Papers by Sabrina Senatore
Towards semantic context-aware drones for aerial scenes understanding
2016 13th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), 2016
Visual object tracking with unmanned aerial vehicles (UAVs) plays a central role in the aerial su... more Visual object tracking with unmanned aerial vehicles (UAVs) plays a central role in the aerial surveillance. Reliable object detection depends on many factors such as large displacements, occlusions, image noise, illumination and pose changes or image blur that may compromise the object labeling. The paper presents a proposal for a hybrid solution that adds semantic information to the video tracking processing: along with the tracked objects, the scene is completely depicted by data from places, natural features, or in general Points of Interest (POIs). Each scene from a video sequence is semantically described by ontological statements which, by inference, support the object identification which often suffers from some weakness in the object tracking methods. The synergy between the tracking methods and semantic technologies seems to bridge the object labeling gap, enhance the understanding of the situation awareness, as well as critical alarming situations.
Study of the Convergence in Automatic Generation of Instance Level Constraints
Advances in Intelligent Systems and Computing, 2015
This work deepens in a methodology to generate Instance Level Constraints for Semi-supervised clu... more This work deepens in a methodology to generate Instance Level Constraints for Semi-supervised clustering by the study of the inherent nature of the data. The methodology executes a partitional clustering algorithm repetitively, so we study its behaviour according to the number of iterations of the clustering. In this scenario we propose three different stopping criteria to determine how many times the partitional clustering algorithm should be executed to obtain reliable instance level constraints. These criteria are experimentally tested under the document clustering problem.
Multi-grained wildfire damage estimation from satellite vegetative scenario by fuzzy decision tree
2022 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)
Applied Sciences
This work is devoted to the analysis of the background seismic noise acquired at the volcanoes (C... more This work is devoted to the analysis of the background seismic noise acquired at the volcanoes (Campi Flegrei caldera, Ischia island, and Vesuvius) belonging to the Neapolitan volcanic district (Italy), and at the Colima volcano (Mexico). Continuous seismic acquisition is a complex mixture of volcanic transients and persistent volcanic and/or hydrothermal tremor, anthropogenic/ambient noise, oceanic loading, and meteo-marine contributions. The analysis of the background noise in a stationary volcanic phase could facilitate the identification of relevant waveforms often masked by microseisms and ambient noise. To address this issue, our approach proposes a machine learning (ML) modeling to recognize the “fingerprint” of a specific volcano by analyzing the background seismic noise from the continuous seismic acquisition. Specifically, two ML models, namely multi-layer perceptrons and convolutional neural network were trained to recognize one volcano from another based on the acquisiti...
Similarity-based SLD resolution and its implementation in an extended Prolog system
10th IEEE International Conference on Fuzzy Systems. (Cat. No.01CH37297)
Abstract This work presents an extension of SLD resolution towards approximate reasoning. The pro... more Abstract This work presents an extension of SLD resolution towards approximate reasoning. The proposed refutation procedure over-comes failures in the unification process by exploiting Similarity relation defined between predicate and constant symbols. This enables to ...
Data-Information-Concept Continuum From a Text Mining Perspective
Encyclopedia of Bioinformatics and Computational Biology, 2019
The recent Web panorama reveals a tangible proliferation of “social” data, in form of posts, opin... more The recent Web panorama reveals a tangible proliferation of “social” data, in form of posts, opinions, feelings, experiences. Most of the available data is unstructured text, unsuitable to be processed by computers, especially due to ambiguity and vagueness of the natural language. Research developments highlight the difficulty in capturing semantics of terms, linguistic expressions, and sentences and their consequent representation as a finite concept. This article presents an open-minded overview of the Text Mining approaches, targeted at transforming unstructured textual data into explicit knowledge, with a special focus on the conceptualization, i.e., the concept identification by analysing syntactic and semantic relations among terms as well as the contextual surrounding information. Different knowledge granulation is described in a layered knowledge model, where the term, the information and the concept represent the basic knowledge granules that cover most Text Mining approaches, in an evolving knowledge continuum.
