reading the article and localness indicating the region coverage of its content. We utilize the distributed representation of words appearing in news articles, and apply the BERT (Bidirectional Encoder Representations from Transformers) to construct our classifiers. The classification results could be used to support users’ decisions on when and where to read the articles. To learn these classifiers, a dataset is automatically generated from historical access logs of a news site. We carried out experiments to validate the proposed method. The experimental results demonstrates that our classifiers are superior to conventional methods.

Query autocompletion (QAC) is an important interactive feature that assists users in formulating queries and saving keystrokes. Due to the convenience it brings to users, QAC has been adopted in many applications, including Web search engines, integrated development environments (IDEs), and mobile devices. For existing QAC methods, users have to manually type delimiters to separate keywords in their inputs. In this paper, we propose a novel QAC paradigm through which users may abbreviate keywords by prefixes and do not have to explicitly separate them. Such paradigm is useful for applications where it is inconvenient to specify delimiters, such as desktop search, text editors, and input method editors. E.g., in an IDE, users may input getnev and we suggest GetNextValue. We show that the query processing method for traditional QAC, which utilizes a trie index, is inefficient under the new problem setting. A novel indexing and query processing scheme is hence proposed to efficiently complete queries. To suggest meaningful results, we devise a ranking method based on a Gaussian mixture model, taking into consideration the way in which users abbreviate keywords, as opposed to the traditional ranking method that merely considers popularity. Efficient top-k query processing techniques are developed on top of the new index structure. Experiments demonstrate the effectiveness of the new QAC paradigm and the efficiency of the proposed query processing method.

The problem of extracting meaningful data through graph analysis spans a range of different fields, such as the internet, social networks, biological networks, and many others. The importance of being able to effectively mine and learn from such data continues to grow as more and more structured data become available. In this paper, we present a simple and scalable semi-supervised learning method for graph-structured data in which only a very small portion of the training data are labeled. To sufficiently embed the graph knowledge, our method performs graph convolution from different views of the raw data. In particular, a dual graph convolutional neural network method is devised to jointly consider the two essential assumptions of semi-supervised learning: (1) local consistency and (2) global consistency. Accordingly, two convolutional neural networks are devised to embed the local-consistency-based and global-consistency-based knowledge, respectively. Given the different data transformations from the two networks, we then introduce an unsupervised temporal loss function for the ensemble. In experiments using both unsupervised and supervised loss functions, our method outperforms state-of-the-art techniques on different datasets.

In contrast to conventional studies of discovering hot spots, by analyzing geo-tagged images on Flickr, we introduce novel methods to discover obscure sightseeing spots that are less well-known while still worth visiting. To this end, we face two new challenges that the classical authority analysis based methods do not encounter: how to discover and rank spots on the basis of 1) popularity (obscurity level) and 2) scenery quality. For the first challenge, we estimate the obscurity level of a spot in accordance with the visiting asymmetry between photographers who are familiar with a target city and those who are not. For the second challenge, the behavior of both viewers who browsed the images and photographers are analyzed per each spot. We also develop an application system to help users to explore sightseeing spots with different geographical granularities. Experimental evaluations and analysis on a real dataset well demonstrate the effectiveness of the proposed methods.

In this paper, we propose and compare three ways of modeling photographers' location familiarity: a social network driven model, a time driven model and a location driven model. Then, the integration of the three models is further discussed. Experimental evaluations and analysis on a real data set consisting of 14,112 images collected from three cities well demonstrate the performance of the proposed classification methods. Many applications could benefit from information about the location familiarity, such as personalized geo-social recommendation, epidemic dispersion, urban computing, and so on.

While many governments have introduced financial schemes to encourage people to invest, it is difficult to understand investment trusts and decide which one to buy. To address that difficulty of understanding, a method for extracting causalities from monthly reports of investment trusts and visualizing them to support a potential investor's understanding of a trust is proposed. First, CRF is used to extract causalities from monthly reports. Note that features of financial reports other than linguistic features are also considered. Next, a causal network is constructed and visualized in consideration of the degrees of influence, frequency, and newness of the extracted causalities. The LOD control method is then applied to present causalities in consideration of the granularity of events appearing in the causal network. The results of a user evaluation demonstrate that proposed method performed better than a baseline method in terms of helping a user's understanding of investment trusts.

