Shalini Batra

A review of ranking approaches for semantic search on Web

With ever increasing information being available to the end users, search engines have become the most powerful tools for obtaining useful information scattered on the Web. However, it is very common that even most renowned search engines return result sets with not so useful pages to the user. Research on semantic search aims to improve traditional information search and retrieval methods where the basic relevance criteria rely primarily on the presence of query keywords within the returned pages. This work is an attempt to explore different relevancy ranking approaches based on semantics which are considered appropriate for the retrieval of relevant information. In this paper, various pilot projects and their corresponding outcomes have been investigated based on methodologies adopted and their most distinctive characteristics towards ranking. An overview of selected approaches and their comparison by means of the classification criteria has been presented. With the help of this comparison, some common concepts and outstanding features have been identified.

A Novel Technique for Call Graph Reduction for Bug Localization

era, software industries are competing for software quality which depends upon the sound software testing phase. To deliver the quality product the most challenging job for software industry is to localize bugs automatically and fix them before release. One of the techniques for automated bug localization is usage of call graph. Since size of the call graph generated is quite large, various call reduction approaches have been proposed. In this paper a novel approach for call graph reduction has been proposed where the size of the call graph is reduced without changing the basic structure and no major loss of the information is incurred. The output generated using the proposed methodology shows promising results.

Web Service Categorization Using Normalized Similarity Score

 Abstract—Service discovery is one of challenging issues in Service-Oriented computing. Currently, most of the existing service discovering and matching approaches are based on keywords-based strategy. However, this method is inefficient and time-consuming. Based on the current dominating mechanisms of discovering and describing Web services with UDDI and WSDL, a novel approach for Web service categorization is proposed, where WSDLs documentation tag is used as only means to describe information pertaining to the entire Web services functionality which is used in conjunction with the current Web service standards, to automatically categorize a Web service into a one of the pre-defined categories. The words are extracted from WSDL of a Web service and Nearest Similarity Score (NSS), a Measure of Semantic Relatedness (MSR) of each word is calculated with every pre-defined category. Total value of all the words is calculated through the NSS and then Web service is assigned a category based on the sum of MSR of all the words provided in the Web service description tag. This work enables automatic semantic categorization of Web services.

Collaborating trust and item-prediction with ant colony for recommendation

Online Recommenders are information filtering systems which works on the implicit or explicit information provided by the users and Collaborative Filtering is most widely used technique for this. But the accuracy of the recommendation process is greatly affected by the sparsity in user-item matrix. Though, collaborative filtering is one of the most promising techniques, it still suffers from the cold start problem due to which it is unable to give recommendations to new users. It is also vulnerable to many attacks like shilling attack, grey sheep, etc. which severely hamper the recommendation systems. A trust-based approach combining trust and swarm intelligence (ant colony) with collaborative filtering has been proposed. It also uses item-based predictions in the process of generating recommendations. Ant Colony exhibit self organizing and distributed properties due to which it is used in real time and constantly changing environment. Trust is updated continuously using pheromone updating strategy of ant colony thus, making the system more accurate. By combining these approaches, effective system is proposed which provide solutions to the above mentioned problems of collaborative filtering and predict whether the user will like the certain item or not. Results have been validated using dataset of movies which is available online.

Fuzzified Cuckoo based Clustering Technique for Network Anomaly Detection

Abstract With the increasing penetration of security threats, the severity of their impact on the underlying network has increased manifold. Hence, a robust anomaly detection technique, Fuzzified Cuckoo based Clustering Technique (F-CBCT), is proposed in this paper which operates in two phases: training and detection. The training phase is supported using Decision Tree followed by an algorithm based on hybridization of Cuckoo Search Optimization and K-means clustering. In the designed algorithm, a multi-objective function based on Mean Square Error and Silhouette Index is employed to evaluate the two simultaneous distance functions namely-Classification measure and Anomaly detection measure. Once the system is trained, detection phase is initiated in which a fuzzy decisive approach is used to detect anomalies on the basis of input data and distance functions computed in the previous phase. Experimental results in terms of detection rate (96.86%), false positive rate (1.297%), accuracy (97.77%) and F-Measure (98.30%) prove the effectiveness of the proposed model.

Privacy-preserving authentication framework using bloom filter for secure vehicular communications

Vehicular ad hoc networks (VANETs) are the future of the intelligent transportation systems (ITS), which aim to improve traffic safety. The received message in VANETs can contain the malicious content that may affect the entire network; hence, these networks are more prone to such attacks. Thus, security is a major consideration before the deployment of such network. In this paper, a secure privacy-preserving authentication framework is proposed, which employs the use of pseudonyms for anonymous communication. A new digital signature scheme and aggregate verification scheme are designed for vehicular communications, and the ID-based signature scheme is used for vehicle-to-RSU communication. The multiple authorities are involved in revealing the identity of the vehicle in case of revocation. The signature verification scheme is improved by the use of bloom filters, and the results achieved by the proposed scheme have been implemented on a simulated environment.

