Accurate forecasting of electricity consumption is crucial for smart grids, enabling dynamic matching of supply and demand in both domestic and industrial use. This paper focuses on predicting electricity usage for individual households over a two-day horizon, a scenario that could facilitate flexible consumption adjustments based on grid conditions, showcasing methods that can also be applied in industrial applications. A significant challenge in this context is the presence of numerous peaks in consumption patterns, which are difficult to predict in terms of timing and magnitude. These peaks resemble outliers and introduce skewness into the target value distribution, ultimately degrading the performance of forecasting models. To address this issue, we empirically evaluate two approaches: example data weighting and square root transformation of target values. Our goal is to assess the effectiveness and limitations of these methods in mitigating the impact of skewed distributions on forecasting performance.

The concept of sustainable mobility is aimed at minimising environmental impacts of transportation systems while meeting the needs of individuals and communities. This includes encouraging citizens to choose sustainable modes of transportation: walking, cycling, public transport, carpooling, and telecommuting. We present an approach at rewarding organisations that actively support the sustainable mobility of their employees, and propose a framework for awarding a sustainable mobility certificate to organisations that fulfil sustainable mobility goals and objectives. The assessment is carried out using a qualitative rule-based multi-criteria model, which considers 50 indicators. Other elements of the certification process include methods for assessing the mobility structure of employees in the organisation and its potential for improvement. In this paper, we present the main components of the proposed certification framework and illustrate its application for assessing the status of sustainable mobility of employees at a Slovenian research institute.

This paper examines the evolution of environmental, social and governance (ESG) reporting by analysing a ten-year corpus of annual reports from FTSE 350 companies. Using BERTopic, an advanced topic modelling technique, we identify and subsequently cluster the most important ESG topics, providing significant insights into the reporting landscape. Our findings show how regulatory changes, such as the Non-Financial Reporting Directive, and major events like Covid-19, influence ESG topic prominence. The disclosure of ESG information is primarily determined by regulatory requirements. This is particularly evident in the fact that companies only disclose the diversity on the board, which is mandatory, but not the diversity and inclusion at other levels of the reporting organisation. Furthermore, our study examines the correlation between ESG scores and topic proportions, showing that extensive disclosure on topics like climate risk and stakeholder engagement is positively associated with higher ESG scores, whereas topics like executive remuneration show negative correlations. Our research contributes to the literature by offering a novel methodological approach to ESG analysis and provides insights into the gaps between reporting standards and practices, relevant to standard-setting bodies and regulators.

In this paper, we present an approach for the assessment of sustainable mobility of employees that is designed to be used in a unified way in different contexts. The approach is intended for assessment and self-assessment of organisations, for supporting decisions regarding sustainable mobility activities and eventually as the basis for awarding organisations with a sustainable mobility certificate. Its main contributions are tailored criteria and parameters for exposing sustainable mobility characteristics of an organisation in terms of current situation and future potential. The paper provides a detailed description and explanation of the methodology and an example of an application in practice. The results indicate that the approach is viable and operational and that the assessment results well represent the situation and provide clear indications of the challenges and the paths to improvement.

For assessing various performance indicators of companies, the focus is shifting from strictly financial (quantitative) publicly disclosed information to qualitative (textual) information. This textual data can provide valuable weak signals, for example through stylistic features, which can complement the quantitative data on financial performance or on Environmental, Social and Governance (ESG) criteria. In this work, we use various multi-task learning methods for financial text classification with the focus on financial sentiment, objectivity, forward-looking sentence prediction and ESG-content detection. We propose different methods to combine the information extracted from training jointly on different tasks; our best-performing method highlights the positive effect of explicitly adding auxiliary task predictions as features for the final target task during the multi-task training. Next, we use these classifiers to extract textual features from annual reports of FTSE350 companies and investigate the link between ESG quantitative scores and these features.

Curriculum learning, especially in robotics, is an active research field aiming to devise algorithms that speed up knowledge acquisition by proposing sequences of tasks an agent should train on. We focus on curriculum generation in reinforcement learning, where various meth-ods are currently compared based on the agent’s performance in terms of rewards on a predefined distribution of target tasks. We want to extend this singular characterization of existing algorithms by introducing metrics inspired by notions from the field of computational creativity. Namely, we introduce surprise, novelty, interest-ingness, and typicality that quantify various aspects of tasks stochastically proposed by the curriculum learning algorithms for the learner to train on. We model proposed tasks with Gaussian mixture models which enable their probabilistic interpretation, and use Hellinger distances between distributions and training rewards in formulation of the proposed metrics. Results are presented for eight curriculum learning algorithms show-casing differences in prioritization of various aspects of task creation and statistically different mean metric values when comparing agent’s best and worst training runs. The latter finding is not only useful for analysis of existing algorithms, but potentially also provides guidance for design of future curriculum learning methods.