In this paper, we propose a novel stakeholder mining mechanism for analyzing bias in news articles by comparing descriptions of stakeholders. Our mechanism is based on the presumption that interests often induce bias of news agencies. As we use the term, a ``stakeholder'' is a participant in an event described in a news article who should have some relationships with other participants in the article. Our approach attempts to elucidate bias of articles from three aspects: stakeholders, interests of stakeholders, and the descriptive polarity of each stakeholder. Mining of stakeholders and their interests is achieved by analysis of sentence structure and the use of Relationship WordNet, a lexical resource that we developed. For analyzing polarities of stakeholder descriptions, we propose an opinion mining method based on the lexical resource Senti WordNet. We also describe an application system we developed for news comparison based on the mining mechanism. This paper presents a user study to validate the proposed methods.

From the viewpoint that most news items report on entities (person, organization and location), we propose a novel stakeholder model to represent and analyze news contents to explore special items in which there is inconsistency in the descriptions. By using this model, we can discover differences in multimedia news items from the perspectives of media types (text, video and audio) and description types (objective, subjective and relationship descriptions). We propose a method of extracting stakeholders as main participants (people, organization, etc.) of the described news event and detect inconsistency to explore the special items by comparing visual and textual descriptions on the exposure level of each stakeholder. A prototype system is implemented and we also show some experimental results to validate the proposed methods.

Protection of user privacy is an important issue of Web applications. Data encryption presents a possible resolution for improving the security level of Web applications. In this paper, we propose a novel mechanism using broadcast encryption and account assignment methods to reduce the decryption cost of Web applications. A notable feature of our mechanism is that the additional function of the application server is not necessary. Moreover, it is easy to apply this method to existing servers to improve their security level. We also show some experimental results to demonstrate the feasibility of our methods.

A search engine that can handle TV programs and Web content in an integrated way is proposed. Conventional search engines have been able to handle Web content and/or data stored in a PC desktop as target information. In the future, however, the target information is expected to be stored in various places such as in hard-disk (HD)/DVD recorders, digital cameras, mobile devices, and even in real space as ubiquitous content, and a search engine that can search across such heterogeneous resources will become essential. Therefore, as a first step towards developing such next-generation search engine, a prototype search system for Web and TV programs is developed that performs integrated search of those content, and that allows chain search where related content can be accessed from each search result. The integrated search is achieved by generating integrated indices for Web and TV content based on vector space model and by computing similarity between the query and all the content described by the indices. The chain search of related content is done by computing similarity between the selected result and all other content based on the integrated indices. Also, the zoom-based display of the search results enables to control media transition and level of details of the contents to acquire information efficiently. In this paper, testing of a prototype of the integrated search engine validated the approach taken by the proposed method.

We propose a new way of browsing bilingual Web sites through concurrent browsing with automatic similar-content synchronization and difference-detection facilities. Our prototype browser system is called the bilingual comparative Web browser (B-CWB) and it concurrently presents bilingual web pages in a way that enables their content of the Web pages to be automatically synchronized. The B-CWB allows users to browse two Web news sites concurrently and compare the similar news articles written in different languages (English and Japanese). The major characteristics of the B-CWB are its content synchronization and difference detection: content synchronization means that user operation (scrolling or clicking) on one Web page automatically invokes not necessarily same operations on the other Web page to preserve similarity of content between the two Web pages. For example, scrolling a Web page may involve passage-level similarity retrieval on the other Web page. Clicking a Web page (and obtaining a new Web page) invokes page-level similarity retrieval within the other Web site pages through the use of a English-Japanese dictionary. Difference detection means that the B-CWB analyzes two similar Web pages shown concurrently to discover the several "differences" between them. This facility is important in comparing two news articles that report the same affairs.

Dynamic content integration of multiple information sources is one way of providing richer content that will satisfy the diverse demands of users. In this paper, we propose an XML-based language to compose synchronized content from Web and video content. The notable features of this language are as follows: (1) dynamic unit identification of content that is composed into synchronized content; and (2) dynamic retrieval of content through pre-defined retrieval criteria. This dynamic identification and retrieval of composable units are based on the author's intentions. Content authors can specify the units of their content that are to be integrated into new content by describing the conditions concerning this content and the conditions concerning the surrounding content. Although the proposed language looks like SMIL (synchronized multimedia integration language), it differs in its dynamic identification and retrieval capabilities. Indeed, the proposed language works just like the meta-mechanism for conventional SMIL. That is, the script written by the proposed language can generate SMIL data as its output.