Cheese-making is a complex process involving numerous stages, with multiple factors contributing and complex interactions occurring among the physicochemical elements involved. Understanding the process and optimizing its stages has attracted the attention of numerous investigations. In recent years, Machine Learning (ML) has established itself as one of the most advanced tools for data analysis and modeling thanks to its ability to capture complex and non-linear patterns. In the area of food science and engineering, these algorithms have started to be used as an alternative to more traditional statistical and mathematical prediction models. This paper explores the main research on ML applied to the study of cheese, from its production stages (i.e., fermentation or coagulation process) to the final product (i.e., detection of adulterations or food fraud). Particularly, we review 42 papers published between January 2014 and January 2025, with the aim of identifying common approaches. First, we present an explanation of the main concepts required to bring these approaches closer to researchers who are not experienced in applying ML. Then, we analyze the selected publications to detail the tasks of interest and the algorithms proposed to solve them. Finally, we detect gaps and opportunities to incorporate ML into future cheese research.

Social media and online platforms play an important role in assessing the impact of a natural disaster, especially in the immediate aftermath of the event. The information from social media demonstrated the potential to accelerate the response to a crisis. Locating social media images is critically important to help the victims immediately, mobilize community support, and provide enhanced situational awareness. However, finding the location of relevant social media images still challenges humans and computational processes. This study introduces a social media image classifier aimed at enhancing crowdsourced geolocation. The model is trained using data annotated by experts in disaster risk management. Its main goal is to reduce the difficulty of geolocating images by detecting those that are ``easy'' to geolocate. We combined the classifier with an ad-hoc crowdsourcing platform and tested it with pictures posted during a crisis. The experimental results indicate that the proposed approach speeds up the geolocation process of social media images while increasing the level of location precision

In emergency situations, social media platforms produce a vast amount of real-time data that holds immense value, particularly in the first 72 h following a disaster event. Despite previous efforts, efficiently determining the geographical location of images related to a new disaster remains an unresolved operational challenge. Currently, the state-of-the-art approach for dealing with these first response mapping is first filtering and then submitting the images to be geolocated to a volunteer crowd, assigning the images randomly to the volunteers. In this work, we extend our previous paper (Ballester et al., 2023) to explore the potential of artificial intelligence (AI) in aiding emergency responders and disaster relief organizations in geolocating social media images from a zone recently hit by a disaster. Our contributions include building two different models in which we try to (i) be able to learn volunteers’ error profiles and (ii) intelligently assign tasks to those volunteers who exhibit higher proficiency. Moreover, we present methods that outperform random allocation of tasks, analyze the effect on the models’ performance when varying numerous parameters, and show that for a given set of tasks and volunteers, we are able to process them with a significantly lower annotation budget, that is, we are able to make fewer volunteer solicitations without losing any quality on the final consensus.

Social media can serve as a valuable source of timely and valuable information about the impacts on people and infrastructure during a disaster. However, due to the lack of geographical data in most social media posts, this information is often underutilized by first responders. This paper proposes and analyses an approach that combines text from social media posts with textual information extracted from images to improve the geolocation of social media data during a given disaster. The implementation incorporates ChatGPT in location prediction. We use real-world dataset from Twitter that represent four different events, including floods and earthquake. The experimental results demonstrate that our proposal improves the location prediction’s quantity and precision. We expect that our findings help policymakers consider the application of the proposed methodology in disaster response.

Social media have the potential to provide timely information about emergency situations and sudden events. However, finding relevant information among the millions of posts being added every day can be difficult, and in current approaches developing an automatic data analysis project requires time and technical skills. This work presents a new approach for the analysis of social media posts, based on configurable automatic classification combined with Citizen Science methodologies. The process is facilitated by a set of flexible, automatic and open-source data processing tools called the Citizen Science Solution Kit. The kit provides a comprehensive set of tools that can be used and personalized in different situations, particularly during natural emergencies, starting from images and text contained in the posts. The tools can be employed by citizen scientists for filtering, classifying, and geolocating the content with a human-in-the-loop approach to support the data analyst, including feedback and suggestions on how to configure the automated tools, and techniques to gather inputs from citizens. Using flooding scenario as a guiding example, this paper illustrates the structure and functioning of the different tools proposed to support citizens scientists in their projects, and a methodological approach to their use. The process is then validated by discussing three case studies based on the Albania earthquake of 2019, the Covid-19 pandemic, and the Thailand floods of 2021. The results suggest that a flexible approach to tools composition and configuration can support a timely setup of an analysis project by citizen scientists, especially in case of emergencies in unexpected locations.

