Search

Motivation: The scientific literature embeds an enormous amount of relational knowledge, encompassing interactions between biomedical entities, like proteins, drugs, and symptoms. To cope with the ever-increasing number of publications, researchers are experiencing a surge of interest in extracting valuable, structured, concise, and unambiguous information from plain texts. With the development of deep learning, the granularity of information extraction is evolving from entities and pairwise relations to events. Events can model complex interactions involving multiple participants having a specific semantic role, also handling nested and overlapping definitions. After being studied for years, automatic event extraction is on the road to significantly impact biology in a wide range of applications, from knowledge base enrichment to the formulation of new research hypotheses. Results: This paper provides a comprehensive and up-to-date survey on the link between event extraction and natural language understanding, focusing on the biomedical domain. First, we establish a flexible event definition, summarizing the terminological efforts conducted in various areas. Second, we present the event extraction task, the related challenges, and the available annotated corpora. Third, we deeply explore the most representative methods and present an analysis of the current state-of-the-art, accompanied by performance discussion. To help researchers navigate the avalanche of event extraction works, we provide a detailed taxonomy for classifying the contributions proposed by the community. Fourth, we compare solutions applied in biomedicine with those evaluated in other domains, identifying research opportunities and providing insights for strategies not yet explored. Finally, we discuss applications and our envisions about future perspectives, moving the needle on explainability and knowledge injection.

Graphical abstract, Today, we are fully immersed into the era of 3D biology. It has been extensively demonstrated that 3D models: (a) better mimic the physiology of human tissues; (b) can effectively replace animal models; (c) often provide more reliable results than 2D ones. Accordingly, anti-cancer drug screenings and toxicology studies based on multicellular 3D biological models, the so-called “-oids” (e.g. spheroids, tumoroids, organoids), are blooming in the literature. However, the complex nature of these systems limit the manual quantitative analyses of single cells’ behaviour in the culture. Accordingly, the demand for advanced software tools that are able to perform phenotypic analysis is fundamental. In this work, we describe the freely accessible tools that are currently available for biologists and researchers interested in analysing the effects of drugs/treatments on 3D multicellular -oids at a single-cell resolution level. In addition, using publicly available nuclear stained datasets we quantitatively compare the segmentation performance of 9 specific tools.

Though the strong evolution of knowledge learning models has characterized the last few years, the explanation of a phenomenon from text documents, called descriptive text mining, is still a difficult and poorly addressed problem. The need to work with unlabeled data, explainable approaches, unsupervised and domain independent solutions further increases the complexity of this task. Currently, existing techniques only partially solve the problem and have several limitations. In this paper, we propose a novel methodology of descriptive text mining, capable of offering accurate explanations in unsupervised settings and of quantifying the results based on their statistical significance. Considering the strong growth of patient communities on social platforms such as Facebook, we demonstrate the effectiveness of the contribution by taking the short social posts related to Esophageal Achalasia as a typical case study. Specifically, the methodology produces useful explanations about the experiences of patients and caregivers. Starting directly from the unlabeled patient’s posts, we derive correct scientific correlations among symptoms, drugs, treatments, foods and so on.

International audience; The paper presents the Olimpiadi di Problem Solving, a mild and inclusive competition aimed to promote computational thinking and general problem-solving in Italian schools. We describe motivation, teaching strategies behind the initiative, as well as its structure, organization and give some sample of the problems proposed to students. We also present some data that show the broad participation of Italian schools to the initiative, and a preliminary analysis of the results obtained by the students in the last 5 editions, which suggests that the competition fosters deep learning of computational thinking knowledge and skills.

A variety of technology enhanced teaching strategies and learning activities have been applied in education, including assessment mechanisms. In this paper, we aim to examine the extent to which peer assessment promotes deep learning and favours the incremental learning of the concepts presented throughout a traditional introductory programming course. In particular, we want to examine how the enactment of the peer-assessor and peer-assessee roles is associated with students’ learning improvements, after enacting recip-rocal peer assessment. We illustrate a case study based on a novel web-based peer assessment tool to improve engagement and learning outcomes in a pro-gramming course. We want to assess whether programming skills are growing through peer review and whether exposure to different programming tech-niques is helpful to students. One of the most remarkable findings of our ex-perience was that students reported that assessing others’ work was an ex-tremely valuable learning activity.

