Abstract
Accurate household profiles (e.g., house type, number of occupants) identification is the key to the successful implementation of behavioral demand response. Currently, supervised learning methods are widely adopted to identify household profiles using smart meter data. Such methods could achieve promising performance in the case of sufficient labeled data but show low accuracy if labeled data is insufficient or even unavailable. However, the acquisition of accurately labeled data (usually obtained by survey) is very difficult, costly, and time-consuming in practice due to various reasons such as privacy concerns. To this end, a semi-supervised learning approach is proposed in this paper to address the above issues. Firstly, 78 preliminary features reflecting the household profiles information are extracted from both time and frequency domain. Secondly, feature selection methods are introduced to select more relevant ones as the input of the identification model from the preliminary features. Thirdly, a transductive support vector machine method is adopted to learn the mapping relation between the input features and the output household profile identification results. Case study on an Irish dataset indicates that the proposed approach outperforms supervised learning methods when only limited labeled data is available. Furthermore, the impacts of different feature selection methods (i.e., Filter, Wrapper and Embedding methods) are also investigated, among which the wrapper method performs best, and the identification accuracy improves with the increase of data resolution.

Abstract
The increasing complexity and dynamism of business environments has led to a significant growth in the risks related to the management of supply chain relationships. Trust and visibility between supply chain partners have been increasingly considered paramount aspects to manage these relationships and reduce risks. This paper aims to analyze and discuss the role of trust and visibility in supply chains, considering the complexity of multi-tier supply chains and multi-aspects visibility. Two cases of the textile sector from Portugal have been studied. After the analysis of the level of visibility and trust, a set of recommendations is provided. © 2022, IFIP International Federation for Information Processing.

Abstract
This paper presents a data market aimed at trading energy forecasts data. The system architecture is built using blockchain as a service, allowing access to data streams and establishing a distributed settlement between stakeholders. Energy Forecasts data is presented as the commodity traded in the market, whose settlement is provided through the blockchain on the basis of the extracted value provided by market stakeholders. Our proposal allows market stakeholders to acquire energy forecasts and pay according to the data accuracy, solving the confidentiality problem of freely sharing data. A data quality reward is introduced, steering the compensation sent to market participants. The data market design is presented and an evaluation campaign is performed, showing that the data market produced functionally valid results in comparison with the results achieved with a central simulated approach. Moreover, results show that the data market architecture is able to scale.

Abstract

Abstract
Artificial Intelligence is a hot topic and Machine Learning is one of the most fluent approaches and practices. The problem with many AI models is that they can be useful for predicting but they are bad at explaining why they behave a certain way. In some contexts, the explanation may even be more important than the prediction itself, mainly in systems in which decisions are made based on their predictions. Therefore, it is increasingly necessary to provide a forecast accompanied by an explanation, when decisions are made automatically. This paper aims to contribute to the solution of problem based on human mobility research, or at least, to be a starting point for its solution.

Abstract
Distributional semantics based on neural approaches is a cornerstone of Natural Language Processing, with surprising connections to human meaning representation as well. Recent Transformer-based Language Models have proven capable of producing contextual word representations that reliably convey sense-specific information, simply as a product of self supervision. Prior work has shown that these contextual representations can be used to accurately represent large sense inventories as sense embeddings, to the extent that a distance-based solution to Word Sense Disambiguation (WSD) tasks outperforms models trained specifically for the task. Still, there remains much to understand on how to use these Neural Language Models (NLMs) to produce sense embeddings that can better harness each NLM's meaning representation abilities. In this work we introduce a more principled approach to leverage information from all layers of NLMs, informed by a probing analysis on 14 NLM variants. We also emphasize the versatility of these sense embeddings in contrast to task-specific models, applying them on several sense-related tasks, besides WSD, while demonstrating improved performance using our proposed approach over prior work focused on sense embeddings. Finally, we discuss unexpected findings regarding layer and model performance variations, and potential applications for downstream tasks.& nbsp;