In the early stages of an emergency, information extracted from social media can support crisis response with evidence-based content. In order to capture this evidence, the events of interest must be first promptly detected. An automated detection system is able to activate other tasks, such as preemptive data processing for extracting event-related information. In this paper, we extend the human-in-the-loop approach in our previous work, TriggerCit, with a machine-learning-based event detection system trained on word count time series and coupled with an automated lexicon building algorithm. We design this framework in a language-agnostic fashion. In this way, the system can be deployed to any language without substantial effort. We evaluate the capacity of the proposed work against authoritative flood data for Nepal recorded over two years.

In the immediate aftermath of natural or man-made disasters, social media plays an essential role in assessing the impact of the event. The images from social media demonstrated the potential to accelerate the response to a crisis. However, finding the exact location of relevant social media images remains a problem for both humans and computer systems. This study presents an automated image classifier aimed at accelerating crowdsourced geolocation. The classifier is trained with data annotated by crisis risk experts and predicts the difficulty in geolocating a photo. The experimental results demonstrate that the proposed approach can predict the geolocating difficulty, thus potentially speed up the geolocation process by presenting volunteers images that are easy to geolocate.

Rapid impact assessment in the immediate aftermath of a natural disaster is essential to provide adequate information to international organisations, local authorities, and first responders. Social media can support emergency response with evidence-based content posted by citizens and organisations during ongoing events. In the paper, we propose TriggerCit: an early flood alerting tool with a multilanguage approach focused on timeliness and geolocation. The paper focuses on assessing the reliability of the approach as a triggering system, comparing it with alternative sources for alerts, and evaluating the quality and amount of complementary information gathered. Geolocated visual evidence extracted from Twitter by TriggerCit was analysed in two case studies on floods in Thailand and Nepal in 2021.

We address the problem of estimating a photo’s geographical location. Success in this estimation enables many impactful applications, like facilitating Disaster Management circumstances. However, this is also a very challenging task. Due to the complexity of the problem, we restrict the area of geolocation to a single city, treating geolocation as a classification problem where the districts of a city are the classes to be distinguished. In this paper, we exploit the Focal Modulation Network that is proven to perform effectively and efficiently in visual modeling for real-world applications. Experimental results on two diverse datasets, crawled from online sources, show the effectiveness of our approach. We can geolocate correctly more than two-thirds of test images from the larger dataset and about one-third from an experimental training dataset of a ten-times smaller size.

Every year more than 150 million people worldwide are affected by natural disasters. As declared by the United Nations Office for the Coordination of Humanitarian Affairs, "The first 72 h after a disaster are crucial; response must begin during that time to save lives". Social media has been demonstrated to be a potential data source to provide actionable data just as a disaster happens and develops, thus allowing emergency responders to better coordinate their activities. However, social media data also presents many challenges regarding data quality and geolocation (i.e., the geographical location of a post). Over the years, several technologies enabled the retrieval and processing of high volumes of data, with artificial intelligence often employed as a replacement for human intelligence for data classification tasks. Nevertheless, the need to deliver high-quality results within a critical response time is still a major challenge. In this tutorial, we will see how crowdsourcing assisted by artificial intelligence can make a significant contribution, especially where critical thinking and decision making are needed, in extracting valuable information from unconventional data sources. The tutorial will introduce the basics for extracting and analyzing information from social media, with a specific focus on retrieving images in an emergency after a natural disaster. We will provide the basics about social media crawling and analysis. A specific focus will be given to finegrained geolocalization of tweets and the combination of AI and crowdsourcing to filter relevant images and confirm or improve geolocations, which are needed to deliver high-quality information. Our experiences with social media analysis (e.g., [5]), geolocalization (e.g., [1, 6]), and crowdsourcing (e.g., [2, 3]) obtained in a recently concluded H2020 project E2mC (Evolution of Emergency Copernicus services) [4] and in the on-going H2020 project Crowd4SDG (Citizen Science for Monitoring Climate Impacts and Achieving Climate Resilience, https:// crowd4sdg.eu) will be illustrated. The objective of the tutorial is to provide an introduction and hands-on experience in some of the tools available in the field of emergency information systems. In particular, we focus on the tools that enable the search and analysis of social media posts, mainly on Twitter but also on other social media. The analysis of posts includes approaches for selecting relevant images based on image contents and text analysis techniques for information extraction. We also show how we could leverage citizen scientists by setting up a crowdsourcing environment, based on the PyBossa open-source platform (https://pybossa. com), and we demonstrate how we evaluate the quality of crowdsourcing results. We will also discuss the methods and processes for using such tools in a sudden emergency to gather different types of information to support first responders and decision makers. The tutorial is intended for participants who represent organizations looking for emergency data, who can benefit from collective intelligence, especially where there is a data gap in their research using traditional data sources, but are skeptical in reliability of this kind information; communities and agencies looking for tools to analyze the data; individuals interested in learning about available tools that can enrich and ensure reliability and usability of data obtained from social media.

