Search

Natural language processing technology has made significant progress in recent years, fuelled by increasingly powerful general language models. This has also inspired a sizeable body of work targeted specifically towards the educational domain, where the creation of questions (both for assessment and practice) is a laborious/expensive effort. Thus, automatic Question-Generation (QG) solutions have been proposed and studied. Yet, according to a recent survey of the educational QG community’s progress, a common baseline dataset unifying multiple domains and question forms (e.g., multiple choice vs. fill-the-gap), including readily available baseline models to compare against, is largely missing. This is the gap we aim to fill with this paper. In particular, we introduce a high-quality dataset in the educational domain, containing over 3,000 entries, comprising (i) multiple-choice questions, (ii) the corresponding answers (including distractors), and (iii) associated passages from the course material used as sources for the questions. Each question is phrased in two forms, normal and cloze (i.e., fill-the-gap), and correct answers are linked to source documents with sentence-level annotations. Thus, our versatile dataset can be used for both question and distractor generation, as well as to explore new challenges such as question format conversion. Furthermore, 903 questions are accompanied by their cognitive complexity level as per Bloom’s taxonomy. All questions have been generated by educational experts rather than crowd workers to ensure they are maintaining educational and learning standards. Our analysis and experiments suggest distinguishable differences between our dataset and commonly used ones for question generation for educational purposes. We believe this new dataset can serve as a valuable resource for research and evaluation in the educational domain. The dataset and baselines are made available to support further research in question generation for education (https://github.com/hadifar/question-generation).

Electric vehicle (EV) charging stations have become prominent in electricity grids in the past few years. Their increased penetration introduces both challenges and opportunities; they contribute to increased load, but also offer flexibility potential, e.g., in deferring the load in time. To analyze such scenarios, realistic EV data are required, which are hard to come by. Therefore, in this article we define a synthetic data generator (SDG) for EV charging sessions based on a large real-world dataset. Arrival times of EVs are modeled assuming that the inter-arrival times of EVs follow an exponential distribution. Connection time for EVs is dependent on the arrival time of EV, and can be described using a conditional probability distribution. This distribution is estimated using Gaussian mixture models, and departure times can calculated by sampling connection times for EV arrivals from this distribution. Our SDG is based on a novel method for the temporal modeling of EV sessions, and jointly models the arrival and departure times of EVs for a large number of charging stations. Our SDG was trained using real-world EV sessions, and used to generate synthetic samples of session data, which were statistically indistinguishable from the real-world data. We provide both (i) source code to train SDG models from new data, and (ii) trained models that reflect real-world datasets.

The renewable energy transition has introduced new electricity tariff structures. With the increased penetration of photovoltaic and wind power systems, users are being charged more for their peak demand. Consequently, peak shaving has gained attention in recent years. In this paper, we investigated the potential of peak shaving through battery storage. The analyzed system comprises a battery, a load and the grid but no renewable energy sources. The study is based on 40 load profiles of low-voltage users, located in Belgium, for the period 1 January 2014, 00:00−31 December 2016, 23:45, at 15 min resolution, with peak demand pricing. For each user, we studied the peak load reduction achievable by batteries of varying energy capacities (kWh), ranging from 0.1 to 10 times the mean power (kW). The results show that for 75% of the users, the peak reduction stays below 44% when the battery capacity is 10 times the mean power. Furthermore, for 75% of the users the battery remains idle for at least 80% of the time; consequently, the battery could possibly provide other services as well if the peak occurrence is sufficiently predictable. From an economic perspective, peak shaving looks interesting for capacity invoiced end users in Belgium, under the current battery capex and electricity prices (without Time-of-Use (ToU) dependency).

Multiple-choice questions (MCQs) are widely used in digital learning systems, as they allow for automating the assessment process. However, owing to the increased digital literacy of students and the advent of social media platforms, MCQ tests are widely shared online, and teachers are continuously challenged to create new questions, which is an expensive and time-consuming task. A particularly sensitive aspect of MCQ creation is to devise relevant distractors, i.e., wrong answers that are not easily identifiable as being wrong. This article studies how a large existing set of manually created answers and distractors for questions over a variety of domains, subjects, and languages can be leveraged to help teachers in creating new MCQs, by the smart reuse of existing distractors. We built several data-driven models based on context-aware question and distractor representations and compared them with static feature-based models. The proposed models are evaluated with automated metrics and in a realistic user test with teachers. Both automatic and human evaluations indicate that context-aware models consistently outperform a static feature-based approach. For our best-performing context-aware model, on average, three distractors out of the ten shown to teachers were rated as high-quality distractors. We create a performance benchmark, and make it public, to enable comparison between different approaches and to introduce a more standardized evaluation of the task. The benchmark contains a test of 298 educational questions covering multiple subjects and languages and a 77k multilingual pool of distractor vocabulary for future research.