The digital construction of cultural heritage promotes communication and sharing of digital cultural resources across time and space. Digital storage serves as the foundation for the digital construction of cultural artifacts. In the digital storage of Dunhuang murals, image stitching plays a critical role in restoring the complete image of the cave murals. Traditional image stitching methods are constrained by the detection accuracy of feature points and are not fit for stitching low-texture murals. Despite deep learning-based image stitching methods, parallax misalignment and ghosting are still prevalent issues. For this reason, we perform the first Dunhuang mural stitching based on deep learning in this paper. This is in response to the need for digitizing and storing Dunhuang murals. Two mural stitching datasets are constructed, and we design a progressive regression image alignment network and a feature differential reconstruction soft-coded seam stitching network. We also introduce a soft-coded seam quality evaluation method. The algorithm presented in this paper achieves state-of-the-art alignment and stitching performance in the mural stitching task through unsupervised learning with a smaller number of model parameters, which provides technical support for the digitization and preservation of Dunhuang murals.

Neural networks have shown great successes in many fields. Due to the complexity of the training pipeline, however, using them in an industrial setting is challenging. In online advertising, the complexity arises from the immense size of the training data, and the dimensionality of the sparse feature space (both can be hundreds of billions). To tackle these challenges, we built TrainSparse (TS), a system that parallelizes the training of neural networks with a focus on efficiently handling large-scale sparse features. In this paper, we present the design and implementation of TS, and show the effectiveness of the system by applying it to predict the ad conversion rate (pCVR), one of the key problems in online advertising. We also compare several methods for dimensionality reduction on sparse features in the pCVR task. Experiments on real-world industry data show that TS achieves outstanding performance and scalability.

Diverse data acquisition methods, such as airborne, Unmanned Aerial Vehicle (UAV), and space-borne sensor bring in large volumes of raster image data. For rapid decision making based on those multi-source raster data, an effective and efficient storage and management system is necessary. This paper mainly focuses on how to design and implement flexible interfaces for GeoRaster. The raster data and their metadata in our work are all stored in Oracle database. Meanwhile, the constructed interfaces can offer such functionalities like image blocking, compressing, pyramid construction, and spatial indexing. To the latter, the constructed system present two connecting mechanisms: Client/Server and Browser/Server. All the systems support data querying, data visualization, data acquisition as well as some other data management operations.

This paper focuses on analyzing and predicting not-answered questions in Community based Question Answering (CQA) services, such as Yahoo! Answers. In CQA services, users express their information needs by submitting natural language questions and await answers from other human users. Comparing to receiving results from web search engines using keyword queries, CQA users are likely to get more specific answers, because human answerers may catch the main point of the question. However, one of the key problems of this pattern is that sometimes no one helps to give answers, while web search engines hardly fail to response. In this paper, we analyze the not-answered questions and give a first try of predicting whether questions will receive answers. More specifically, we first analyze the questions of Yahoo Answers based on the features selected from different perspectives. Then, we formalize the prediction problem as supervised learning – binary classification problem and leverage the proposed features to make predictions. Extensive experiments are made on 76,251 questions collected from Yahoo! Answers. We analyze the specific characteristics of not-answered questions and try to suggest possible reasons for why a question is not likely to be answered. As for prediction, the experimental results show that classification based on the proposed features outperforms the simple word-based approach significantly.

Community discovery on large-scale linked document corpora has been a hot research topic for decades. There are two types of links. The first one, which we call d2d-link, indicates connectiveness among different documents, such as blog references and research paper citations. The other one, which we call u2u-link, represents co-occurrences or simultaneous participations of different users in one document and typically each document from u2u-link corpus has more than one user/author. Examples of u2u-link data covers email archives and research paper co-authorship networks. Community discovery in d2d-link data has achieved much success, while methods for that in u2u-link data either make no use of the textual content of the documents or make oversimplified assumptions about the users and the textual content. In this paper we propose a general approach of community discovery for u2u-link data, i.e., multiple user data, by placing topical variables on multiple authors' participations in documents. Experiments on a research proceeding co-authorship corpus and a New York Times news corpus show the effectiveness of our model.

This paper is concerned with community discovery in textual interaction graph, where the links between entities are indicated by textual documents. Specifically, we propose a Topical Link Model(TLM), which leverages Hierarchical Dirichlet Process(HDP) to introduce hidden topical variable of the links. Other than the use of links, TLM can look into the documents on the links in detail to recover sound communities. Moreover, TLM is a nonparametric model, which is able to learn the number of communities from the data. Extensive experiments on two real world corpora show TLM outperforms two state-of-the-art baseline models, which verify the effectiveness of TLM in determining the proper number of communities and generating sound communities.

This paper is concerned with the study of information retrieval (IR) on Accumulative Social Descriptions (ASDs). ASDs refer to Web texts that accumulated by many Web users describing certain Web resources, such as anchor texts, search logs and social annotations. There have been some studies working on leveraging ASDs for improving search performance in both internet and intranet. However, to the best of our knowledge, no prior study has concerned the specific generation features of ASDs, which are the focus point of this paper. Specifically, we consider the generation features from two perspectives, the generation processes and the generated distributions. Further, three probabilistic IR models are derived based on them. The three models are first demonstrated with one toy dataset and then empirically evaluated with two real datasets: an internet dataset consisting of 90,295 Web pages, with 25,845,818 social annotations crawled from Del.icio.us and 31,320,005 pieces of anchor texts crawled through Yahoo! API, and an intranet dataset consisting of 179,835 Web pages with 1,245,522 annotations dumped from the intranet tagging system in IBM, named as Dogear. Extensive experimental results show that the proposed methods, which fully leverage the generation features of ASDs, improve the performance of both internet and intranet search significantly.

Web document could be seen to be composed of textual content as well as social metadata of various forms (e.g., anchor text, search query and social annotation), both of which are valuable to indicate the semantic content of the document. However, due to the free nature of the web, the two streams of web data suffer from the serious problems of noise and sparseness, which have actually become the major challenges to the success of many web mining applications. Previous work has shown that it could enhance the content of web document by integrating anchor text and search query. In this paper, we study the problem of exploring emergent social annotation for document enhancement and propose a novel reinforcement framework to generate "social representation" of document. Distinguishing from prior work, textual content and social annotation are enhanced simultaneously in our framework, which is achieved by exploiting a kind of mutual reinforcement relationship behind them. Two convergent models, social content model and social annotation model, are symmetrically derived from the framework to represent enhanced textual content and enhanced social annotation respectively. The enhanced document is referred to as Social Document or sDoc in that it could embed complementary viewpoints from many web authors and many web visitors. In this sense, the document semantics is enhanced exactly by exploring social wisdom. We build the framework on a large Del.icio.us data and evaluate it through three typical web mining applications: annotation, classification and retrieval. Experimental results demonstrate that social representation of web document could boost the performance of these applications significantly.