Social media have the potential to provide timely information about emergency situations and sudden events. However, finding relevant information among millions of posts being posted every day can be difficult, and developing a data analysis project usually requires time and technical skills. This study presents an approach that provides flexible support for analyzing social media, particularly during emergencies. Different use cases in which social media analysis can be adopted are introduced, and the challenges of retrieving information from large sets of posts are discussed. The focus is on analyzing images and text contained in social media posts and a set of automatic data processing tools for filtering, classification, and geolocation of content with a human-in-the-loop approach to support the data analyst. Such support includes both feedback and suggestions to configure automated tools, and crowdsourcing to gather inputs from citizens. The results are validated by discussing three case studies developed within the Crowd4SDG H2020 European project.

Over the last decade, hundreds of thousands of volunteers have contributed to science by collecting or analyzing data. This public participation in science, also known as citizen science, has contributed to significant discoveries and led to publications in major scientific journals. However, little attention has been paid to data quality issues. In this work we argue that being able to determine the accuracy of data obtained by crowdsourcing is a fundamental question and we point out that, for many real-life scenarios, mathematical tools and processes for the evaluation of data quality are missing. We propose a probabilistic methodology for the evaluation of the accuracy of labeling data obtained by crowdsourcing in citizen science. The methodology builds on an abstract probabilistic graphical model formalism, which is shown to generalize some already existing label aggregation models. We show how to make practical use of the methodology through a comparison of data obtained from different citizen science communities analyzing the earthquake that took place in Albania in 2019.

This data set contains crowdsourced classification and damage assessment of images of an earthquake extracted from social media.A data set of 907 images posted on Twitter related to the 2019 Albanian Earthquake, that are filtered and pre-classified using an automated technique is cross-validated for accuracy by two different crowds. One, digital humanitarian volunteers using the crowdsourcing platform CROWD4EMS and another, paid micro-taskers of the Amazon Mechanical Turk. In order to compare and evaluate the efficiency and accuracy of the volunteers and the paid micro taskers, ground truth is established with the help of a team of experts, who validated the same set of data.Parameters considered for volunteer contributions: The dataset was imported to the Crowd4EMS platform for Crowd contribution. In the forum, each volunteer will see the image to be validated along with the tweet text and the link to the original tweet. The user has to validate whether the given image is relevant or irrelevant to the disaster. In case of doubt, the user can refer to the tutorial explaining the relevance or skip the task. Once the image's relevance is validated, the user will be asked to label the severity of the impact, as seen in the image. The Automated algorithm has pre-classified the images as severe and minimal damage. The Crowd4EMS platform lets the volunteer label them as 'severe damage,' moderate damage',' minimal damage', and' no damage'. Each task has to be answered at least three times, and the final consensus is taken as per the inter-rater agreement.Parameters considered for micro-taskers contribution: The dataset was imported to the Amazon Mechanical Turk platform for Crowd contribution. In the platform, each worker will see only the image that is to be categorised as follows: The user has to validate whether the given image depicts severe damage, moderate damage, minimal damage, no damage or irrelevant to the disaster. Each task has to be answered at least ten times, and the final consensus is taken as per the inter-rater agreement.Acknowledgements: We want to thank Muhammad Imran of Qatar Computing Research Institute for sharing their pre-filtered social media imagery dataset on the Albanian earthquake from the Artificial Intelligence for Disaster Response (AIDR) Platform. We would also like to extend our gratitude to the volunteers for their contribution on the Crowd4EMS Platform.

