Abstract
The Inven!RA architecture is an approach for online tracking of progression towards learning objectives, from analytics of distributed learning activities, provided by multiple third parties. However, there are few examples on how to implement such third-party learning activities, known as Activity Provider modules. We followed the Inven!RA architecture interfacing specification to create and implement two sample learning activities: a technical documentation analysis activity and an Arduino microelectronics programming activity. Integration tests with an Inven!RA architecture prototype confirmed the adequacy of this implementation. Thus, these samples provide clarification on how to design and develop Inven!RA Activity Provider modules.;A arquitetura Inven!RA é uma abordagem para o acompanhamento online da evolução face a objetivos de aprendizagem, através de dados analíticos de atividades de aprendizagem distribuídas, proporcionadas por um leque variado de entidades externas. Como são escassos os exemplos de implementação destas atividades de aprendizagem externas, designadas por módulos de Prestadores de Atividades,seguimos a especificação de interfaces da arquitetura para criar e implementar dois exemplos de atividades de aprendizagem: uma atividade de análise de documentação técnica e uma de programação de microeletrónica com Arduino. Testes de integração com um protótipo da Inven!RA confirmaram a adequação destas implementações. Consequentemente, proporcionam clarificação quanto à forma de conceber e desenvolver módulos de Prestadores de Atividades para a arquitetura Inven!RA.

Abstract
Adaptive immune receptor repertoires (AIRR) are key targets for biomedical research as they record past and ongoing adaptive immune responses. The capacity of machine learning (ML) to identify complex discriminative sequence patterns renders it an ideal approach for AIRR-based diagnostic and therapeutic discovery. So far, widespread adoption of AIRR ML has been inhibited by a lack of reproducibility, transparency and interoperability. immuneML (immuneml.uio.no) addresses these concerns by implementing each step of the AIRR ML process in an extensible, open-source software ecosystem that is based on fully specified and shareable workflows. To facilitate widespread user adoption, immuneML is available as a command-line tool and through an intuitive Galaxy web interface, and extensive documentation of workflows is provided. We demonstrate the broad applicability of immuneML by (1) reproducing a large-scale study on immune state prediction, (2) developing, integrating and applying a novel deep learning method for antigen specificity prediction and (3) showcasing streamlined interpretability-focused benchmarking of AIRR ML.

Abstract
Index structures were often used to optimise fetch operations to external storage devices (secondary memory). Nowadays, this also holds for increasingly large amounts of data residing in main-memory (primary memory). Within this scope, this work focuses on index structures that efficiently insert, query and delete valid-time data from very large datasets. This work performs a comparative study on the performance of the Interval B+ tree (IB+ tree) and the Improved Interval B+ tree (I2B+ tree): a variant that improves the time-efficiency of the deletion operation by reducing the number of traversed nodes to access siblings. We performed an extensive analysis of the performance of two operations: insertions and deletions, on both index structures, using multiple datasets with growing volumes of data, distinct temporal distributions and tree parameters (time-split alpha and node order). Results confirm that the I2B+ tree globally outperforms the IB+ tree, since, on average, deletion operations are 7% faster, despite insertions requiring 2% more time. Furthermore, results also allowed to determine the key factors that augment the performance difference on deletions between both trees. Copyright © 2021 by Author/s and Licensed by Veritas Publications Ltd., UK.

Abstract
It has been a quarter of a century since the publication of the first edition of the IEEE International Conference on Computer Supported Cooperative Work in Design (CSCWD) held in 1996 in Beijing, China. Despite some attempts to empirically examine the evolution and identity of the field of CSCW and its related communities and disciplines, the scarcity of scientometric studies on the IEEE CSCWD research productivity is noteworthy. To fill this gap, this study reports on an exploratory quantitative analysis of the literature published in the IEEE CSCWD conference proceedings with the purpose of visualizing and understanding its structure and evolution for the 2001-2019 period. The findings offer valuable insights into the paper and author distribution, country and citation-level productivity indicators, degree of collaboration, and collaboration index. Through this analysis we also expect to get an initial overview of the IEEE CSCWD conference concerning the main topics being presented, most cited papers, and variances in the number of keywords, full-text views, and references.

Abstract
Despite decades of research and development in named entity resolution, dealing with name ambiguity is still a challenging issue for many bibliometric-enhanced information retrieval (IR) tasks. As new bibliographic datasets are created as a result of the upward growth of publication records worldwide, more problems arise when considering the effects of errors resulting from missing data fields, duplicate entities, misspellings, extra characters, etc. As these concerns tend to be of large-scale, both the general consistency and the quality of electronic data are largely affected. This paper presents an approach to handle these name ambiguity problems through the use of crowdsourcing as a complementary means to traditional unsupervised approaches. To this end, we present "AuthCrowd", a crowdsourcing system with the ability to decompose named entity disambiguation and entity matching tasks. Experimental results on a real-world dataset of publicly available papers published in peer-reviewed venues demonstrate the potential of our proposed approach for improving author name disambiguation. The findings further highlight the importance of adopting hybrid crowd-algorithm collaboration strategies, especially for handling complexity and quantifying bias when working with large amounts of data.

Abstract
In this paper, we present and discuss an innovative approach to solve Job Shop scheduling problems based on machine learning techniques. Traditionally, when choosing how to solve Job Shop scheduling problems, there are two main options: either use an efficient heuristic that provides a solution quickly, or use classic optimization approaches (e.g., metaheuristics) that take more time but will output better solutions, closer to their optimal value. In this work, we aim to create a novel architecture that incorporates reinforcement learning into scheduling systems in order to improve their overall performance and overcome the limitations that current approaches present. It is also intended to investigate the development of a learning environment for reinforcement learning agents to be able to solve the Job Shop scheduling problem. The reported experimental results and the conducted statistical analysis conclude about the benefits of using an intelligent agent created with reinforcement learning techniques. The main contribution of this work is proving that reinforcement learning has the potential to become the standard method whenever a solution is necessary quickly, since it solves any problem in very few seconds with high quality, approximate to the optimal methods.