Sensing multi-agent system for anomaly detection on crop fields exploiting the phenological and historical context
2021 International Conference on Emerging Techniques in Computational Intelligence (ICETCI), 2021
IoT technology spread led to the development of smart solutions for Precision Agriculture, employ... more IoT technology spread led to the development of smart solutions for Precision Agriculture, employing multiple smart sensors to acquire and process data to support vegetation monitoring and crucial tasks such as seeding, irrigation, etc. to improve crop quality and production. However, the gathering of data from multiple devices requires a data integration stage that strictly depends on the context and features of the environment, including the type of environment, species, and phenology of the area considered. To this purpose, this paper introduces a multi-agent model that allows a swarm of IoT devices to perform environmental monitoring and anomaly detection on Regions of Interest (ROIs) accomplishing several tasks, including harmonization of spectral images taken from different sources, phenology data extraction about ROIs to build contextual knowledge over the ROIs and anomaly detection through vegetation index classification. A case study in a simulated real-time environment demonstrates the potential of the model to promptly alert humans about ROIs affected by critical vegetation, burned areas, and ROIs that can be at risk after critical events occurred in their surroundings.
Empowering UAV scene perception by semantic spatio-temporal features
2018 IEEE International Conference on Environmental Engineering (EE), 2018
The use of unmanned aerial vehicles (UAVs) is becoming a key asset in different application domai... more The use of unmanned aerial vehicles (UAVs) is becoming a key asset in different application domains: from the military to surveillance tasks; to filming and journalism to shipping and delivery; to disaster monitoring to rescue operation and healthcare. One of the most desirable UAV capabilities is a human-like scenario understanding, i.e., the object recognition and interactions with other objects and with the environment, through the scene evolution, in order to get a high-view scenario description. The paper presents a semantic-enhanced approach for UAV-based surveillance systems. The video analysis is extended and enriched with semantic high level data to provide a global view of the video scenes. Semantic Web technologies provide the expressive power to describe semantically scenes appearing in the videos. The synergy between the video tracking methods and the semantic web technologies provides a new high-level human-like interpretation of the scenario. The approach focuses on the event understanding at semantic level: it is coded as spatio-temporal relation which joins fixed or mobile objects, with respect to a given temporal sequence of video frames. The system is composed of two macro components: one devoted to the tracking activities, i.e., the object identification and classification, the other enriches tracking data semantically, where the ontology-based scenario model is the bridge between the components. A reasoning component applied to the semantic knowledge, extracted from the scenario, infers new statements that describe the detected events occurring in the video.
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021
Precision agriculture systems collect spectral images from satellites, from which vegetation indi... more Precision agriculture systems collect spectral images from satellites, from which vegetation indices (VIs) can be assessed to monitor vegetation and soil condition. It requires a near-daily data acquisition to perform robust crop monitoring and data analysis. Satellites provide a periodic data acquisition that need a further data integration using multiple satellite sources along with camera-equipped drones to achieve an accurate data collection on a selected area. Moreover, VIs are not enough for a proper vegetation evaluation of the monitored areas due to differences among cultivars, the phenological season in which the vegetation is evaluated, the latitude of the areas, etc. This article introduces a system model to detect anomalies regarding the vegetation and soil conditions according to the area phenology and the historical vegetation trends. The system collects spectral images of the regions of interest (ROIs) from satellites and drones, harmonized to calculate VIs and feeds a dataset of near-daily high-resolution integrated images. The harmonic analysis allows phenological data extraction about the ROIs, hence the territorial observation model (TOM) has been extended to represent phenological stages and build knowledge on the ROIs and their phenology that is stored on a triple store. The system selects the VI values, calculated during the learned growing seasons of the ROIs, and classifies them to detect vegetation anomalies affecting those ROIs. The collected knowledge can be used by end-users (e.g., agronomists, experts, etc.) to analyze the anomalies correlated to historical results and vegetation trends.
