Skip to main content
. 2014 Feb 26;9(2):e89318. doi: 10.1371/journal.pone.0089318

Figure 3. Retrieval performance according to Inline graphic-value threshold.

Figure 3

Mean and standard error of the AP scores when a Inline graphic-value threshold is used. For each data set, we display the performance for two alternatives of our framework, namely with or without the TF-IDF step. Notice that, for every Inline graphic-value threshold (x-axis), the two framework alternatives are displayed side-by-side.