About String Theory
(back)String Theory is, above all a Quantum Theory of Gravity. So we need to briefly describe what is a Quantum Theory, and what is Gravity. These two are the most important fundamental theories of the 20th century, both nicely applicable to many different phenomena, but both waiting since they were settled to become compatible under a more general framework.
Gravity is the theory that describes the way masses (or energy in general) atract other masses. The best formulation available at moment is still Einstein's General Theory of Relativity, dating back to 1916. Einstein's theory relates the effect of gravity interaction to spacetime curvature (grosso modo, masses attract because they curve the space around them, and then all bodies follow descend paths in this curved space). Gravity can be successfully applied to large scale situations, like planets orbiting the sun, Galaxies and even to the whole universe (recall that the Big Bang is a prediction of General Relativity).
Quantum mechanics typically applies to much smaller scales, from atoms to nuclei, to subnuclear physics. Despite being formulated as a general framework (rather than as being tied to specific interactions), it has proven amazingly useful to describe the other three interactions we know in Nature: electromagnetic, weak nuclear interaction, and strong nuclear interaction. You can read all sorts of paradoxes, weird predictions, and counter-intuitive phenomena that lie at the very definition of Quantum Mechanics... but, man, they are all true (won't enter into a phylosophical discussion of how to define true, but let me just point out a fact that quantifies its quantity of truth: it is capable of predicting that a certain property you can measure about an electron is whatever number with twelve decimal digits right!).
The problem with these two theories is that they themselves are capable of telling you that there's something wrong with them when put together. And unfortunately, you need to put them together sometimes, specially when you want to ask questions like 'what happened in the Big Bang?' or 'what happens to Black holes when they evaporate' or even 'what is all this dark matter we cannot see that makes the universe expand every time faster?'.
Thus, the problem of unification of both theories is perhaps the major one in physics today. Here's where String Theory appears, as one of the most solid candidates for unification. Indeed, it is the only candidate that has proven mathematical consistency when unifying them (at least in what is called perturbation theory, with good indications that consistency holds beyond this regime). Thus String Theory has turned a bit the problem of 'look for a theory' into a problem of 'solve this particular theory at least enough to answer my questions'. And that's still where we are, trying to extract the answers from a vast theory that requires better tools to find them...