Phil Plait in Slate:
The star is called KIC 8462852, and it’s one of more than a hundred thousand stars that was observed by NASA’s Kepler mission. Kepler stared at these stars, looking for dips in their brightness. These very slight dimmings can be due to many factors, but one is if the star has planets, and one (or more) of them orbits the star in such a way that it passes directly in front of the star as seen from Earth. If it does—what we call a transit—we see a tiny diminution of starlight, usually by less than a percent.
Thousands of exoplanets have been found this way. Usually the planet is on a short orbit, so the dip we see is periodic, repeating every few days, weeks, or months, depending on the size of the planet’s orbit.
KIC 8462852 is a star somewhat more massive, hotter, and brighter than the Sun. It’s about 1,500 light-years away, a decent distance, so it’s too faint to see with the naked eye. The Kepler data for the star are pretty bizarre: There are dips in the light, but they aren’t periodic. They can be very deep; one dropped the amount of starlight by 15 percent, and another by a whopping 22 percent!
Straight away, we know we’re not dealing with a planet here. Even a Jupiter-sized planet only blocks roughly 1 percent of this kind of star’s light, and that’s about as big as a planet gets. It can’t be due to a star, either; we’d see it if it were. And the lack of a regular, repeating signal belies both of these as well. Whatever is blocking the star is big, though, up to half the width of the star itself!
Also, it turns out there are lots of these dips in the star’s light. Hundreds. And they don’t seem to be periodic at all. They have odd shapes to them, too. A planet blocking a star’s light will have a generally symmetric dip; the light fades a little, remains steady at that level, then goes back up later.
More here.