Exploiting a multi-device knowledge meshing to agent-based activity tracking
2020 IEEE Symposium Series on Computational Intelligence (SSCI), 2020
Nowadays, systems of systems, composed of multiple cooperative smart devices, reached popularity ... more Nowadays, systems of systems, composed of multiple cooperative smart devices, reached popularity in many areas, including surveillance, digital forensics, agriculture and more. The analysis of great amounts of data coming from different sources could be very time consuming for humans, so they require automatic tools that help them to monitor vast and complex environments to detect anomalous situations. To this purpose, this paper introduces an agent-based model to allow IoT systems to monitor outdoor environments and detect suspicious or critical situations. The agent-modeling allows the accomplishment and coordination of various tasks, including object detection and data collection achieved through tracking, environmental context detection by using frame classification and semantic segmentation, contextual knowledge generation and activity detection through ontology reasoning. Finally, the agents report to humans about what happened by estimating the scene criticality through a fuzzy controller. A case study shows the potential of the whole framework and experiences evaluate the skills of the framework for activity detection.
Customized Advertising in E-Commerce Services Provision
Journal of Internet Technology, 2008
While traditional marketing is product-focused, e-commerce market is specifically oriented to the... more While traditional marketing is product-focused, e-commerce market is specifically oriented to the customers. The challenge in e-commerce is to actually get the Internet user's interest and customize answers according to his requirements. Recommendations and hints in a form of banner ads or advertising messages not only provide the opportunity of selling products, but also help enterprises to target the right customers with the proper products or services. Starting from this key point, this paper introduces a web-centric system for e-commerce services, which offers ad-hoc advertisements to users during their navigation. In order to address these issues, the system analyzes the asset marketing of the web enterprises for connect advertising to the targeted customer groups. Through an agent-based architecture and fuzzy techniques, the system provides a simple support to an e-commerce market mediation.
A UAV-Driven Surveillance System to Support Rescue Intervention
In recent years, the intelligent surveillance systems have attracted many application domains, du... more In recent years, the intelligent surveillance systems have attracted many application domains, due to the increasing demand on security and safety. Unmanned Areal Vehicles (AUVs) represent the reliable, low-cost solution for mobile sensor node deployment, localization, and collection of measurements.
Emotional Concept Extraction Through Ontology-Enhanced Classification
Metadata and Semantic Research, 2019
Capturing emotions affecting human behavior in social media bears strategic importance in many de... more Capturing emotions affecting human behavior in social media bears strategic importance in many decision-making fields, such as business and public policy, health care, and financial services, or just social events. This paper introduces an emotion-based classification model to analyze the human behavior in reaction to some event described by a tweet trend. From tweets analysis, the model extracts terms expressing emotions, and then, it builds a topological space of emotion-based concepts. These concepts enable the training of the multi-class SVM classifier to identify emotions expressed in the tweets. Classifier results are “softly” interpreted as a blending of several emotional nuances which thoroughly depicts people’s feeling. An ontology model captures the emotional concepts returned by classification, with respect to the tweet trends. The associated knowledge base provides human behavior analysis, in response to an event, by a tweet trend, by SPARQL queries.
2002 IEEE World Congress on Computational Intelligence. 2002 IEEE International Conference on Fuzzy Systems. FUZZ-IEEE'02. Proceedings (Cat. No.02CH37291)
The rapid growth of Web resources makes very difficult the task of Web search engines. Neverthele... more The rapid growth of Web resources makes very difficult the task of Web search engines. Nevertheless powerful search crawlers have been developed to aid in locating unfamiliar document (by means o f category, contents or subjeet based approaches), often queries r e t u r n inconsistent results. The main lack o f Web searching i s in the deduction capability: nowadays Web searching put much attention in matching user's queries that are too weak to cope with the user's expressiveness. First attempts in extending searching towards deduction capability are essentially based on two-valued logic and standard probability theory. The complexity of the problem (8.4 million o f Web sites), the features of the space domain (unstructured data, immature standards) demand a strong deviation from this trend. This work presents some results stemmed from a research projects where different technologies (in particular mobile agents and approximate reasoning) have been merged into an operational architecture suitable for Web searchinweb discovering. This paper discusses a different approach to Web searching where the input to the retrieval process i s described through a Web page. The system reacts to this kind of query by returning a set o f Web pages that reflect a similar context and deals with related arguments.
