Vikram Goyal

UP-Hist Tree: An Efficient Data Structure for Mining High Utility Patterns from Transaction Databases

High-utility itemset mining is an emerging research area in the field of Data Mining. Several algorithms were proposed to find high-utility itemsets from transaction databases and use a data structure called UP-tree for their working. However, algorithms based on UP-tree generate a lot of candidates due to limited information availability in UP-tree for computing utility value estimates of itemsets. In this paper, we present a data structure named UP-Hist tree which maintains a histogram of item quantities with each node of the tree. The histogram allows computation of better utility estimates for effective pruning of the search space. Extensive experiments on real as well as synthetic datasets show that our algorithm based on UP-Hist tree outperforms the state of the art pattern-growth based algorithms in terms of the total number of candidate high utility itemsets generated that needs to be verified.

EcoTop: an economic model for dynamic processing of top-k queries in mobile-P2P networks

This work addresses the processing of top-k queries in mobile ad hoc peer to peer (M-P2P) networks using economic schemes. Our proposed economic model, designated as EcoTop, issues economic rewards to the mobile peers, which send relevant data items (i.e., those that contribute to the top-k query result), and penalizes peers for sending irrelevant items, thereby incentivizing the optimization of communication traffic. The main contributions of our work are three-fold. First, we propose the EcoTop economic model for efficient top-k query processing in M-P2P networks. Second, we propose two schemes, namely ETK and ETK+, for assigning rewards/penalties to peers and for enabling peers to re-evaluate the scores of their data items for item re-ranking purposes. Third, we conduct a performance study, which demonstrates that EcoTop is indeed effective in improving the performance of top-k queries, while minimizing the communication traffic. Notably, our novel economic incentive model also discourages free-riding in M-P2P networks

Open Source Social Media Analytics for Intelligence and Security Informatics Applications

Open-Source Intelligence OSINT is intelligence collected and inferred from publicly available and overt sources of information. Open-Source social media intelligence is a sub-field within OSINT with a focus on extracting insights from publicly available data in Web 2.0 platforms like Twitter micro-blogging website, YouTube video-sharing website and Facebook social-networking website. In this paper, we present an overview of Intelligence and Security Informatics ISI applications in the domain of open-source social media intelligence. We present technical challenges and introduce basic Machine Learning based framework, tools and techniques within the context of open-source social media intelligence using two case-studies. The focus of the paper is on mining free-form textual content present in social media websites. In particular we describe two important application: online radicalization and civil unrest. In addition to covering basic concepts and applications, we discuss open research problem, important papers, publication venues, research results and future directions.

Malafide intension based detection of privacy violation in information system

In the past few years there has been an increased focus on privacy issues for Information Systems. This has resulted in concerted systematic work focused on regulations, tools and enforcement. Despite this, privacy violations still do take place. Therefore there is an increased need to develop efficient methods to detect privacy violations. After a privacy violation has taken place, the post-event diagnostics should make use of any post-event information which might be available. This information (malafide intention) might play a decisive role in determining violations. In this paper we propose one such framework which makes use of malafide intentions. The framework is based on the hypothesis that any intrusion/unauthorized access has a malafide intention always associated with it and is available in a post-event scenario. We hereby propose that by analyzing the privacy policies and the available malafide intention, it is possible to detect probable privacy violations.

Query rewriting for detection of privacy violation through inferencing

When a privacy violation is detected the intension behind the violation is revealed. We refer to this as a malafide intension and the information revealed as the target. The target can be expressed using an SQL-like syntax. In sophisticated privacy attacks the target of the attack may not have been directly accessed but inferred from other pieces of information by exploiting functional dependencies present in the application domain. In this paper we present an efficient algorithm to rewrite the malafide intension query attributes which will return the minimal set of attribute from which the target can be derived. The attribute sets returned by algorithm can derive the target using functional dependencies (algorithm is sound) and furthermore if any minimal set can derive the target using functional dependencies then it will be returned by the algorithm (algorithm is complete).

