Abstract Dynamic networks are ubiquitous in many domains for modelling evolving graph-structured data and detecting changes allows us to understand the dynamic of the domain represented. A category of computational solutions is represented by the pattern-based change detectors (PBCDs), which are non-parametric unsupervised change detection methods based on observed changes in sets of frequent patterns over time. Patterns have the ability to depict the structural information of the sub-graphs, becoming a useful tool in the interpretation of the changes. Existing PBCDs often rely on exhaustive mining, which corresponds to the worst-case exponential time complexity, making this category of algorithms inefficient in practice. In fact, in such a case, the pattern mining process is even more time-consuming and inefficient due to the combinatorial explosion of the sub-graph pattern space caused by the inherent complexity of the graph structure. Non-exhaustive search strategies can represent a possible approach to this problem, also because not all the possible frequent patterns contribute to changes in the time-evolving data. In this paper, we investigate the viability of different heuristic approaches which prevent the complete exploration of the search space, by returning a concise set of sub-graph patterns (compared to the exhaustive case). The heuristics differ on the criterion used to select representative patterns. The results obtained on real-world and synthetic dynamic networks show that these solutions are effective, when mining patterns, and even more accurate when detecting changes.

Contract clause recognition is the process of discriminating legal clauses from ordinary sentences in contracts, and categorizing them accordingly. It is one of the most crucial tasks performed by human experts, typically lawyers from legal offices, when analyzing contracts, often for negotiation purposes. When manually executed, contract clause recognition is often time-consuming and error-prone. Traditional solutions based on machine learning leverage two-stepped approaches relying on anomalous contract clause detection and classification. Unfortunately, these approaches fail to deliver accurate results. In this paper, we propose a holistic approach, based on openset recognition algorithms, to the problem of contract clause recognition. Experimental results on benchmark data prove that the proposed solution is effective.

Large and medium-sized manufacturing companies are concerned with maintaining their supplier registries with well-reputed suppliers sourced over time. Every supplier periodically undergoes rigorous qualification processes where procurement officers assess, among other factors, the supplier document compliance status. To this end, procurement officers periodically ask suppliers, via digital e-procurement platforms, for qualification documents of different categories. Conversely, suppliers promptly answer by handing out such documents that can, maliciously or inadvertently, be wrong. When wrong qualification documents remain undetected, and the associated suppliers are qualified, a threat to the overall business arises: procurement officers may entrust purchase orders to not compliant suppliers with unpredictable performances. Our claim is that equipping e-procurement platforms with document recognition based on supervised open-set recognition (OSR) could mitigate the problem. In particular, we deem OSR solutions suitable for supplier qualification document recognition due to their simultaneous abilities of i) recognizing documents belonging to relevant categories and ii) rejecting those belonging to unknown categories. Quantitative and qualitative results from a real-world case study in partnership with an Italian manufacturing company show that the proposed solution is viable.