In mathematics vectors are abstract entities that like the vectors used in physics to represent forces can be summed or multiplied by a scalar (i.e. a number).
A vector space is roughly speaking a set of vectors that contains every sum of any two vectors in it and every multiple of every vector in it.
If we have two vector spaces $V$ and $W$ we can build up a new vector space $V\times W$ considering pairs of vectors: an element in $V\times W$ will be a pair
$$
(v,w)\qquad v\in V,\quad w\in W.
$$
The tensor product $V\otimes W$ is a different and a bit subtler way to pair vectors in $V$ with vectors in $W$. A basic element in $V\otimes W$ is something denoted
$$
v\otimes w.
$$
The main difference between pairs and tensors is that while the pairs
$$
(\lambda v,w)\qquad
(v,\lambda w)
$$
give different elements in $V\times W$ there is an actual equality
$$
(\lambda v)\otimes w=v\otimes(\lambda w)
$$
as elements in $V\otimes W$ (in the above $\lambda$ is any scalar).
The basic reason to introduce tensor products is to reduce, so to speak, the theory of multilinear functions to the theory of linear functions. For instance, there's a characterization of the determinant of a square matrix that uses the language of tensor products.