This paper extends the graphical and formal language of UML-B to provide the ability to model probabilities. Discrete probabilities, interval probabilities, and stochastic delays are added to the UML-B's state-machine syntax, and their corresponding semantics are defined in Event-B. In addition, as a secondary contribution, UML-B (probabilistic) state-machine models are defined as MDP (Markov Decision Process) models in order to provide a means of quantitative verification in PRISM (Probabilistic Symbolic Model Checker). As an important feature of the proposed method, it does not change the Event-B syntax or semantics. To evaluate this work, as a case study, the Zeroconf protocol will be modeled in the extended UML-B using the Rodin tool, and its Event-B counterpart is converted to a PRISM model. The results of evaluations indicate that this study's additions provide the capability of modeling and verification of probabilistic and stochastic systems.

Users usually want to gather information about what they are interested in, which could be achieved by entity augmentation using a vast amount of web tables. Existing techniques assume that web tables are entity-attribute binary tables. As for tables having multiple columns to be augmented, they will be split into several entity-attribute binary relations, which would cause semantic fragmentation. Furthermore, the result table consolidated by binary relations will suffer from entity inconsistency and low precision. The objective of our research is to return a consistent result table for entity augmentation when given a set of entities and attribute names. In this paper we propose a web information gathering framework based on consistent entity augmentation. To ensure high consistency and precision of the result table we propose that answer tables for building result table should have consistent matching relationships with each other. Instead of splitting tables into pieces we regard web tables as nodes and consistent matching relationships as edges to make a consistent clique and expand it until its coverage for augmentation query reaches certain threshold gamma. It is proved in this paper that a consistent result table could be built by considering tables in consistent clique to be answer tables. We tested our method on four real-life datasets, compared it with different answer table selection methods and state-of-the-art entity augmentation technique based on table fragmentation as well. The results of a comprehensive set of experiments indicate that our entity augmentation framework is more effective than the existing method in getting consistent entity augmentation results with high accuracy and reliability.

Materialized views are important for optimizing Business Intelligence (BI) systems when they are designed without data cubes. Selecting candidate queries from large number of queries for materialized views is a challenging task. Most of the work done in the past involves finding out frequent queries from the past workload and creating materialized views from such queries by either manually analyzing workload or using approximate string matching algorithms using query text. Most of the existing methods suggest complete queries but ignore query components such as sub queries for creation of materialized views. This paper presents a novel method to determine on which queries and query components materialized views can be created to optimize aggregate and join queries by mining database of query execution plans which are in the form of binary trees. The proposed algorithm showed significant improvement in terms of more number of optimized queries because it is using the execution plan tree of the query as a basis of selection of query to be optimized using materialized views rather than choosing query text which is used by traditional methods. For selecting a correct set of queries to be optimized using materialized views, the paper proposes efficient specialized frequent tree component mining algorithm with novel heuristics to prune search space. These frequent components are used to determine the possible set of candidate queries for creation of materialized views. Experimentation on standard, real and synthetic data sets, and also the theoretical basis, proved that the proposed method is able to optimize a large number of queries with less number of materialized views and showed a significant improvement in performance compared to traditional methods.

We aim to speed up approximate keyword matching with the use of a lightweight, fixed-size block of data for each string, called a fingerprint. These work in a similar way to hash values; however, they can be also used for matching with errors. They store information regarding symbol occurrences using individual bits, and they can be compared against each other with a constant number of bitwise operations. In this way, certain strings can be deduced to be at least within the distance k from each other (using Hamming or Levenshtein distance) without performing an explicit verification. We show experimentally that for a preprocessed collection of strings, fingerprints can provide substantial speedups for k = 1, namely over 2.5 times for the Hamming distance and over 30 times for the Levenshtein distance. Tests were conducted on synthetic and real-world English and URL data.

The ease of creating digital content coupled with technological advancements allows institutions and organizations to further embrace distance learning. Teaching materials also receive attention, because it is difficult for the student to obtain adequate didactic material, being necessary a high effort and knowledge about the material and the repository. This work presents a framework that enables the automatic metadata generation for materials available in educational video repositories. Each module of the framework works autonomously and can be used in isolation, complemented by another technique or replaced by a more appropriate approach to the field of use, such as repositories with other types of media or other content.

