Agostino Forestiero

A single pass algorithm for clustering evolving data streams based on swarm intelligence

Existing density-based data stream clustering algorithms use a two-phase scheme approach consisting of an online phase, in which raw data is processed to gather summary statistics, and an offline phase that generates the clusters by using the summary data. In this article we propose a data stream clustering method based on a multi-agent system that uses a decentralized bottom-up self-organizing strategy to group similar data points. Data points are associated with agents and deployed onto a 2D space, to work simultaneously by applying a heuristic strategy based on a bio-inspired model, known as flocking model. Agents move onto the space for a fixed time and, when they encounter other agents into a predefined visibility range, they can decide to form a flock if they are similar. Flocks can join to form swarms of similar groups. This strategy allows to merge the two phases of density-based approaches and thus to avoid the computing demanding offline cluster computation, since a swarm represents a cluster. Experimental results show that the bio-inspired approach can obtain very good results on real and synthetic data sets.

An adaptive flocking algorithm for performing approximate clustering

This paper presents an approach based on an adaptive bio-inspired method to make state of the art clustering algorithms scalable and to provide them with an any-time behavior. The method is based on the biology-inspired paradigm of a flock of birds, i.e. a population of simple agents interacting locally with each other and with the environment. The flocking algorithm provides a model of decentralized adaptive organization useful to solve complex optimization, classification and distributed control problems. This approach avoids the sequential search of canonical clustering algorithms and permits a scalable implementation.The method is applied to design two novel clustering algorithms based on the main principles of two popular clustering algorithms: DBSCAN and SNN. This apporach can identify clusters of widely varying shapes and densities and is able to extract an approximate view of the clusters whenever it is required. Both the algorithms have been evaluated on synthetic and real world data sets and the impact of the flocking strategy on performance has been evaluated.

A grid portal for solving geoscience problems using distributed knowledge discovery services

This paper describes our research effort to employ Grid technologies to enable the development of geoscience applications by integrating workflow technologies with data mining resources and a portal framework in unique work environment called MOSE. Using MOSE, a user can easily compose and execute geo-workflows for analyzing and managing natural disasters such as landslides, earthquakes, floods, wildfires, etc. MOSE is designed to be applicable both for the implementation of response strategies when emergencies occur and for disaster prevention. It takes advantage of the standardized resource access and workflow support for loosely coupled software components provided by web/grid services technologies. The integration of workflows with data mining services significantly improves data analysis. Geospatial data management and mining are critical areas of modern-day geosciences research. An important challenge for geospatial information mining is the distributed nature of the data. MOSE provides knowledge discovery services based on the WEKA data mining library and novel distributed data mining algorithms for spatial data analysis. A P2P bio-inspired algorithm for distributed spatial clustering as an example of distributed knowledge discovery service for intensive data analysis is presented. A real case application for the analysis of landslide hazard areas in the Campania Region near the Sarno area shows the advantages of using the portal.

Self-organizing anomaly detection in data streams

Many distributed systems continuously gather, produce and elaborate data, often as data streams that can change over time. Discovering anomalous data is fundamental to obtain critical and actionable information such as intrusions, faults, and system failures. This paper proposes a multi-agent algorithm to detect anomalies in distributed data streams. As data items arrive from whatever sources, they are associated with bio-inspired agents and randomly disseminated onto a virtual space. The loaded agents move on the virtual space in order to form a group following the flocking algorithm. The agents group on the basis of a predefined concept of similarity of their associated objects. Only the agents associated to similar objects form a flock, whereas the agents associated with objects dissimilar to each other do not group in flocks. Anomalies are objects associated with isolated agents or objects associated with agents belonging to flocks having a few number of elements. Swarm intelligence features of the approach, such as adaptivity, parallelism, asynchronism, and decentralization, make the algorithm scalable to very large data sets and very large distributed systems. Experimental results for real and synthetic datasets confirm the validity of the proposed model.