2020 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), 2020
Precision agriculture employs IoT devices to smartly monitoring plant vegetation and support food... more Precision agriculture employs IoT devices to smartly monitoring plant vegetation and support food production. Precision agriculture is highly required to improve product quality and better suit the requirements of the market. Among the IoT devices, Unmanned Aerial Vehicles (UAVs), can be equipped with many sensors that allow precise assessments of plant stress by flying over the plots. Notwithstanding the great benefits introduced, IoT devices may suffer from some issues. Many devices provide data in different formats on the same task, therefore they need solutions to integrate data and support a more thorough crop monitoring. This paper introduces a multitier architecture to deal with IoT-based intelligent monitoring, as well as an implementation of the architecture through multiagent modeling of the IoT devices for precision agriculture. The introduced model allows data acquisition from various sources (i.e., IoT devices), an ontology-based integration of data provided by the devices and a knowledge integration process to deal with domain-specific applications.
IEEE Systems Journal, 2019
Unmanned vehicle systems are often teleoperated, semiautonomous, and strictly dependent on human ... more Unmanned vehicle systems are often teleoperated, semiautonomous, and strictly dependent on human operators. In complex and dynamic environments, unmanned vehicles should be autonomous in a stricter sense, which means they should exhibit a human-like behavior to be capable of accurately perceiving the environment; understanding the situation, locating and interacting with environmental elements; and reporting solutions to humans. In order to address these desiderata, a modeling of a proactive, context-aware unmanned system is presented. Precisely, the system framework is designed for an unmanned aerial vehicle (UAV) that flies over an area, and collects data in the form of video frames, sensor values, etc. It recognizes situations, senses scene object and environment data, acquires the awareness about the evolving scenes, and, finally, takes a decision based on the perception of the overall scenario. The system design is based on two primary building blocks: 1) the semantic web technologies that provide the high-level object description in the tracked scenario, and 2) the fuzzy cognitive map model that provides the cognitive accumulation of spatial knowledge in order to discern specific situations that need a decision. Although the paper presents a UAV-based surveillance system model, its applicability is shown based on a realistic case study (viz., broken car on the highway); moreover, several possible scenario configurations have been simulated to assess the criticality level perceived by the system (UAV) in a given situation and to validate the effective response/decision in the case of critical situations.
IEEE Access, 2019
Video scene understanding is leading to an increased research investment in developing artificial... more Video scene understanding is leading to an increased research investment in developing artificial intelligence technologies, pattern recognition, and computer vision, especially with the advance in sensor technologies. Developing autonomous unmanned vehicles, able to recognize not just targets appearing in a scene but a complete scene the targets are involved in (describing events, actions, situations, etc.) is becoming crucial in the recent advanced intelligent surveillance systems. At the same time, besides these consolidated technologies, the Semantic Web Technologies are also emerging, yielding seamless support to the high-level understanding of the scenes. To this purpose, the paper proposes a systematic ontology modeling to support and improve video content analysis, by generating a comprehensive highlevel scene description, achieved by semantic reasoning and querying. The ontology schema comes from as an integration of new and existing ontologies and provides some design pattern guideline to get a highlevel description of a whole scenario. It starts from the description of basic targets in the video scenario, thanks to the support of video tracking algorithms and target classification; then provides a higher level interpretation, compounding event-driven target interactions (for local activity comprehension), to reach gradually an abstraction high level that enables a concise and complete scenario description.
