Dynamic Diverse Summarisation in Heterogeneous Graph Streams: a Comparison between Thesaurus/Ontology-based and Embeddings-based Approaches

Nowadays, there is a lot of attention drawn in smart environments, like Smart Cities and Internet of Things. These environments generate data streams that could be represented as graphs, which can be analysed in real-time to satisfy user or application needs. The challenges involved in these environments, ranging from the dynamism, heterogeneity, continuity, and high-volume of these real-world graph streams create new requirements for graph processing algorithms. We propose a dynamic graph stream summarisation system with the use of embeddings that provides expressive graphs while ensuring high usability and limited resource usage. In this paper, we examine the performance comparison between our embeddings-based approach and an existing thesaurus/ontology-based approach (FACES) that we adapted in a dynamic environment with the use of windows and data fusion. Both approaches use conceptual clustering and top-k scoring that can result in expressive, dynamic graph summaries with limited resources. Evaluations show that sending top-k fused diverse summaries, results in 34% to 92% reduction of forwarded messages and redundancy-awareness with an F-score ranging from 0.80 to 0.95 depending on the k compared to sending all the available information without top-k scoring. Also, the summaries' quality follows the agreement of ideal summaries determined by human judges. The summarisation approaches come with the expense of reduced system performance. The thesaurus/ontology-based approach proved 6 times more latency-heavy and 3 times more memory-heavy compared to the most expensive embeddings-based approach while having lower throughput but provided slightly better quality summaries.

A Framework for Exploration and Visualization of SPARQL Endpoint Information

Widely accepted standards, such as the Resource Description Framework, have provided unified ways for data provision aiming to facilitate the exchange of information between machines. This information became of interest to a wider audience due to its volume and variety but the available formats are posing significant challenges to users with limited knowledge of the Semantic Web. The SPARQL query language alleviates this barrier by facilitating the exploration of this information and many data providers have created dedicated SPARQL endpoints for their data. Many efforts have been dedicated to the development of systems that will provide access and support the exploration of these endpoints in a semantically correct and user friendly way. The main challenge of such approaches is the diversity of the information contained in the endpoints, which renders holistic or schema specific solutions obsolete. We present here an integrated platform that supports the users to the querying, exploration and visualization of information contained in SPARQL endpoints. The platform handles each query result independently based only on its characteristics, offering an endpoint and data schema agnostic solution. This is achieved through a Decision Support System, developed based on a knowledge base containing information experimentally collected from many endpoints, that allows us to provide case-specific visualization strategies for SPARQL query results based exclusively on features extracted from the result.

The Leaf Function of Graphs Associated with Penrose Tilings

In graph theory, the question of fully leafed induced subtrees has recently been investigated by Blondin Massé et al in regular tilings of the Euclidian plane and 3-dimensional space. The function LG that gives the maximum number of leaves of an induced subtree of a graph $G$ of order $n$, for any $n\in \N$, is called leaf function. This article is a first attempt at studying this problem in non-regular tilings, more specifically Penrose tilings. We rely not only on geometric properties of Penrose tilings, that allow us to find an upper bound for the leaf function in these tilings, but also on their links to the Fibonacci word, which give us a lower bound. Our approach rely on a purely discrete representation of points in the tilings, thus preventing numerical errors and improving computation efficiency. Finally, we present a procedure to dynamically generate induced subtrees without having to generate the whole patch surrounding them.

Algorithms for Workflow Satisfiability Problem with User-independent Constraints

The Workflow Satisfiability Problem (WSP) is a problem of interest in access control of information security. In its simplest form, the problem coincides with the Constraint Satisfiability Problem, where the number of variables is usually much smaller than the number of values. Wang and Li (ACM Trans. Inf. Syst. Secur. 2010) were the first to study the WSP as a problem parameterized by the number of variables. Their paper initiated very fruitful research surveyed by Cohen, Crampton, Gutin and Wahlström (2017). In this paper, we overview more recent WSP algorithmic developments and discuss computational experiments performed on two new testbeds of WSP instances. These WSP instances are closer to real-world ones than those by Karapetyan et al. (JAIR 2019). One of the two testbeds is generated using a novel iterative approach for obtaining computationally hard WSP instances.