Route following based on visual landmark matching may require many models to cover all different situations. This paper describes a system that is able to adapt template's modelling parameters to environmental conditions (lighting, shadows, etc.) by a genetic learning technique. In addition, the mobile robot self-localisation is obtained by a stereo approach that uses the centres of matching in the two images to solve in a simple way the correspondence problem in the 3D position estimation. The experimental results show that the tracking robustness is improved, while using a small set of templates. © 1998 Elsevier Science B.V. All rights reserved.

The paper describes a prototypical system for optimal landmark acquisition and selection. Our landmark-learning approach does not require any type of environment model to be supplied to the robot in advance, and represents a step towards robots interacting with real environments. The approach is fitted and tested for the TMGA system (P. Zingaretti et al., 1998), which the authors developed for landmark tracking by adaptive, stereo template matching. Two complementary strategies, properly managed, are followed to construct a suitable subset of landmarks: the selection of the more discriminant landmarks and the selection of the landmarks that are more invariant in a neighbourhood. The robustness of the TMGA system in analysing the discriminant power of each landmark and the analysis of the disparity map and of the spatial activity maps of the stereo images are used for identifying discriminant and invariant landmarks. The experimental results show that the number of matching failures, and consequent landmark changes during the following of a route is comparable with (not much greater than) those obtained using an a-priori subset.

The rising auto usage deriving from growth in jobs and residential population is making traffic congestion less tolerable in urban and suburban areas. This results in air pollution, energy waste and unproductive and unpleasant consumption of people’s time. Public transport cannot be the only answer to this increasing transport demand. Car pooling has emerged to be a viable possibility for reducing private car usage in congested areas. Its actual practice requires a suitable information system support and, most important, the capability of effectively solving the underlying combinatorial optimization problem. This paper presents an application of the ANTS approach, one of the approaches which follow the Ant Colony Optimization (ACO) paradigm, to the car pooling optimization problem. Computational results are presented both on datasets derived from the literature about problems similar to car pooling and on real-world car pooling instances.

With the aim to individualise human-computer interaction, an Intelligent Tutoring System (ITS) has to keep track of what and how the student has learned. Hence, it is necessary to maintain a Student Model (SM) dealing with complex knowledge representation, such as incomplete and inconsistent knowledge and belief revision. With this in view, the main objective of this paper is to present and discuss the student modelling approach we have adopted to implement Pitagora 2.0, an ITS based on a co-operative learning model, and designed to support teaching-learning activities in a Euclidean Geometry context. In particular, this approach has led us to develop two distinct modules that cooperate to implement the SM of Pitagora 2.0. The first module resembles a "classical" student model, in the sense that it maintains a representation of the current student knowledge level, which can be used by the teacher in order to tune its teaching strategies to the specific student needs. In addition, our system contains a second module that implements a virtual partner, called companion. This module consists of a computational model of an "average student" which cooperates with the student during the learning process. The above mentioned module calls for the use of machine learning algorithms that allow the companion to improve in parallel with the real student. Computational results obtained when testing this module in simulation experiments are also presented. © 1995 Kluwer Academic Publishers.

The paper describes a novel comprehensive approach to image-contrast enhancement in the spatial domain. Instead of defining another transformation function our strategy consists of adopting a general functional form, able to map different transformation functions, and in using a learning technique to select the parameter values that are optimal for the image being processed. First, local measures of spatial activity are assigned to each pixel of the image. Second, the local contrast value for each pixel is computed according to a function which is based on human visual response. Third, the parameters of a comprehensive contrast-enhancement function are selected by a genetic algorithm on the basis of the spatial activity of the image resulting from the transformation. The validity of the proposed technique is confirmed both perceptually, that is, higher fitness values correspond to the images that have been judged better by human observers, and by comparative evaluations of our algorithm with respect to classical methods. © 1999 IEEE.