Traditional search engines rely on keyword-based matching, recovering the documents which present... more Traditional search engines rely on keyword-based matching, recovering the documents which present some occurrences of the input keywords, but ignore at all the data meaning of the retrieved documents. Thus, long lists of pages links are returned but actually only a handful of pages contain reference to relevant web resources and meet the needs of users. The exigency of major awareness in the interpretation of web data yields new approaches and methodologies for improving the web search and retrieval, by taking into account the context of information, related to the user query. This work presents an approach for supporting the user in the Web search activity: it achieves the interpretation of the input query and, on the basis of the the local knowledge, replies by providing (links of) web pages which are more relevant to the content meaning of the input query. The approach combines intrinsic potential of the agent-based paradigm with the modeling of knowledge through techniques of soft computing. The agents encode the semantics of data, by exploiting ontologies, in order to grasp the actual query meaning. The information elicited by the query interpretation represents an add-on, aimed at augmenting the system knowledge, exploited in the discovery of web pages which match the user request.
Capturing Digest Emotions by Means of Fuzzy Linguistic Aggregation
Studies in Computational Intelligence, 2016
Distilling sentiments and moods hidden in the written (natural) language is a challenging issue w... more Distilling sentiments and moods hidden in the written (natural) language is a challenging issue which attracts research and commercial interests, aimed at studying the users behavior on the Web and evaluating the public attitudes towards brands, social events, political actions. The understanding of the written language is a very complicated task: sentiments and opinions are concealed in the sentences, typically associated to adjectives and verbs; then the intrinsic meaning of some textual expressions is not amenable to rigid linguistic patterns. This work presents a framework for detecting sentiment and emotion from text. It exploits an affective model known as Hourglass of Emotions, a variant of Plutchik’s wheel of emotions. The model defines four affective dimensions, each one with some activation levels, called ‘sentic levels’ that represent an emotional state of mind and can be more or less intense, depending on where they are placed in the corresponding dimension. Our approach draws from the Computational Intelligence area to provide a conceptual setting to sentiment and emotion detection and processing. The novelty is the fuzzy linguistic modeling of the Hourglass of Emotions: dimensions are modeled as fuzzy linguistic variables, whose linguistic terms are the sentic levels (emotions). This linguistic modeling naturally enables the use of fuzzy linguistic aggregation operators (from Computing with Words paradigm), such as LOWA (Linguistic Ordered Weighted Averaging) that inherently accomplishes an aggregation of the emotions in order to get an emotional expression that synthesizes a set of emotions associated with different sentic levels and activation intensities. The whole process for the emotion detection and synthesis is described through its main tasks, from the text parsing up to emotions extraction, returning a predominant emotion, associated with each dimension of the Hourglass of Emotions. An ad-hoc ontology has been designed to integrate lexical information and relations, along with the Hourglass model.
Intelligent Decision Technologies, 2007
The continuing growth of the Internet contents makes difficult the information access, inducing t... more The continuing growth of the Internet contents makes difficult the information access, inducing the task of information retrieval highly critical. The search engines often return a huge quantity of Web data, which is irrelevant to the input query. The emergency of personalization in the web search activities demands stable synergies for retrieving relevant information which meets the user needs. This work proposes an agent-based system for supporting customized Web searches. The system replies to a typical web query providing ad-hoc user-profiled links to web pages. the basis of a learning activity, which constitute an initial knowledge The agents collect locally the knowledge during an initial user querying/answering interaction phase and then interpret the meaning of collected information, by exploiting ontologies: they discover new semantic correlations among query terms, in order to refine the description of queries. These queries are used in the web search to provide more relevant replies, which reflect the user preferences and interest. This proposal represents a valid support for evidence-based applications, where sensitive contexts such as health care, medicine require high quality and unambiguous information in the specialized lexical domain.
Uploads
Papers by Sabrina Senatore