The paper describes an approach for indirect data-based assessment and use of user preferences in an unobtrusive sensor-based coaching system with the aim of improving coaching effectiveness. The preference assessments are used to adapt the reasoning components of the coaching system in a way to better align with the preferences of its users. User preferences are learned based on data that describe user feedback as reported for different coaching messages that were received by the users. The preferences are not learned directly, but are assessed through a proxy—classifications or probabilities of positive feedback as assigned by a predictive machine learned model of user feedback. The motivation and aim of such an indirect approach is to allow for preference estimation without burdening the users with interactive preference elicitation processes. A brief description of the coaching setting is provided in the paper, before the approach for preference assessment is described and illustrated on a real-world example obtained during the testing of the coaching system with elderly users.

Forward-looking sentences are often a subject of studies of financial texts. Detection of such sentences is usually performed with wordlists of inclusive and exclusive keywords that are used as indicators of the forward-looking nature of the sentences at hand. In this paper we describe our assessment of potential improvements of forward-looking sentence detection wordlists by combining them together and by extending them with neighboring words in word-vector representations. Our current results indicate that simple combinations and straightforward extensions of wordlists with vector-space representation neighbors might not be suitable for FLS detection without further methodological improvements.

This paper presents tools and data sources collected and released by the EMBEDDIA project, supported by the European Union’s Horizon 2020 research and innovation program. The collected resources were offered to participants of a hackathon organized as part of the EACL Hackashop on News Media Content Analysis and Automated Report Generation in February 2021. The hackathon had six participating teams who addressed different challenges, either from the list of proposed challenges or their own news-industry-related tasks. This paper goes beyond the scope of the hackathon, as it brings together in a coherent and compact form most of the resources developed, collected and released by the EMBEDDIA project. Moreover, it constitutes a handy source for news media industry and researchers in the fields of Natural Language Processing and Social Science.

Computational creativity seeks to understand computational mechanisms that can be characterized as creative. The creation of new concepts is a central challenge for any creative system. In this article, we outline different approaches to computational concept creation and then review conceptual representations relevant to concept creation, and therefore to computational creativity. The conceptual representations are organized in accordance with two important perspectives on the distinctions between them. One distinction is between symbolic, spatial and connectionist representations. The other is between descriptive and procedural representations. Additionally, conceptual representations used in particular creative domains, such as language, music, image and emotion, are reviewed separately. For every representation reviewed, we cover the inference it affords, the computational means of building it, and its application in concept creation.

Social isolation is an important determinant of elderly people's health and well-being. Modern technologies could be a powerful ally in combating social isolation. However, instead of replacing human relationships they should help build them within a person's most natural social circles. This paper presents a framework for development of a technological coaching solution that can safeguard or even boost the everyday social life of the elderly. The modus operandi of this system spans from unobtrusive data collection through data processing and situation assessment of social behaviour, to selection and rendering of the most appropriate coaching actions to the elderly person including through members of their social circles. The assessment and decision-making process also integrates three important groups of external factors which influence the solution. These are: (i) personal profiles of the elderly and those members of the social circles who participate in the coaching process; (ii) objective external environment and (iii) the quality of the coaching actions. The latter is a crucial element of the system's learning abilities.

Computational creativity seeks to understand computational mechanisms that can be characterized as creative. The creation of new concepts is a central challenge for any creative system. In this paper, we outline different approaches to computational concept creation and then review conceptual representations relevant to concept creation, and therefore to computational creativity. The conceptual representations are organized in accordance with two important perspectives on the distinctions between them. One distinction is between symbolic, spatial and connectionist representations. The other is between descriptive and procedural representations. Additionally, conceptual representations used in particular creative domains, i.e. language, music, image and emotion, are reviewed separately. For every representation reviewed, we cover the inference it affords, the computational means of building it, and its application in concept creation.

This paper presents and critically discusses an approach for knowledge modelling and reasoning in a system for monitoring and coaching of senior adults. We present a modular architecture of the system and a detailed description of the modelling methodology which originates from the field of multi-criteria decision modelling and differs from the commonly used ones in this problem domain. The methodology has several characteristics that make it fit well to the purpose in this application and initial insights from potential users are positive. A discussion of the suitability of the proposed methodology for knowledge representation and reasoning in the given problem domain is provided, with an outline of its potential benefits and drawbacks and a comparison with the ontological approach.

