Web archives already hold together more than 534 billion files and this number continues to grow as new initiatives arise. Searching on all versions of these files acquired throughout time is challenging, since users expect as fast and precise answers from web archives as the ones provided by current web search engines.

This talk discusses how to improve the search effectiveness of web archives through the creation of novel ranking features and ranking models that exploit the temporal dimension of archived data. A temporal-dependent ranking framework that exploits the variance of web characteristics over time is proposed. Based on the assumption that closer periods are more likely to hold similar web characteristics, this framework learns multiple models simultaneously, each tuned for a specific period. Experimental results show significant improvements over the search effectiveness of single-models created from all data independently of its time, using state-of-the-art learning-to-rank technology. This talk will also address ensemble approaches of ranking models.