With the coming up of plethora of web applications and technologies like sensors, IoT, cloud computing, etc., the data generation resources have increased exponentially. Stream processing requires real time analytics of data in motion and that too in a single pass. This paper proposes a framework for hourly analysis of streamed data using Bloom filter, a probabilistic data structure where hashing is done by using a combination of double hashing and partition hashing; leading to less inter-hash function collision and decreased computational overhead. When size of incoming data is not known, use of Static Bloom filter leads to high collision rate if data flow is too much, and wastage of storage space if data is less. In such cases it is difficult to determine the optimal Bloom filter parameters (m, k) in advance, thus a target threshold for false positives (f_p) cannot be guaranteed. To accommodate the growing data size, one of the major requirements in Bloom filter is that filter size m should grow dynamically. For predicting the array size of Bloom filter Kalman filter has been used. It has been experimentally proved that proposed Adaptable Bloom Filter (ATBF) efficiently performs peak hour analysis, server utilization and reduces the time and space required for querying dynamic datasets.

In computing, duplicate data detection refers to identifying duplicate copies of repeating data. Identifying duplicate data items in streamed data and eliminating them before storing, is a complex job. This paper proposes a novel data structure for duplicate detection using a variant of stable Bloom filter named as FingerPrint Stable Bloom Filter (FP-SBF). The proposed approach uses counting Bloom filter with fingerprint bits along with an optimization mechanism for duplicate detection. FP-SBF uses d-left hashing which reduces the computational time and decreases the false positives as well as false negatives. FP-SBF can process unbounded data in single pass, using k hash functions, and successfully differentiate between duplicate and distinct elements in O(k+1) time, independent of the size of incoming data. The performance of FP-SBF has been compared with various Bloom Filters used for stream data duplication detection and it has been theoretically and experimentally proved that the proposed approach efficiently detects the duplicates in streaming data with less memory requirements.

The popularity of cloud computing is increasing by leaps and bounds. To cope with resource demands of increasing number of cloud users, the cloud market players establish large sized data centers. The huge energy consumption by the data centers and liability of fulfilling Quality of Service (QoS) requirements of the end users have made resource allocation a challenging task. In this paper, energy and QoS aware resource allocation approach which employs Antlion optimization for allocation of resources to virtual machines (VMs) is proposed. It can operate in three modes, namely power aware, performance aware, and balanced mode. The proposed approach enhances energy efficiency of the cloud infrastructure by improving the utilization of resources while fulfilling QoS requirements of the end users. The proposed approach is implemented in CloudSim. The simulation results have shown improvement in QoS and energy efficiency of the cloud.

The aim of the vertex clique covering problem (CCP) is to cover the vertices of a graph with as few cliques as possible. We analyse the iterated greedy (IG) algorithm for CCP, which was previously shown to provide strong empirical results for real-world networks. It is demonstrated how the techniques of analysis for randomised search heuristics can be applied to IG, and several practically relevant results are obtained. We show that for triangle-free graphs, IG solves CCP optimally in expected polynomial time. Secondly, we show that IG finds the optimum for CCP in a specific case of sparse random graphs in expected polynomial time with high probability. For Barabási-Albert model of scale-free networks, which is a canonical model explaining the growth of social, biological or computer networks, we obtain that IG obtains an asymptotically optimal approximation in polynomial time in expectation. Last but not least, we propose a slightly modified variant of IG, which guarantees expected polynomial-time convergence to the optimum for graphs with non-overlapping triangles.

In this paper we propose very efficient itemset representation for frequent itemset mining from transactional databases. The combinatorial number system is used to uniquely represent frequent k-itemset with just one integer value, for any k ≥ 2. Experiments show that memory requirements can be reduced up to 300 %, especially for very low minimal support thresholds. Further, we exploit combinatorial number schema for representing candidate itemsets during iterative join-based approach. The novel algorithm maintains one-dimensional array rank, starting from k = 2nd iteration. At the index r of the array, the proposed algorithm stores unique integer representation of the r-th candidate in lexicographic order. The rank array provides joining of two candidate k-itemsets to be O(1) instead of O(k) operation. Additionally, the rank array provides faster determination which candidates are contained in the given transaction during the support count and test phase. Finally, we believe that itemset ranking by combinatorial number system can be effectively integrated into pattern-growth algorithms, that are state-of-the-art in frequent itemset mining, and additionally improve their performances.