Sharif Digital Repository

Mining search engine query log is a new method for evaluating web site link structure and information architecture. In this paper we propose a new query-URL co-clustering for a web site useful to evaluate information architecture and link structure. Firstly, all queries and clicked URLs corresponding to particular web site are collected from a query log as bipartite graph, one side for queries and the other side for URLs. Then a new content free clustering is applied to cluster queries and URLs concurrently. Afterwards, based on information entropy, clusters of URLs and queries will be used for evaluating link structure and information architecture respectively. Data sets of different web... 

A Comprehensive speech database is one of the important tools for developing speech recognition systems; these tools are necessary for telephony recognition, too. Although adequate databases for direct speech recognizers exist, there is not an appropriate database for telephony speech recognizers. Most methods suggested for solving this problem are based on building new databases which tends to consume much time and many resources; or they used a filter which simulates circuit switch behavior to transform direct databases to telephony ones, in this case resulted databases have many differences with real telephony databases. In this paper we introduce a framework for creating telephony speech... 

Currently, many daily human activities such as education, commerce and talks are carried out through the Internet. In cases such as the registering in websites, some hackers write programs to make automatic false enrolments which waste the resources of the website while this may even stop the entire website from working. Therefore, it is necessary to tell apart human users from computer programs which is known as CAPTCHA (Completely Automated Public Turing test to tell Computers and Human Apart). CAPTCHA methods are mainly based on the weak points of OCR (Optical Character Recognition) systems while using them are undesirable to human users. In this paper a method has been presented for... 

Motivated by the problem of automatically composing network accessible services, such as those on the World Wide Web, this paper proposes an approach to exploit all semantic information available for semantic web services to complete this task. For semantic web services, we propose a prioritized and limited list of other web services, with composability measure value based on feasibility of being succeeding service in a composition to fulfill the request. Due to the fact that in semantic web services, semantic information- all description and signatures- is defined by ontology language, we use some of the mapping rules to estimate a correlation between services for composability. Then we... 

So far several methods have been proposed for clustering the web. On the other hand, many algorithms have been developed for clustering the relational data, but their usage for the Web is to be investigated. One main category of such algorithms is density based methods providing high quality results. In this paper first, a new density based algorithm is introduced and then it is compared with other algorithms of this category. The proposed algorithm has some interesting properties and capabilities such as hierarchical clustering and sampling, making it suitable for clustering the web data. © Springer-Verlag Berlin Heidelberg 2007  

Clustering is one of the most important approaches for mining and extracting knowledge from the web. In this paper a method for clustering the web data is presented which using a Bayesian network, finds appropriate representatives for each of the clusters. Having those representatives, we can create more accurate clusters. Also the contents of the web pages are converted into vectors which firstly, the number of dimensions is reduced, and secondly the orthogonality problem is solved. Experimental results show about the high quality of the resultant clusters. © Springer-Verlag Berlin Heidelberg 2007  

Recent years have witnessed an explosion in the availability of news articles on the World Wide Web. In addition, organizing the results of a news search facilitates the user(s) in overviewing the returned news. In this work, we have focused on the label-based clustering approaches for news meta-search engines, and which clusters news articles based on their topics. Furthermore, our engine for NEws meta-Search REsult Clustering (NeSReC) is implemented along. NeSReC takes queries from the users and collect the snippets of news which are retrieved by The Altavista News Search Engine for the queries. Afterwards, it performs the hierarchical clustering and labeling based on news snippets in a... 

Graph data structures are fundamental for studying connected entities. With an increase in the number of applications where data is represented as graphs, the problem of graph generation has recently become a hot topic. However, despite its significance, conditional graph generation that creates graphs with desired features is relatively less explored in previous studies. This paper addresses the problem of class-conditional graph generation that uses class labels as generation constraints by introducing the Class Conditioned Graph Generator (CCGG). We built CCGG by injecting the class information as an additional input into a graph generator model and including a classification loss in its...