Entropy-based approach to missing-links prediction
Abstract Link-prediction is an active research field within network theory, aiming at uncovering missing connections or predicting the emergence of future relationships from the observed network structure. This paper represents our contribution to the stream of research concerning missing links pred...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2018-07-01
|
| Series: | Applied Network Science |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1007/s41109-018-0073-4 |
| _version_ | 1818359895849697280 |
|---|---|
| author | Federica Parisi Guido Caldarelli Tiziano Squartini |
| author_facet | Federica Parisi Guido Caldarelli Tiziano Squartini |
| author_sort | Federica Parisi |
| collection | DOAJ |
| description | Abstract Link-prediction is an active research field within network theory, aiming at uncovering missing connections or predicting the emergence of future relationships from the observed network structure. This paper represents our contribution to the stream of research concerning missing links prediction. Here, we propose an entropy-based method to predict a given percentage of missing links, by identifying them with the most probable non-observed ones. The probability coefficients are computed by solving opportunely defined null-models over the accessible network structure. Upon comparing our likelihood-based, local method with the most popular algorithms over a set of economic, financial and food networks, we find ours to perform best, as pointed out by a number of statistical indicators (e.g. the precision, the area under the ROC curve, etc.). Moreover, the entropy-based formalism adopted in the present paper allows us to straightforwardly extend the link-prediction exercise to directed networks as well, thus overcoming one of the main limitations of current algorithms. The higher accuracy achievable by employing these methods - together with their larger flexibility - makes them strong competitors of available link-prediction algorithms. |
| first_indexed | 2024-12-13T20:52:10Z |
| format | Article |
| id | doaj.art-3d565655c4dd475180809385be3d4474 |
| institution | Directory Open Access Journal |
| issn | 2364-8228 |
| language | English |
| last_indexed | 2024-12-13T20:52:10Z |
| publishDate | 2018-07-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Applied Network Science |
| spelling | doaj.art-3d565655c4dd475180809385be3d44742022-12-21T23:31:51ZengSpringerOpenApplied Network Science2364-82282018-07-013111510.1007/s41109-018-0073-4Entropy-based approach to missing-links predictionFederica Parisi0Guido Caldarelli1Tiziano Squartini2IMT School for Advanced Studies LuccaIMT School for Advanced Studies LuccaIMT School for Advanced Studies LuccaAbstract Link-prediction is an active research field within network theory, aiming at uncovering missing connections or predicting the emergence of future relationships from the observed network structure. This paper represents our contribution to the stream of research concerning missing links prediction. Here, we propose an entropy-based method to predict a given percentage of missing links, by identifying them with the most probable non-observed ones. The probability coefficients are computed by solving opportunely defined null-models over the accessible network structure. Upon comparing our likelihood-based, local method with the most popular algorithms over a set of economic, financial and food networks, we find ours to perform best, as pointed out by a number of statistical indicators (e.g. the precision, the area under the ROC curve, etc.). Moreover, the entropy-based formalism adopted in the present paper allows us to straightforwardly extend the link-prediction exercise to directed networks as well, thus overcoming one of the main limitations of current algorithms. The higher accuracy achievable by employing these methods - together with their larger flexibility - makes them strong competitors of available link-prediction algorithms.http://link.springer.com/article/10.1007/s41109-018-0073-489.75.Hc; 89.65.Gh; 02.50.Tt |
| spellingShingle | Federica Parisi Guido Caldarelli Tiziano Squartini Entropy-based approach to missing-links prediction Applied Network Science 89.75.Hc; 89.65.Gh; 02.50.Tt |
| title | Entropy-based approach to missing-links prediction |
| title_full | Entropy-based approach to missing-links prediction |
| title_fullStr | Entropy-based approach to missing-links prediction |
| title_full_unstemmed | Entropy-based approach to missing-links prediction |
| title_short | Entropy-based approach to missing-links prediction |
| title_sort | entropy based approach to missing links prediction |
| topic | 89.75.Hc; 89.65.Gh; 02.50.Tt |
| url | http://link.springer.com/article/10.1007/s41109-018-0073-4 |
| work_keys_str_mv | AT federicaparisi entropybasedapproachtomissinglinksprediction AT guidocaldarelli entropybasedapproachtomissinglinksprediction AT tizianosquartini entropybasedapproachtomissinglinksprediction |