Search

Because of the increasing variety of user terminals, there is a great demand for the Web pages adapt to various resolution terminals with good readability. This paper describes an effective approach to display a Web page in various resolution terminals with a realistic layout. According to this approach, multiple-level navigation views are automatically generated based on the layout information extracted from the original Web pages. The navigation views can help the users using low resolution terminals to see the brief contents of the large Web pages with high resolution. A case study is introduced in this paper to display how to apply this approach, and show that it is an effective solution for low-resolution terminals to display Web pages that is designed for high-resolution terminals. [ABSTRACT FROM AUTHOR]

Document clustering is necessary for information retrieval, Web data mining, and Web data management. To support very high dimensionality and the sparsity of document feature, the model-based clustering has been proved to be an intuitive choice for document clustering. However, the current model-based algorithms are prone to generating the skewed clusters, which influence the quality of clustering seriously. In this paper, the reasons of skew generating are examined and determined as the inappropriate initial model, and the interaction between the decentralization of estimation samples and the over-generalized cluster model. An effective clustering skew prevention method (ESPClust) is proposed to focus on the last reason. To break this interaction, for each cluster, ESPClust automatically selects a part of documents that most relevant to its corresponding class as the estimation samples to re-estimate the cluster model. Based on the ESPClust, two algorithms with respect to the quality and efficiency are provided for different kinds of applications. Compared with balanced model-based algorithms, the ESPClust method has less restrictions and more applicability. The experiments show that the ESPClust can avoid the clustering skew in a great degree and its Macro-F1 measure outperforms the previous methods' measure. [ABSTRACT FROM AUTHOR]

To evaluate document-to-document relevance is very important to many advanced applications such as IR, text mining and natural language processing. Since it is very hard to define document relevance in a mathematic way on account of users' uncertainty, the concept of topical relevance is widely accepted by most of research fields. It suggests that a document relevance model should explain whether the document representation describes its topical contents and the matching method reveals the topical differences among the documents. However, the current document-to-document relevance models, such as vector space model, string distance, don't put explicitly emphasis on the perspective of topical relevance. This paper exploits a document language model to represent the document topical content and explains why it can reveal the document topics and then establishes two distributional similarity measure based on the document language model to evaluate document-to-document relevance. The experiment on the TREC testing collection is made to compare it with the vector space model, and the results show that the Kullback-Leibler divergence measure with Jelinek-Mercer smoothing outperforms the vector space model significantly. [ABSTRACT FROM AUTHOR]

Due to the high-dimension and sparseness properties of documents, clustering the similar documents together is a tough task. The most popular document clustering method K-Means has the shortcoming of its cluster intra-dissimilarity, i.e. inclining to clustering unrelated documents together. One of the reasons is that all objects (documents) in a cluster produce the same influence to the mean of the cluster. SOM (Self Organizing Map) is a method to reduce the dimension of data and display the data in low dimension space, and it has been applied successfully to clustering of high-dimensional objects. The scalar factor is an important part of SOM. In this paper, an optimized K-Means algorithm is proposed. It introduces the scalar factor from SOM into means during K-Means assignment stage for controlling the influence to the means from new objects. Experiments show that the optimized K-Means algorithm has more F-Measure and less Entropy of clustering than standard K-Means algorithm, thereby reduces the intra-dissimilarity of clusters effectively. [ABSTRACT FROM AUTHOR]

IR-style keyword-based search on XML document has become the most common tool for XML query, as users need not to know the structural information of the target XML document before constructing a query. For a keyword-based search engine for XML document, the key issue is how to return some sets of meaningfully related nodes to user's query efficiently. An ordinary solution of current approaches is to store the relationship of each pair of nodes in an XML document to an index. Obviously, this will lead to serious storage overhead. In this paper, we propose an enhanced inverted index structure (PN-Inverted Index) that stores path information in addition to node ID, and import and extend the concept of LCA to PLCA. Efficient algorithms with these concepts are designed to check the relationship of arbitrary number of nodes. Compared with existing approaches, our approach need not create additional relationship index but just utilize the existing inverted index that is much common for IR-style keyword search engine. Experimental results show that with the promise of returning meaningful answers, our search engine offers great performance benefits. Although the size of the inverted index is increased, the total size of indices of search engine is smaller than the existing approaches. [ABSTRACT FROM AUTHOR]