Location-based services refer to services that use location as primary input. But accessing user's location by an adversary invites issues of privacy breach. Instead of specific location coordinates, its surrounding area known as cloaking region is revealed in order to get the service. K anonymity technique of location privacy ensures that at least K-1 users should be included within a specific cloaked region. Researches have established that on combining K anonymity with the idea of including similar users together in a cloaked region provides stringent privacy (especially from background and heterogeneity attacks). This work quantifies the amount of privacy gain attained through, opting-for users with similar profiles instead of random users. The quantification is done by using KL divergence. Values of KL divergence of user profiles have been calculated for different cloaking regions containing similar and random users. Low KL divergence values depict privacy gains up to 33% for users with similar profiles.

This paper explores the social quality (goodness) of community structures formed across Twitter users, where social links within the structures are estimated based upon semantic properties of user-generated content (corpus). We examined the overlap of the community structures of the constructed graphs, and followership-based social communities, to find the social goodness of the links constructed. Unigram, bigram and LDA content models were empirically investigated for evaluation of effectiveness, as approximators of underlying social graphs, such that they maintain the community social property. Impact of content at varying granularities, for the purpose of predicting links while retaining the social community structures, was investigated. 100 discussion topics, spanning over 10 Twitter events, were used for experiments. The unigram language model performed the best, indicating strong similarity of word usage within deeply connected social communities. This observation agrees with the phenomenon of evolution of word usage behavior, that transform individuals belonging to the same community tending to choose the same words, made by [1], and raises a question on the literature that use, without validation, LDA for content-based social link prediction over other content models. Also, semantically finer-grained content was observed to be more effective compared to coarser-grained content.

Existing systems deliver high accuracy and F1-scores for detecting paraphrase and semantic similarity on traditional clean-text corpus. For instance, on the clean-text Microsoft Paraphrase benchmark database, the existing systems attain an accuracy as high as 0:8596. However, existing systems for detecting paraphrases and semantic similarity on user-generated short-text content on microblogs such as Twitter, comprising of noisy and ad hoc short-text, needs significant research attention. In this paper, we propose a machine learning based approach towards this. We propose a set of features that, although well-known in the NLP literature for solving other problems, have not been explored for detecting paraphrase or semantic similarity, on noisy user-generated short-text data such as Twitter. We apply support vector machine (SVM) based learning. We use the benchmark Twitter paraphrase data, released as a part of SemEval 2015, for experiments. Our system delivers a paraphrase detection F1-score of 0.717 and semantic similarity detection F1-score of 0.741, thereby significantly outperforming the existing systems, that deliver F1-scores of 0.696 and 0.724 for the two problems respectively. Our features also allow us to obtain a rank among the top-10, when trained on the Microsoft Paraphrase corpus and tested on the corresponding test data, thereby empirically establishing our approach as ubiquitous across the different paraphrase detection databases.

With the wide spread usage of LBS, convenience has reached on the finger tips of mobile users, but on the other side, it has escalated many security and privacy concerns. In this paper we address the location K-anonymity problem using fuzzy spatio-temporal attributes, a new perspective of looking at privacy issue in location privacy. In the context of LBSs and mobile clients, location K-anonymity refers to K-anonymous usage of location information. A novel approach for determining location disclosure based on fuzzy attributes of spatio-temporal context is proposed which in turn will give us a value of K for K-anonymity purpose. Spatio-temporal fuzzy attributes for privacy issues are identified and Fuzzy Inference System (FIS) is implemented that takes these attributes as input and generates location disclosure value. Using Location disclosure value, K is computed for K-anonymity to ensure privacy. This value of K is directly based on current spatio temporal context and is valid for all users present in that context. Further, an exhaustive rule base of fuzzy rules is generated based on responses obtained by conducting survey on the potential users who frequently use POI (Point of Interest) services. Later on, fuzzy rules for FIS rule base are extracted using Fuzzy C Means (FCM) clustering technique. Using the rules extracted through FCM, the size of rule base is reduced and the performance of the FIS is evaluated. Number of rules in rule base is decreased for scalability and efficiency purposes. Root Mean Square Error (RMSE) for every reduced set is computed and compared with initial exhaustive rule base. It is observed that size of rule base can be decreased to a considerable extent.

