2013
Conference article
Restricted
On multidimensional network measures
Magnani M, Monreale A, Rossetti G, Giannotti F
Networks, i.e., sets of interconnected entities, are ubiquitous, spanning disciplines as diverse as sociology, biology and computer sci- ence. The recent availability of large amounts of network data has thus provided a unique opportunity to develop models and analysis tools ap- plicable to a wide range of scenarios. However, real-world phenomena are often more complex than existing graph data models. One relevant ex- ample concerns the numerous types of social relationships (or edges) that can be present between individuals in a social network. In this short pa- per we present a uni ed model and a set of measures recently developed to represent and analyze network data with multiple types of edges.
Magnani M, Monreale A, Rossetti G, Giannotti F
Networks, i.e., sets of interconnected entities, are ubiquitous, spanning disciplines as diverse as sociology, biology and computer sci- ence. The recent availability of large amounts of network data has thus provided a unique opportunity to develop models and analysis tools ap- plicable to a wide range of scenarios. However, real-world phenomena are often more complex than existing graph data models. One relevant ex- ample concerns the numerous types of social relationships (or edges) that can be present between individuals in a social network. In this short pa- per we present a uni ed model and a set of measures recently developed to represent and analyze network data with multiple types of edges.
2013
Other
Restricted
"You Know Because I Know": a multidimensional network approach to human resources problem
Coscia M, Rossetti G, Pennacchioli D, Ceccarelli D, Giannotti F
Finding talents, often among the people already hired, is an endemic challenge for organizations. The social networking revolution, with online tools like Linkedin, made possible to make explicit and accessible what we perceived, but not used, for thousands of years: the exact position and ranking of a person in a network of professional and personal connections. To search and mine where and how an employee is positioned on a global skill network will enable organizations to find unpredictable sources of knowledge, innovation and know-how. This data richness and hidden knowledge demands for a multidimensional and multiskill approach to the network ranking problem. Multidimensional networks are networks with multiple kinds of relations. To the best of our knowledge, no network-based ranking algorithm is able to handle multidimensional networks and multiple rankings over multiple attributes at the same time. In this paper we propose such an algorithm, whose aim is to address the node multi-ranking problem in multidimensional networks. We test our algorithm over several real world networks, extracted from DBLP and the Enron email corpus, and we show its usefulness in providing less trivial and more flexible rankings than the current state of the art algorithms.Project(s): DATA SIM
Coscia M, Rossetti G, Pennacchioli D, Ceccarelli D, Giannotti F
Finding talents, often among the people already hired, is an endemic challenge for organizations. The social networking revolution, with online tools like Linkedin, made possible to make explicit and accessible what we perceived, but not used, for thousands of years: the exact position and ranking of a person in a network of professional and personal connections. To search and mine where and how an employee is positioned on a global skill network will enable organizations to find unpredictable sources of knowledge, innovation and know-how. This data richness and hidden knowledge demands for a multidimensional and multiskill approach to the network ranking problem. Multidimensional networks are networks with multiple kinds of relations. To the best of our knowledge, no network-based ranking algorithm is able to handle multidimensional networks and multiple rankings over multiple attributes at the same time. In this paper we propose such an algorithm, whose aim is to address the node multi-ranking problem in multidimensional networks. We test our algorithm over several real world networks, extracted from DBLP and the Enron email corpus, and we show its usefulness in providing less trivial and more flexible rankings than the current state of the art algorithms.Project(s): DATA SIM
2013
Conference article
Open Access
The three dimensions of social prominence
Pennacchioli D, Rossetti G, Pappalardo L, Pedreschi D, Giannotti F, Coscia M
One classic problem denition in social network analysis is the study of diusion in networks, which enables us to tackle problems like favoring the adoption of positive technologies. Most of the attention has been turned to how to maximize the number of in uenced nodes, but this approach misses the fact that dierent scenarios imply dierent dif- fusion dynamics, only slightly related to maximizing the number of nodes involved. In this paper we measure three dierent dimensions of social prominence: the Width, i.e. the ratio of neighbors in uenced by a node; the Depth, i.e. the degrees of separation from a node to the nodes perceiv- ing its prominence; and the Strength, i.e. the intensity of the prominence of a node. By dening a procedure to extract prominent users in complex networks, we detect associations between the three dimensions of social prominence and classical network statistics. We validate our results on a social network extracted from the Last.Fm music platform.DOI: 10.1007/978-3-319-03260-3_28