Unsupervised image segmentation is a complicated task due to the lack of appropriate measures for judging the quality of segmentation results. At present, most segmentation results are evaluated visually and qualitatively. In a previous paper [1], we presented a system that performs image segmentation without using neither ground-truth information nor human judgement. A genetic algorithm is used to efficiently search the set of parameters that maximises a segmentation quality criterion based on the integration of information from region boundaries and edge pixels. The performance of that system is degraded by the need of binarizing the edge maps, resulting in an information loss that may even hinder the convergence of the GA. In this paper we describe the solutions adopted to overcome this problem and to increase the objectiveness of segmentation results. Basically, human visual criteria are firstly embodied in an automatic image-contrast enhancement-process. The connectivity-code intermediate representation [2] and saliency-enhancing operations derived from the psychology field are then used to increase the significance of edge-pixels that have a small intensity but a high perceptual importance. Both computer simulated and real images have been used to test the performance of our adaptive segmentation system. The increased objectiveness of the segmentation results is confirmed both perceptually and by higher fitness values than those obtained by the previous system.

We describe an approach to image segmentation based on a two-layer module that is executed until a good segmentation is achieved, providing an evolution of previous segmentation results at each execution. The first layer performs a global segmentation of an image of decreasing area at each evolution by adopting a genetic algorithm learning technique to select segmentation parameters that give better results. The second layer provides the input to the next evolution by selecting the segmented regions that need further optimisation. A main goal of our system is to perform the segmentation without using neither ground-truth information nor human judgement. Thus, edge detection is performed to assess the performance of region segmentation and to guide the evolution of segmentation. Experimental results are consistent with what is observed visually.

Genetic Algorithms and Classifier Systems are often used in biologic-like and evolutionary behaviors’ simulations. The basic example is Wood7 Wilson’s world. In this environment it is interesting to study some problems: Can evolve the cooperative behaviors of organisms present in the world? How and when do the behaviors evolve? Some preliminary results show the conditions under that cooperative behavior rules are developing rapidly. Particularly we have pointed out the likely of following observations: a. The cooperative behavior develops more easily if the initial population start from the same point. b. It exists some thresholds under that the cooperative behavior can’t evolve; these thresholds depend to the population size.

The use of knowledge graphs (KGs) in advanced applications is constantly growing, as a consequence of their ability to model large collections of semantically interconnected data. The extraction of relational facts from plain text is currently one of the main approaches for the construction and expansion of KGs. In this paper, we introduce a novel unsupervised and automatic technique of KG learning from corpora of short unstructured and unlabeled texts. Our approach is unique in that it starts from raw textual data and comes to: i) identify a set of relevant domain-dependent terms; ii) extract aggregate and statistically significant semantic relationships between terms, documents and classes; iii) represent the accurate probabilistic knowledge as a KG; iv) extend and integrate the KG according to the Linked Open Data vision. The proposed solution is easily transferable to many domains and languages as long as the data are available. As a case study, we demonstrate how it is possible to automatically learn a KG representing the knowledge contained within the conversational messages shared on social networks such as Facebook by patients with rare diseases, and the impact this can have on creating resources aimed to capture the “voice of patients”.

A system for landmark tracking by a template matching approach is described. Route following based on landmarks may require many models to cover all different situations, so a genetic algorithm learning technique is used to adapt modelling parameters to environmental conditions (lighting, shadows, reflexes, etc.) during the tracking. In addition, the mobile robot self-localisation is obtained by a stereo approach that uses the centres of matching in the two images to solve in a simple way the correspondence analysis in the 3D position estimation. The experimental results show that the tracking robustness is improved when the adaptive template matching is used for landmark tracking.

The aim of this paper is to evaluate the student learning about Computer Science subjects. A questionnaire based on ordinal scored items has been submitted to the students through a computer automated system. The data collected have been analyzed by using a latent variable model for ordinal data within the Item Response Theory framework. The scores obtained from the model allow to classify the students according to the reached competence.

This paper describes the tracking of an object in a sequence of video-recorded images by a template matching approach based on a genetic algorithm learning technique to generate adaptive templates. The system is intended for real images. Contrary to a controlled environment, where the template may be generated from a noise free image and all successive frames do not show great changes, we hypothesise a set of images acquired at possible different environmental conditions. The varying environmental conditions greatly influence the matching, but they cannot be coded/used because they are not known a priori. In addition, 2D projections of 3D objects may vary during the sequence. Despite these difficulties, the need for robust systems that can locate objects in frame sequences is rapidly increasing due to the explosion in the amount of visual data in the communications systems.