High Utility Rare Itemset Mining over Transaction Databases

High-Utility Rare Itemset (HURI) mining finds itemsets from a database which have their utility no less than a given minimum utility threshold and have their support less than a given frequency threshold. Identifying high-utility rare itemsets from a database can help in better business decision making by highlighting the rare itemsets which give high profits so that they can be marketed more to earn good profit. Some two-phase algorithms have been proposed to mine high-utility rare itemsets. The rare itemsets are generated in the first phase and the high-utility rare itemsets are extracted from rare itemsets in the second phase. However, a two-phase solution is inefficient as the number of rare itemsets is enormous as they increase at a very fast rate with the increase in the frequency threshold. In this paper, we propose an algorithm, namely UP-Rare Growth, which uses UP-Tree data structure to find high-utility rare itemsets from a transaction database. Instead of finding the rare itemsets explicitly, our proposed algorithm works on both frequency and utility of itemsets together. We also propose a couple of effective strategies to avoid searching the non-useful branches of the tree. Extensive experiments show that our proposed algorithm outperforms the state-of-the-art algorithms in terms of number of candidates.

Efficient Trajectory Cover Search for Moving Object Trajectories

Given a set of query locations and a set of query keywords, a Trajectory Cover (CT) query over a repository of mobile trajectories returns a minimal set of trajectories that maximally covers the query keywords and are also spatially close to the query locations. Processing CT queries over mobile trajectories requires substantially different algorithms than those for location range queries. The main contributions of this work are three-fold. First, we introduce a notion of Trajectory Cover that enables mobile users to get most relevant trajectories of their interest to plan their route of travel. Second, we show that CT search is an NP-hard problem and we present a greedy algorithm that combines spatial proximity and keyword proximity with an efficient filter. We use the promising value filter to select better candidate trajectories for inclusion in the CT and prune out non-promising trajectories in the Greedy search process. We also develop a lower bound and upper bound mechanism to efficiently compute the promising value of each trajectory, allowing our promising value based greedy algorithm to scale to large trajectory databases. Finally, we conduct a performance study to demonstrate that our greedy algorithm is efficient in evaluating trajectory cover queries.

Malafide Intension and its mapping to Privacy Policy Purposes for Masquerading

In presence of a robust privacy infrastructure, an attacker can fulfil his purpose (malafide intension) only by masquerading it with bonafide purposes besides other authentication parameters. We address the issue of masquerading of purpose for a malafide intension by defining the mapping from a malafide intension to bonafide purposes in this paper. An understanding of such a mapping can facilitate both a hacker (assist him in masquerading) and a forensic expert to investigate malafide accesses. Determination of these bonafide purposes may help speed up the violation detection if the user accesses log has listed bonafide purpose with each user access. The bonafide purposes can be determined in data-independent (without accessing the database) or data-dependent (database access is required) mode. In this paper we define a mapping of a malafide intension to bonafide purposes in data-independent mode

TiDE: Template-Independent Discourse Data Extraction

The problem of Discourse Data Extraction focuses on identifying comments and reviews from social networking websites. Existing approaches for Discourse Data extraction are either template-dependent or they are limited to comment-posting-structure discovery. We are not aware of any technique that extracts the detailed comment information like comment text, commenter and discussion structure from the comment page. In this paper, we present a template-independent two step approach, namely TiDE, which extracts the discourse data such as comments, reviews, posts and structural relationship among them. In the first step, we parse the input comment page to prepare a Document Object Model tree and then find the location of discourse data in the tree using the concept of Path-Strings. The outputs of the first step are Comment Blocks and these Comment Blocks are leveraged in second step to extract the comments, commenter and discussion structure. Experimental studies on 19 well known Discourse websites having different templates show that our Comment Block discovery is more adaptable than the existing posting-structure discovery technique. We are able to extract 97 % of comment-text and 79 % commenter information which is significant compared to state of the art techniques. We also show the usefulness of TiDE by building a news comment crawler.

Efficient Skyline Itemsets Mining

Utility Mining (UM) in context of Market Basket Analysis consists of mining itemsets from a transaction database guided by optimizing utility. For example, UM consists of extracting all itemsets in a transaction database having utility above a user-defined minimum threshold or mining Top-K high utility itemset. Similarly, Frequent Itemset Mining (FIM) finds frequent patterns using a frequency threshold. However, none of these pattern mining methods determine patterns that are interesting in both the aspects of utility and frequency. In addition these methods require a user to specify respective thresholds. In this paper, we present a novel framework for mining a new pattern called as Utility-Frequency Skyline Pattern. We formalize our problem as a pattern search problem and propose an efficient technique on recently proposed popular data structure called as UP Tree (Utility-Pattern Tree). The proposed algorithm consists of two phases called as Filter and Refine. In the Filter phase, a set of candidate itemsets are mined, which are then verified finally in the Refine phase. We study the effectiveness of our proposed algorithm along with two heuristics and conclude that our proposed method is efficient.