This work is about speeding up retrieval in Case-Based Reasoning (CBR) for large-scale case bases (CBs) comprised of temporally related cases in metric spaces. A typical example is a CB of electronic health records where consecutive sessions of a patient forms a sequence of related cases. k-Nearest Neighbors (kNN) search is a widely used algorithm in CBR retrieval. However, brute-force kNN is impossible for large CBs. As a contribution to efforts for speeding up kNN search, we introduce an anytime kNN search methodology and algorithm. Anytime Lazy kNN finds exact kNNs when allowed to run to completion with remarkable gain in execution time by avoiding unnecessary neighbor assessments. For applications where the gain in exact kNN search may not suffice, it can be interrupted earlier and it returns best-so-far kNNs together with a confidence value attached to each neighbor. We describe the algorithm and methodology to construct a probabilistic model that we use both to estimate confidence upon interruption and to automatize the interruption at desired confidence thresholds. We present the results of experiments conducted with publicly available datasets. The results show superior gains compared to brute-force search. We reach to an average gain of 87.18% with 0.98 confidence and to 96.84% with 0.70 confidence.

This work has been funded in part by the project PlayS and, by the project ´ Innobrain - New Technologies for the Innovation in Cognitive Stimulation and Rehabilitation, COMRDI-15-1-0017 (RIS3CAT comunitats); and the project SGR, 2017-SGR-172 from Generalitat de Catalunya; and FEDER funds.

dIn the Play&Sing project, we are developing an AI platform to support home-based self-training interventions for chronic stroke patients. A large percentage of patients suffering from this disease show motor deficits that clearly hinder their daily activities and diminish their quality of life. In this project we are proposing and testing a new Music Supported Therapy (MST) to induce upper limb motor recovery. With the help of a tablet-based application and a small musical keyboard, we are developing an AI platform to support home-based MST. Specifically, the role of AI algorithms is to support therapists and to boost user engagement by personalizing the interventions according to patient needs and preferences. AI algorithms will provide the therapists with hindsight and foresight tools. In the proposed MST, patients are performing 30 training sessions of 45 minutes with a frequency of 3 sessions per week. In this paper we present our platform and preliminary experiments conducted at a pilot phase.

Case-Based Reasoning (CBR) is a lazy learning method and, being such, when a new query is made to a CBR system, the swiftness of its retrieval phase proves to be very important for the overall system performance. The availability of ubiquitous data today is an opportunity for CBR systems as it implies more cases to reason with. Nevertheless, this availability also introduces a challenge for the CBR retrieval since distance calculations become computationally expensive. A good example of a domain where the case base is subject to substantial growth over time is the health records of patients where a query is typically an incremental update to prior cases. To deal with the retrieval performance challenge in such domains where cases are sequentially related, we introduce a novel method which significantly reduces the number of cases assessed in the search of exact nearest neighbors (NNs). In particular, when distance measures are metrics, they satisfy the triangle inequality and our method leverages this property to use it as a cutoff in NN search. Specifically, the retrieval is conducted in a lazy manner where only the cases that are true NN candidates for a query are evaluated. We demonstrate how a considerable number of unnecessary distance calculations is avoided in synthetically built domains which exhibit different problem feature characteristics and different cluster diversity.

When AI technologies are applied to real-world problems, it is often difficult for developers to anticipate all the knowledge needed. Previous research has shown that introspective reasoning can be a useful tool for helping to address this problem in case-based reasoning systems, by enabling them to augment their routine learning of cases with learning to make better use of their cases, as problem-solving experience reveals deficiencies in their reasoning process. In this paper we present a new introspective model for autonomously improving the performance of a CBR system by reasoning about system problem solving failures. We illustrate its benefits with experimental results from tests in an industrial design application.

Being able to predict the performance of a Case-Based Reasoning system against a set of future problems would provide invaluable information for design and maintenance of the system. Thus, we could carry out the needed design changes and maintenance tasks to improve future performance in a proactive fashion. This paper proposes a novel method for identifying regions in a case base where the system gives low confidence solutions to possible future problems. Experimentation is provided for RoboSoccer domain and we argue how encountered regions of dubiosity help us to analyse the case base and the reasoning mechanisms of the given Case-Based Reasoning system.