Syed Khairuzzaman Tanbeer

PUF-Tree: A Compact Tree Structure for Frequent Pattern Mining of Uncertain Data

Many existing algorithms mine frequent patterns from traditional databases of precise data. However, there are situations in which data are uncertain. In recent years, researchers have paid attention to frequent pattern mining from uncertain data. When handling uncertain data, UF-growth and UFP-growth are examples of well-known mining algorithms, which use the UF-tree and the UFP-tree respectively. However, these trees can be large, and thus degrade the mining performance. In this paper, we propose (i) a more compact tree structure to capture uncertain data and (ii) an algorithm for mining all frequent patterns from the tree. Experimental results show that (i) our tree is usually more compact than the UF-tree or UFP-tree, (ii) our tree can be as compact as the FP-tree, and (iii) our mining algorithm finds frequent patterns efficiently.

Finding Strong Groups of Friends among Friends in Social Networks

Over the past few years, the rapid growth and the exponential use of social digital media has led to an increase in popularity of social networks and the emergence of social computing. In general, social networks are structures made of social entities (e.g., individuals) that are linked by some specific types of interdependency such as friendship. Most users of social media (e.g., Face book, Google+, Linked In, My Space, Twitter) have many linkages in terms of friends, connections, and/or followers. Among all these linkages, some of them are more important than another. For instance, some friends of a user may be casual ones who acquaintances met him at some points in time, whereas some others may be friends that care about him in such a way that they frequently post on his wall, view his updated profile, send him messages, invite him for events, and/or follow his tweets. In this paper, we apply data mining techniques to social networks to help users of the social digital media to distinguish these important friends from a large number of friends in their social networks.

Fast tree-based mining of frequent itemsets from uncertain data

Over the past two decades, numerous algorithms have been proposed for mining frequent itemsets from precise data. However, there are situations in which data are uncertain. In recent years, tree-based algorithms have been proposed to mine frequent itemsets from uncertain data. While the key success of tree-based algorithms for mining precise data is due to the compactness of a tree structure in capturing precise data, the corresponding tree structure in capturing uncertain data may not be so compact. In this paper, we propose a novel tree structure for capturing uncertain data such that it is as compact as the tree for capturing precise data. Moreover, we also propose two fast algorithms that use this compact tree structure to mine frequent itemsets. Experimental results showed the compactness of our tree and the effectiveness of our algorithms in mining frequent itemsets from uncertain data.

Interactive Mining of Strong Friends from Social Networks and Its Applications in E-Commerce

Social networks are generally made of individuals who are linked by some types of interdependencies such as friendship. Most individuals in social networks have many linkages in terms of friends, connections, and/or followers. Among these linkages, some of them are stronger than others. For instance, some friends may be acquaintances of an individual, whereas others may be friends who care about him or her (e.g., who frequently post on his or her wall). In this study, we integrate data mining with social computing to form a social network mining algorithm, which helps the individual distinguish these strong friends from a large number of friends in a specific portion of the social networks in which he or she is interested. Moreover, our mining algorithm allows the individual to interactively change his or her mining parameters. Furthermore, we discuss applications of our social mining algorithm to organizational computing and e-commerce

Interactive discovery of influential friends from social networks

Social networks, which are made of social entities (e.g., individual users) linked by some specific types of interdependencies such as friendship, have become popular to facilitate collaboration and knowledge sharing among users. Such interactions or interdependencies can be dependent on or influenced by user characteristics such as connectivity, centrality, weight, importance, and activity in the networks. As such, some users in the social networks can be considered as highly influential to others. In this article, we propose a computational model that integrates data mining with social computing to help users discover influential friends from a specific portion of the social networks that they are interested in. Moreover, our social network analysis and mining model also allows users to interactively change their mining parameters (e.g., scopes of their interested portions of the social networks).

Finding Popular Friends in Social Networks

The emergence of social computing enables users to intersect social behaviour with computing systems and to create social conventions as well as social contexts through the use of software and technology. Social networking sites have become popular to facilitate collaboration and knowledge sharing between users. A rich set of information is embedded in these social media data. In this paper, we propose a data mining algorithm to help users to find popular friends in social networks. This can be considered as an application of data mining and social computing.

Finding groups of friends who are significant across multiple domains in social networks

Social networking websites such as Facebook, LinkedIn, Twitter, and Weibo have been used for collaboration and knowledge sharing between users. The mining of social network data has become an important topic in data mining and computational aspects of social networks. Nowadays, it is not uncommon for most users in a social network to have many friends and in multiple social domains. Among these friends, some groups of friends are more significant than others. In this paper, we introduce a data mining technique that helps social network users find groups of friends who are significant across multiple domains in social networks.

Mining social networks for significant friend groups

The emergence of Web-based communities and hosted services such as social networking sites has facilitated collaboration and knowledge sharing between users. Hence, it has become important to mine this vast pool of data in social networks, which are generally made of users linked by some specific interdependency such as friendship. For any user, some groups of his friends are more significant than others. In this paper, we propose a tree-based algorithm to mine social networks to help these users to distinguish their significant friend groups from all the friends in their social networks.

Mining popular patterns from transactional databases

Since the introduction of the frequent pattern mining problem, researchers have extended frequent patterns to different useful patterns such as cyclic, emerging, periodic and regular patterns. In this paper, we introduce popular patterns, which captures the popularity of individuals, items, or events among their peers or groups. Moreover, we also propose (i) the Pop-tree structure to capture the essential information for the mining of popular patterns and (ii) the Pop-growth algorithm for mining popular patterns. Experimental results showed that our proposed tree structure is compact and space efficient and our proposed algorithm is time efficient.

Finding Diverse Friends in Social Networks

Social networks are usually made of users linked by friendship, which can be dependent on (or influenced by) user characteristics (e.g., connectivity, centrality, weight, importance, activity in the networks). Among many friends of these social network users, some friends are more diverse (e.g., more influential, prominent, and/or active in a wide range of domains) than other friends in the networks. Recognizing these diverse friends can provide valuable information for various real-life applications when analyzing and mining huge volumes of social network data. In this paper, we propose a tree-based mining algorithm that finds diverse friends, who are highly influential across multiple domains, in social networks.