Research Article
An Effective Recommender Algorithm for Cold-Start Problem in Academic Social Networks
Table 5
One-way ANOVA tests for examining the difference of means between recommender algorithms.
(a) ANOVA test on precision value. |
| | Rnd | CL | CB | ECSN |
| Rnd | 1 | | | | CL | 0.03735 | 1 | | | CB | 0.04395* | 0.00660 | 1 | | ECSN | 0.07847* | 0.04111* | 0.03452* | 1 |
|
|
(b) ANOVA test on recall value. |
| | Rnd | CL | CB | ECSN |
| Rnd | 1 | | | | CL | 0.4932 | 1 | | | CB | 0.05807 | 0.00875 | 1 | | ECSN | 0.02330 | 0.02602 | 0.03477 | 1 |
|
|
(c) ANOVA test on fallout value. |
| | Rnd | CL | CB | ECSN |
| Rnd | 1 | | | | CL | 0.00346* | 1 | | | CB | 0.00431* | 0.0085 | 1 | | ECSN | 0.00741* | 0.00395* | 0.00310 | 1 |
|
|
(d) ANOVA test on value. |
| | Rnd | CL | CB | ECSN |
| Rnd | 1 | | | | CL | 0.04902* | 1 | | | CB | 0.05765* | 0.00863 | 1 | | ECSN | 0.10444* | 0.05542* | 0.04679* | 1 |
|
|
The mean difference is significant at the 0.05 level.
|