Twitter-based Influenza Surveillance: An Analysis of the 2016-2017 and 2017-2018 Seasons in Italy

Influenza surveillance through social media data is becoming an important research topic because it could enhance the capabilities of official surveillance systems in monitoring the outbreak of seasonal flu, by providing healthcare organization with improved situational awareness. In this paper, the two influenza seasons 2016-2017 and 2017-2018, restricted to Italy, are investigated by analyzing the tweets posted by users regarding influenza-like illness. Two types of analysis are performed. The first studies the correlation between the tweets containing the most frequent flu related words with the data provided by the Italian InfluNet surveillance system. The second one examines the sentiment of people on the medicines used to heal flu. We show that there is a strict correlation between the reports published on the InfluNet system, and the contents posted by Twitter users about their symptoms and health state. Moreover, we found that the sentiment expressed by people regarding the treatment, in terms of medicines, taken to heal seems rather negative.

Bio-inspired algorithm for outliers detection

An essential activity to obtain valuable information to identify, for example, intrusions, faults, system failures, etc, is outliers detection. This paper proposes a bio-inspired algorithm able to detect anomaly data in distributed systems. Each data object is associated with a mobile agent that follows the well-known bio-inspired algorithm of flocking. The agents are randomly disseminated onto a virtual space where they move autonomously in order to form one or more flocks. Through a tailored similarity function, the agents associated with similar objects join in the same flock, whereas, the agents associated with dissimilar objects do not join in any flock. The objects associated with isolated agents or associated with agents grouped into flock with a number of entities lower than a given threshold, represent the outliers. Experimental results on synthetic and real data sets confirm the validity of the approach.

A Smart Discovery Service in Internet of Things Using Swarm Intelligence

The Internet of Things (IoT) has brought to a significant growing of data produced, and therefore, new models and approaches are needed to investigate these “big data” in terms of volume, velocity and variability. IoT services can be considered a dynamic content, including data sources and middleware infrastructures. An effective solution to manage dynamic contents are Content Delivery Networks (CDNs), but, in dynamic and large systems as IoT environment, their limits emerge, therefore, decentralized approaches and algorithms have to be designed and employed. This paper proposes SmartFinder, a swarm based algorithm to build a CDN based discovery service in pervasive and dynamic environment as IoT. The CDN servers are represented with metadata obtained through a locality preserving hash function. A swarm of mobile agents move the metadata and, by applying of tailored probability functions, achieve a logical organization of the servers. The outcome is a sorted overlay network that allows content and services discovery operations faster. Experimental results show the effectiveness of the approach.

A Multi-agent Approach for Intrusion Detection in Distributed Systems

Detecting anomalous data is essential to obtain critical and actionable information such as intrusions, faults, and system failures. In this paper an agent-based clustering algorithm to detect anomalies in a distributed system, is introduced. Each data object, independently of which source it arrives, is associated with a mobile agent following the flocking algorithm, a self-organizing bio-inspired computational model. The agents are randomly disseminated onto a virtual space where they move in order to form a flock. Thanks to a tailored similarity function the agents that are associated with similar objects form a flock, whereas the agents that are associated with objects dissimilar (outliers/anomalies) to each other do not group in flocks. Preliminarily experimental results confirm the validity of the proposed approach.

AIRS: Ant-Inspired Recommendation System

The goal of recommendation systems is to produce a set of meaningful suggestions for a group of users that can be useful for them. This paper introduces a multi-agent algorithm that builds a distributed recommendation system by exploiting nature-inspired techniques. The recommendable resources are recognized through a metadata represented of a bit string obtained by the application of a locality preserving hash function that maps similar resources into similar strings. Each agent works independently to replicate and wisely relocate the metadata. The agent operations are led by the application of ad-hoc probability functions. The outcome of this collective work will be a sorted logical overlay network that allows a fast recommendation service. Experimental analysis shows how the logical reorganization of metadata achieved by the agents can improve the performances of the recommendation system.

Distributed Anytime Clustering Using Biologically Inspired Systems

In this paper, we propose a biologically-inspired algorithm for clustering distributed data in a peer-to-peer network with a small world topology. The method proposed is based on a set of locally executable flocking algorithms that use a decentralized approach to discover clusters by an adaptive nearest-neighbor non-hierarchical approach and the execution, among the peers, of an iterative self-labeling strategy to generate global labels with which identify the clusters of all peers. We have measured the goodness of our flocking search strategy on performance in terms of accuracy and scalability. Furthermore, we evaluated the impact of small world topology in terms of reduction of iterations and messages exchanged to merge clusters.