In this study, we experimentally assess the potential of informal and unregulated communication to contribute to predictive models of financial markets indicators. The data sources that were analyzed are unregulated parts of yearly reports of the companies of the DOW30 index, text of tweets that mention these companies, data from financial statements, and stock market data about stock prices and volume. We conducted correlation analysis of descriptive and target features and an analysis of impacts of descriptive features to predictive power of models for regression and classification. The results indicate that overall the studied features only weakly describe the complex and noisy target phenomena and that also the linguistic features can contribute to phenomena models, particularly the features that represent expressions of sentiment, both in tweets and annual reports.

Churn prediction is the practice of assigning a probability to the event of a customer ending his contract with a service provider. Traditional data mining approaches to churn prediction in telecommunications industry are based on detecting patterns from customer contractual information, traffic related data, bills and payments, CRM data and customer service logs. The study presented in this paper has employed various machine learning approaches and assessed their performances using the data of a European mobile operator. The feature importance rankings which were used for feature selection yielded also some initial guidelines for acting on churn prevention in practice.

Computational creativity (CC) is a multidisciplinary research field, studying how to engineer software that exhibits behavior that would reasonably be deemed creative. This paper shows how composition of software solutions in this field can effectively be supported through a CC infrastructure that supports user-friendly development of CC software components and workflows, their sharing, execution, and reuse. The infrastructure allows CC researchers to build workflows that can be executed online and be easily reused by others through the workflow web address. Moreover, it enables the building of procedures composed of software developed by different researchers from different laboratories, leading to novel ways of software composition for computational purposes that were not expected in advance. This capability is illustrated on a workflow that implements a Concept Generator prototype based on the Conceptual Blending framework. The prototype consists of a composition of modules made available as web services, and is explored and tested through experiments involving blending of texts from different domains, blending of images, and poetry generation.

The aim of this work is to reproduce the approach to detecting semantic orientations in economic texts that was presented in the paper Good Debt or Bad Debt: Detecting Semantic Orientations in Economic Texts by Malo et al. The approach employs the Linearized Phrase Structure model for sentence level classification of short economic texts into a positive, negative or neutral category from investor’s perspective and yields state-of-the-art results. The proposed method employs both rule based linguistic models and machine learning. Where possible we follow the same approach as described in the original paper, with some documented modifications. Our solution is simplified in at least two aspects, but its performance is comparable to the original and overall remains better than the reported results of other benchmark algorithms mentioned in the original paper. The differences between the two models and results are described in detail and lead to conclusion that the original approach is to a large extent repeatable and that our simplified version does not overly sacrifice performance for generalizability.

Computational Creativity is a field of Articial Intelligence that addresses processes that would be deemed creative if performed by a human. The field has been very active since 1999, and is now an established research field with its own International Conference on Computational Creativity (ICCC) conference series founded in 2010. This paper brifley surveys the field of Computational Creativity (CC) that is based on the analysis of ICCC conference papers, followed by a more detailed presentation of projects and selected contributions of Slovenian researchers to the field.

We describe a novel slogan generator that employs bisociation in combination with the selection of stylistic literary devices. Advertising slogans are a key marketing tool for every company and a memorable slogan provides an advantage on the market. A good slogan is catchy and unique and projects the values of the company. To get an insight in construction of such slogans, we first analyze a large corpus of advertising slogans in terms of alliteration, assonance, consonance and rhyme. Then we develop an approach for constructing slogans that contain these stylistic devices which can help make the slogans easy to remember. At the same time, we use bisociation to imprint a unique message into the slogan by allowing the user to specify the original and bisociated domains from where the generator selects the words. These word sets are first expanded with the help of FastText embeddings and then used to fill in the empty slots in slogan skeletons generated from a database of existing slogans. We use a language model to increase semantical cohesion of generated slogans and a relevance evaluation system to score the slogans by their connectedness to the selected domains. The evaluation of generated slogans for two companies shows that even if slogan generation is a hard problem, we can find some generated slogans that are suitable for the use in production without any modification and a much larger number of slogans that are positively evaluated according to at least one criteria (e.g., humor, catchiness).

Numerous visual programming platforms support the generation, execution and reuse of constructed scientic workows. However, there has been little effort devoted to building creative software blending systems, capable of composing novel workows by autonomously combining individual software components or even entire workows originally designed for solving tasks in different research elds. Based on the review of relevant computational creativity research and of contemporary web platforms for workow construction, this paper denes the desired functionality of a software blending system. Considering the required autonomy of the system and the workow complexity limitations, we investigate the necessary conditions for the implementation of a creative blending system within the existing visual programming platforms.