Project(s): DATA SIM
Metrics:
Pennacchioli D, Rossetti G, Pappalardo L, Pedreschi D, Giannotti F, Coscia M
One classic problem denition in social network analysis is the study of diusion in networks, which enables us to tackle problems like favoring the adoption of positive technologies. Most of the attention has been turned to how to maximize the number of in uenced nodes, but this approach misses the fact that dierent scenarios imply dierent dif- fusion dynamics, only slightly related to maximizing the number of nodes involved. In this paper we measure three dierent dimensions of social prominence: the Width, i.e. the ratio of neighbors in uenced by a node; the Depth, i.e. the degrees of separation from a node to the nodes perceiv- ing its prominence; and the Strength, i.e. the intensity of the prominence of a node. By dening a procedure to extract prominent users in complex networks, we detect associations between the three dimensions of social prominence and classical network statistics. We validate our results on a social network extracted from the Last.Fm music platform.DOI: 10.1007/978-3-319-03260-3_28
Project(s): DATA SIM
Metrics:
See at:
www.michelecoscia.com 
| doi.org 
| CNR IRIS | CNR IRIS | link.springer.com
2013
Conference article
Open Access
"You know because I know": a multidimensional network approach to human resources problem
Coscia M, Rossetti G, Pennachioli D, Ceccarelli D, Giannotti F
Finding talents, often among the people already hired, is an endemic challenge for organizations. The social networking revolution, with online tools like Linkedin, made possible to make explicit and accessible what we perceived, but not used, for thousands of years: the exact position and ranking of a person in a network of professional and personal connections. To search and mine where and how an employee is positioned on a global skill network will enable organizations to find unpredictable sources of knowledge, innovation and know- how. This data richness and hidden knowledge demands for a multidimensional and multiskill approach to the network ranking problem. Multidimensional networks are networks with multiple kinds of relations. To the best of our knowledge, no network-based ranking algorithm is able to handle multidimensional networks and multiple rankings over multiple attributes at the same time. In this paper we propose such an algorithm, whose aim is to address the node multi-ranking problem in multidimensional networks. We test our algorithm over several real world networks, extracted from DBLP and the Enron email corpus, and we show its usefulness in providing less trivial and more flexible rankings than the current state of the art algorithms.DOI: 10.1145/2492517.2492537
DOI: 10.48550/arxiv.1305.7146
Project(s): DATA SIM
Metrics:
Coscia M, Rossetti G, Pennachioli D, Ceccarelli D, Giannotti F
Finding talents, often among the people already hired, is an endemic challenge for organizations. The social networking revolution, with online tools like Linkedin, made possible to make explicit and accessible what we perceived, but not used, for thousands of years: the exact position and ranking of a person in a network of professional and personal connections. To search and mine where and how an employee is positioned on a global skill network will enable organizations to find unpredictable sources of knowledge, innovation and know- how. This data richness and hidden knowledge demands for a multidimensional and multiskill approach to the network ranking problem. Multidimensional networks are networks with multiple kinds of relations. To the best of our knowledge, no network-based ranking algorithm is able to handle multidimensional networks and multiple rankings over multiple attributes at the same time. In this paper we propose such an algorithm, whose aim is to address the node multi-ranking problem in multidimensional networks. We test our algorithm over several real world networks, extracted from DBLP and the Enron email corpus, and we show its usefulness in providing less trivial and more flexible rankings than the current state of the art algorithms.DOI: 10.1145/2492517.2492537
DOI: 10.48550/arxiv.1305.7146
Project(s): DATA SIM
Metrics:
See at:
arXiv.org e-Print Archive | dl.acm.org | doi.org | doi.org | CNR IRIS | CNR IRIS
2013
Conference article
Restricted
Measuring tie strength in multidimensional networks
Rossetti G, Pappalardo L, Pedreschi D
Online social networks have allowed us to build massive net- works of weak ties: acquaintances and nonintimate ties we use all the time to spread information and thoughts. Conversely, strong ties are people we really trust, persons most like us and whose social circles tightly overlap with our own. Unfortunately, social media do not incorporate tie strength in the creation and management of relationships, and treat all users the same: friend or stranger, with little or nothing in between. In the current work, we address the challenging issue of detecting on online social networks the strong and intimate ties from the huge mass of such mere social contacts. In order to do so, we propose a novel multidimensional definition of tie strength which exploits the existence of multiple online social links between two individuals. We test our definition on a multidimensional network constructed over users in Foursquare, Twitter and Facebook, analyzing the structural role of strong e weak links, and the correlations with the most common similarity measures.Project(s): DATA SIM
