A model of melt segregation and extraction within rising plumes is proposed. It is based on two-phase porous flow within the partially molten region, combined with only three extraction parameters. The conservation equations of mass, momentum, and energy are solved for a two-phase melt matrix system. As a rising hot mantle region (plume) reaches the asthenosphere, decompression melting occurs, and the melt begins to percolate with respect to the matrix. Accumulation layers form, which might be the locus for the formation of buoyancy-driven propagating dikes. As dike propagation requires a minimum dike length, melt extraction is parameterized by dex, $varphi$1, and $varphi$2. Here dex is the critical thickness of the partially molten layer in which a critical melt fraction $varphi$2 is exceeded. If this condition is met within a certain region of the melt source region, melt might be extracted from that region in the form of one or several propagating dikes, leaving behind a region of residual melt fraction $varphi$1. This simple extraction model is tested in one dimension for rising hot mantle flow. Depending on the chosen extraction parameters, multiple extraction events may be observed with a characteristic episodicity and a saw-tooth-like depth distribution. Exploring the parameter space shows that for values of dex and $varphi$2 of a few kilometers and a few melt percent, respectively, typical extraction cycles have the order of 103--104 years, and they extract melt volumes per surface area of 50 to several hundred meters each. Tentatively assuming that eruptions are tied to mantle ascent at depth, the model is applied to observed eruption frequencies and multiple extraction depths, and values for $varphi$2 of about 2% and dex of 3--5 km are derived.