Eleven Point Precision-Recall Curve

Definition

The 11-point precision-recall curve is a graph plotting the interpolated precision of an information retrieval (IR) system at 11 standard recall levels, that is, {0.0, 0.1, 0.2, …, 1.0}. The method for interpolation is detailed below. The graph is widely used to evaluate IR systems that return ranked documents, which are common in modern search systems.

Key Points

In a precision-recall graph, precision is plotted as a function of recall. Assume a search system has retrieved n documents. Then for each kϵ [0, n], examine the set of top k retrieved documents and calculate the precision and recall for the set. These values are then plotted on a graph of precision versus recall. Figure 1 shows a typical precision-recall curve. It has a saw-toothed shape because the recall remains the same as k while the precision drops when the (k+1)th retrieved document is not relevant.

Eleven Point Precision-Recall Curve, Fig. 1

Uninterpolated precision-recall curve