P-skip graph : an efficient data structure for peer to peer network

Peer-to-peer networks display interesting characteristics of fast queries, updation, deletion, fault-tolerance etc., while lacking any central authority. Adjacency Matrix, Skip-Webs, Skip-Nets, Skip-List, Distributed Hash Table, and many more data structures form the candidature for peer-to-peer networks, of which, Skip-Graph (evolved version of skip-list) displays one of the best characteristics as it help to search and locate a node in a peer-to-peer network efficiently with time complexity being O(log n). However when a hotspot node is searched and queried again and again, the Skip-Graph does not learn or adapt to the situation and still searches traditionally with O(log n) complexity. In this paper we propose a new data structure P-skip graph, a modified version of Skip graph, which reduces the search time of a hot spot node drastically from initial time of O(log n). Results provided by Simulations of a skip graph-based Peer-to-Peer application demonstrate that the proposed approach can in fact effectively decrease the search time to O(1).

Using LSI and its variants in Text Classification

Latent Semantic Indexing (LSI), a well known technique in Information Retrieval has been partially successful in text retrieval and no major breakthrough has been achieved in text classification as yet. A significant step forward in this regard was made by Hofmann[3], who presented the probabilistic LSI (PLSI) model, as an alternative to LSI. If we wish to consider exchangeable representations for documents and words, PLSI is not successful which further led to the Latent Dirichlet Allocation (LDA) model [4]. A new local Latent Semantic Indexing method has been proposed by some authors called “Local Relevancy Ladder-Weighted LSI” (LRLW-LSI) to improve text classification [5]. In this paper we study LSI and its variants in detail , analyze the role played by them in text classification and conclude with future directions in this area.

An efficient multi-party scheme for privacy preserving collaborative filtering for healthcare recommender system

Abstract Patient oriented decision-making in medical domains can enhance the efficiency of the modern healthcare recommender system provided the data scattered across different geographical regions is collected, mined and analyzed efficiently. Different sites, having Arbitrary Distributed Data (ADD) of healthcare services at various nodes can collaborate with each other to generate customer’s preference leading to mutual advantage and overcoming of the issues related to insufficient ratings of various medical services. However, due to privacy, financial and legal issues, different parties defer from sharing their confidential data. If the parties are assured of data confidentiality, they might agree for fruitful collaboration. Few existing studies proposed Privacy Preserving Collaborative Filtering (PPCF) on ADD, but these techniques considered only two parties. Moreover, the computation cost of off-line model generation process is high since these techniques use homomorphic encryption techniques. To fill these gaps, this paper propose PPCF scheme on ADD based on multi-party random masking and polynomial aggregation techniques. In the proposal, two phases are considered namely as: off-line model generation and online prediction generation. Three protocols have been considered for privacy preservation so that analysis of each protocol is performed separately. The Paillier homomorphic encryption system is also used to calculate the length of vector X securely, so that only additive property of homomorphic encryption is used. Analysis of the proposed scheme has been done for security, accuracy, coverage and performance on healthcare and Movieslens datasets. It has been experimentally demonstrated that the proposed scheme maintains data owner’s confidentiality, and privacy measure so that it does not affect the accuracy of prediction generation on integrated data. Comparative analysis of the proposed scheme has also been done with other related schemes based on off-line and online computation overheads. The results obtained demonstrated that the proposed scheme has significant improvement by a factor of 36% (approx) with respect to the aforementioned parameters.

Bloom filter based optimization scheme for massive data handling in IoT environment

Abstract With the widespread popularity of big data usage across various applications, need for efficient storage, processing, and retrieval of massive datasets generated from different applications has become inevitable. Further, handling of these datasets has become one of the biggest challenges for the research community due to the involved heterogeneity in their formats. This can be attributed to their diverse sources of generation ranging from sensors to on-line transactions data and social media access. In this direction, probabilistic data structures (PDS) are suitable for large-scale data processing, approximate predictions, fast retrieval and unstructured data storage. In conventional databases, entire data needs to be stored in memory for efficient processing, but applications involving real time in-stream data demand time-bound query output in a single pass. Hence, this paper proposes Accommodative Bloom filter (ABF), a variant of scalable bloom filter, where insertion of bulk data is done using the addition of new filters vertically. Array of m bits is divided into b buckets of l bits each and new filters of size ‘ m ∕ k ′ are added to each bucket to accommodate the incoming data. Data generated from various sensors has been considered for experimental purposes where query processing is done at two levels to improve the accuracy and reduce the search time. It has been found that insertion and search time complexity of ABF does not increase with increase in number of elements. Further, results indicate that ABF outperforms the existing variants of Bloom filters in terms of false positive rates and query complexity, especially when dealing with in-stream data.