With the increase in number of available interesting locations, it becomes difficult for users to find interesting ones, thus imposes a need for recommender systems to suggest interesting locations. Further, to ease the user's decision making, the amount of supplementary information, such as right time to visit, weather conditions, traffic condition, right mode of transport, crowdedness, security alerts, etc., may be annotated with the list of recommended locations. This paper explores the possibility of enriching tourist locations using crowd sourcing approach, which can be used by Tourist Spot Recommender System (TSRS) for mobile users. Proposed crowd sourcing system focuses on getting work done from the crowd currently available at the location under consideration. In proposed system, the contributed information is not limited to ones available on blogs, web pages and sensor-readings from the device etc., but includes proactively-generated user's opinions and perspectives, that are processed to offer immediate knowledge. Our system works in collaboration with a TSRS, takes the list of locations to be recommended to the current user and performs just-in-time information enrichment for those selected set of locations. We have implemented a prototype of proposed systems using java android software development toolkit and evaluated this system by 76 real users.

The increasing trend of embedding positioning capabilities (e.g., GPS) in mobile devices facilitates the widespread use of Location Based Services. For such applications to succeed, privacy and confidentiality are key issues. Over all privacy will have to be managed through a combination of technology, legislation, corporate policy, and social norms. There are many data privacy schemes including Zero Knowledge Proof (ZKP) those can be used for location privacy. ZKP is also useful for removing the bottleneck problem introduced by use of trusted third party. In the earlier work, authors proposed the concept of middleware architecture in which, request and response are not routed through middleware for every transaction. This has reduced the dependency on middleware. This paper presents correspondence between the authentication techniques used in above said architecture and zero knowledge proof technique. Use of the concept of zero knowledge proof for authentication and authorization in the domain of location based services is also explored.

The geographical areas that are considered to be popular and interesting are called Region of Interest (ROI). There are multiple sources that can be used for erecting the ROIs such as user trajectory, POI databases, internet news etc. A tourist spot, historical region, monuments, a forest reserve, city, state or country's administrative boundaries are considered as the ROI objects. The interesting facts of the regions can be used for on-the-spot infotainment (information + entertainment) while user is walking, driving or just sitting idle on the flight. One may want the information in different level of details based on the interests, travel direction, and speed. The best way to get the recent information is going through the internet news, discussion forums, and encyclopedias on the web. The challenge is that there is no universal database available that contains the region information along with the location based search capabilities. The maintaining the granularity of the information based on the size or the details of the region adds more challenges. The broader idea of our work is to use the existing LBS infrastructure to track the user and achieve other navigation objectives in the system. However, the freely available up-to-date internet infrastructure is used as the information source. Initially, the user location is determined by the conventional methods. In order to relate the location with the web content, we use the semantic labels associated to the underlying location. The semantic labels are fetched by using reverse geocoding that returns the local attractions, street city, state and country names etc. Then these labels are used to search the associated detailed content in the web. The region data is maintained in two forms i.e. locally populated database and the web databases. The locally populated database can be updated on the preconfigured time interval in order to avoid data staleness problem. We have used Wikipedia as the internet data source in our prototype. The further research is in progress extracting the ROI information from the POI databases with their spatial and non-spatial attributes.

A new approach for correcting multiple errors and detecting an additive overflow in the Residue Number System (RNS) is suggested. It works with the code whose redundancy is in the form of magnitude indices. The residue representation of a number with magnitude index is reconsidered. The RNS with magnitude index were first studied by Sasaki16and Rao15and then by Barsi and Maestrini5, 6. The range of a given RNS is divided into intervals of equal width and the magnitude of a number X is defined as a integer locating X into one of such intervals. We have proposed algorithm which detects and corrects multiple errors in residue number. The algorithms for special cases viz., single burst residue error and single residue error are also suggested. Some of the advantages are pointed out over the existing approaches.

This paper proposes a scheme for the representation and processing of fractions in a residue system. The scheme is based on a mixed radix representation of a fraction in a residue, system. The algorithms for basic arithmetic operations of addition, subtraction, and multiplication involving fractions are developed and are shown to provide some improvement over an existing method. Application of these algorithms to division of two integers in the residue system has been shown.

A necessary and sufficient condition for sign detection in Non-Redundant Residue Number System by reducing the information of a residue digit has been obtained. The function to reduce the information of a residue digit xpcorresponding to a modulus mphas been assumed to be periodic with the period length equation, where equation and equation. A sequential method for determining the sign of a number is shown to demonstrate the applicability of the results thus proved.

This paper is concerned with the algebraic sign detection of a number in the Symmetric Residue Number System, A new approach has been suggested which completely avoid the time consuming process of the Symmetric Mixed Radix Conversion (SMRC). An algorithm based on the above approach implementable in parallel for sign detection is also presented. The heriware representation of the above algorithm is shown. The time and hardware complexity required for the process have also bean computed.