We present the analysis of microlensing event OGLE-2006-BLG-284 , which has a lens system that consists of two stars and a gas giant planet with a mass ratio of q _ { p } = ( 1.26 \pm 0.19 ) \times 10 ^ { -3 } to the primary .
The mass ratio of the two stars is q _ { s } = 0.289 \pm 0.011 , and their projected separation is s _ { s } = 2.1 \pm 0.7 AU , while the projected separation of the planet from the primary is s _ { p } = 2.2 \pm 0.8 AU .
For this lens system to have stable orbits , the three-dimensional separation of either the primary and secondary stars or the planet and primary star must be much larger than that these projected separations .
Since we do not know which is the case , the system could include either a circumbinary or a circumstellar planet .
Because there is no measurement of the microlensing parallax effect or lens system brightness , we can only make a rough Bayesian estimate of the lens system masses and brightness .
We find host star and planet masses of M _ { L 1 } = 0.35 ^ { +0.30 } _ { -0.20 } M _ { \odot } , M _ { L 2 } = 0.10 ^ { +0.09 } _ { -0.06 } M _ { \odot } , and m _ { p } = 144 ^ { +126 } _ { -82 } { M _ { \oplus } } , and the K -band magnitude of the combined brightness of the host stars is K _ { L } = 19.7 ^ { +0.7 } _ { -1.0 } .
The separation between the lens and source system will be \sim 90 mas in mid-2020 , so it should be possible to detect the host system with follow-up adaptive optics or Hubble Space Telescope observations .