Rossetti G, Pappalardo L, Pedreschi D
Online social networks have allowed us to build massive net- works of weak ties: acquaintances and nonintimate ties we use all the time to spread information and thoughts. Conversely, strong ties are people we really trust, persons most like us and whose social circles tightly overlap with our own. Unfortunately, social media do not incorporate tie strength in the creation and management of relationships, and treat all users the same: friend or stranger, with little or nothing in between. In the current work, we address the challenging issue of detecting on online social networks the strong and intimate ties from the huge mass of such mere social contacts. In order to do so, we propose a novel multidimensional definition of tie strength which exploits the existence of multiple online social links between two individuals. We test our definition on a multidimensional network constructed over users in Foursquare, Twitter and Facebook, analyzing the structural role of strong e weak links, and the correlations with the most common similarity measures.Project(s): DATA SIM
2013
Conference article
Restricted
Quantification trees
Milli L, Monreale A, Rossetti G, Giannotti F, Pedreschi D, Sebastiani F
In many applications there is a need to monitor how a population is distributed across different classes, and to track the changes in this distribution that derive from varying circumstances; an example such application is monitoring the percentage (or "prevalence") of unemployed people in a given region, or in a given age range, or at different time periods. When the membership of an individual in a class cannot be established deterministically, this monitoring activity requires classification. However, in the above applications the final goal is not determining which class each individual belongs to, but simply estimating the prevalence of each class in the unlabeled data. This task is called quantification. In a supervised learning framework we may estimate the distribution across the classes in a test set from a training set of labeled individuals. However, this may be suboptimal, since the distribution in the test set may be substantially different from that in the training set (a phenomenon called distribution drift). So far, quantification has mostly been addressed by learning a classifier optimized for individual classification and later adjusting the distribution it computes to compensate for its tendency to either under- or over-estimate the prevalence of the class. In this paper we propose instead to use a type of decision trees (quantification trees) optimized not for individual classification, but directly for quantification. Our experiments show that quantification trees are more accurate than existing state-of-the-art quantification methods, while retaining at the same time the simplicity and understandability of the decision tree framework.DOI: 10.1109/icdm.2013.122
Metrics:
Milli L, Monreale A, Rossetti G, Giannotti F, Pedreschi D, Sebastiani F
In many applications there is a need to monitor how a population is distributed across different classes, and to track the changes in this distribution that derive from varying circumstances; an example such application is monitoring the percentage (or "prevalence") of unemployed people in a given region, or in a given age range, or at different time periods. When the membership of an individual in a class cannot be established deterministically, this monitoring activity requires classification. However, in the above applications the final goal is not determining which class each individual belongs to, but simply estimating the prevalence of each class in the unlabeled data. This task is called quantification. In a supervised learning framework we may estimate the distribution across the classes in a test set from a training set of labeled individuals. However, this may be suboptimal, since the distribution in the test set may be substantially different from that in the training set (a phenomenon called distribution drift). So far, quantification has mostly been addressed by learning a classifier optimized for individual classification and later adjusting the distribution it computes to compensate for its tendency to either under- or over-estimate the prevalence of the class. In this paper we propose instead to use a type of decision trees (quantification trees) optimized not for individual classification, but directly for quantification. Our experiments show that quantification trees are more accurate than existing state-of-the-art quantification methods, while retaining at the same time the simplicity and understandability of the decision tree framework.DOI: 10.1109/icdm.2013.122
Metrics:
See at:
CNR IRIS | CNR IRIS | xplorestaging.ieee.org