ABSTRACT A competency‐based approach supported by personalized learning paths and prompt feedback accelerates skill development by continuously adapting to learners' needs and maintaining high levels of engagement. Capturing and understanding learner competency development through interaction data offers the potential for early intervention and optimized educational design, yet introduces challenges related to scalability and complexity. We present TracePath , a novel graph‐based framework that models learner trajectories as directed graphs, where nodes correspond to competencies or learner states and edges denote transitions such as validation or rejection events. This approach uncovers common learning pathways, identifies bottlenecks, and supports predictive analytics. At the core, a generic metamodel formalizes Competency Transition Graphs (CTGs), enabling comprehensive graph‐based analytics implemented over a hybrid polystore architecture that integrates both relational and NoSQL databases. Our design decouples data extraction from graph exploration, allowing efficient querying, clustering, and pattern matching to deliver timely and explainable learning insights. Empirical validation using real‐world data from the écri+ e‐certification project demonstrates TracePath's effectiveness in providing scalable, dynamic, and low‐latency learning analytics to support personalized education.

Computer-assisted learning is widely discussed in the literature to aid the comprehension of SQL queries (Structured Query Language) in higher education. However, it is difficult for educators/instructors to track, monitor and analyze students’ learning situation due to the higher education massification, and institutions with large classes. Consequently, we need to provide for educators a learning dashboard to monitor and analyze the digital traces issued from students during the practice learning in SQL course. We propose a system called LSQL (Learning Analytics for SQL) that is a solution based on the learning analytics’ methodology. To this end, we propose (i) learning environment dedicated to help students understand the syntax and logic of SQL and getting data issued from these students during online SQL lab work, (ii) trace model which is designed to more effectively represent and capture the complex interactions/actions carried by a student during practice learning activities in virtual/remote laboratories, and (iii) learning analytics dashboard for educators to visualize the statistics and metrics that represent the students’ behavior, and control the students progress in SQL skills to enhance the teaching activities. Tool support is fully available.

Cost-based optimization is a promising paradigm that relies on execution queries to enable fast and efficient ex- ecution reached by the database cost model (CM) during query processing/optimization. While a few database management systems (DBMS) already have support for mathematical CMs, developing such a CMs embedded or hard-coded for any DBMS remains a challenging and error-prone task. A generic interface must support a wide range of DBMS independently of the internal structure used for extending and modifying their signature; be efficient for good responsiveness. We propose a solution that provides a common set of parameters and cost primitives allowing intercepting the signature of the internal cost function and changing its internal parameters and configuration options. Therefore, the power of abstraction allows one to capture the designers/develop- ers intent at a higher level of abstraction and encode expert knowledge of domain-specific transformation in order to construct complex CMs, receiving quick feedback as they calibrate and alter the specifications. Our contribution relies on a generic CM interface supported by Model-Driven Engineering paradigm to create cost functions for database operations as intermediate specifications in which more optimization concerning the performance are delegated by our framework and that can be compiled and executed by the target DBMS. A proof-of-concept prototype is implemented by considering the CM that exists in PostgreSQL optimizer.

Summary Recently artificial intelligence techniques in the database community have become a driver for many database applications. The proposed solution adopting AI in the core database shows that incorporating AI improves the query processing and the self‐tuning of database systems. In traditional systems, self‐tuning database systems are commonly addressed with heuristics to suggest the physical structures (e.g., creation of indexes and materialized views) that enable the fastest execution of queries. However, existing designer tools do not explain/justify how the system behaves and the reasoning behind tuning activities. Moreover, these tools do not keep the database administrator (DBA) in the loop of the optimization process to trust some of the automatic tuning decisions. To address this problem, we introduce a framework called Explain‐Tun that enables to predict and explain self‐tuning actions with transparent strategy from historical data using two explicit models, that is, decision tree and random forests. First, we propose AI‐based DBMS to explain how to select physical structures and provide decision rules extracted by machine learning (ML) as a designed plug‐gable component. Second, a goal‐oriented model to keep DBA in the loop of the optimization process in order to manipulate ML models as CRUD entities. Finally, we evaluate our approach on three use cases, results show that bridging the DBA's experience and ML make sense in tuning database systems.

Abstract Designing a database cost model is one of the main research topics related to the physical design phase. It follows the evolution of database technology in order to evaluate and quantify the performance metrics (e.g., response time, energy consumption, etc.). Therefore, it makes the community researchers sensitive to the generated results. However, reusing and comparing database cost models require extracting related information manually from the research publications. This process is error‐prone and time‐consuming. Unfortunately, many researchers claim the difficulty of surveying and reproducing cost models already published in several/journal articles and/or reports. This difficulty is due to the absence of a process describing the cost model itself formally as well as the context of its utilization. This article presents an approach enabling the extraction of cost models information (context, parameters, features, etc.) as a set of orchestrated services. These services are implemented using natural language processing and machine‐learning techniques via a work‐flow pipeline inspired by DevOps practices. We illustrate our approach on a case study to stress the feasibility and benefits of our proposal by emphasizing the reproduction and automatization facilities.

Abstract A major challenge for many database management tasks including admission control, query scheduling, progress monitoring and self‐driving data storage systems is to enhance queries performances which are based on computational models known as database cost models. One of the most challenging aspects of developing accurate database cost models is identifying their parameters and capturing their relationships, consequently we can derive the query execution cost on the basis of a specific database hosted on a given platform. Furthermore, the highly dynamic workload (i.e., a set of queries) and the query execution variation lead to performance degradation risk, therefore cost models need to be improved by considering newer software configuration and future workload characteristics. In this article, we propose a framework called DeepCM that is based on a min–max optimization for building robust database cost model against uncertainty parameters. Furthermore, our framework is based on Robust Deep Neural Networks to build database cost models that guarantee a high accuracy regardless of variations from software configuration and workload characteristics. Several experiments have been done to evaluate the robustness of produced cost models and findings show that DeepCM provides a high cost model prediction accuracy and stable performance.

The development of database systems and applications requires the use of metrics to evaluate the quality and the efficiency of each phase, especially as regards the physical phase, where logical, physical and hardware optimizations are mainly used. Since the 1980s, a large range of cost models has been proposed. Each cost model is dedicated to the calculation of specific metrics and mainly pertains to specific target database, the workload, Database Management Systems (DBMS), deployment platforms, etc. Augmenting and improving reuse in complex systems is increasingly recognized as a crucial as it contributes to increasing the quality of the target systems, shortening engineering development time, and to paring down costs for the engineering of typically highly customer-specific solutions. In this paper, we propose a cost models language called CostDL that allows the description of metrics related to the most sensitive database system characteristics. An implementation of CostDL and its usage through a running example are also provided.

We are currently witnessing an explosion of advances in database technology, that cover all phases of database application design: non-functional requirements, conceptual modeling, logical modeling, deployment, physical design and exploitation. Researchers and engineers cooperate to integrate these advances in the database design. Their proposed solutions have to be confronted with similar studies whose results have been either published in scientific papers or on specific websites such as that of TPC (the Transaction Processing Council). Recently, several researchers have highlighted difficulties in reproducing the results of existing studies. As a consequence, certain research communities require that the environment and the results of the testing activities have to be published in order to facilitate their reproduction, to publish their simulator environments, in order to allow evaluators to perform by their own the experiments. This is quite similar to the Volkswagen fake pollution controls scandal. In this paper, we firstly, advocate the transparency of testing in the database field. Secondly, we propose the use of a repository dedicated to store testing environment and results. The environment includes used data sets, deployment platform, non-functional requirements, used algorithms, hypotheses, etc. and the results completed with their measurement units.