In chess, the minority attack refers to a paradoxical strategy which was discovered during the Classical era of chess, and is defined as an attack by a smaller number of pawns against a greater number in a particular area of the chessboard. Generally speaking, according the the Classical tenets, a majority of pawns in one area of the board confers an advantage in that area. A queenside pawn majority, for example, confers an advantage in the endgame due to the possibility of creating a passed pawn on the side of the board, far away from the sphere of influence of the opponent's king.
However, it was discovered that in the opening and middle game, for example in openings such as the Queen's Gambit Declined, it was possible for the side with a minority of pawns to advance them with the intention of attacking the opponent's static pawn majority. This was discovered to have the effect of (if pursued to its logical conclusion of exchanging the advanced pawns) leaving one remaining opposing pawn on that side of the board. In an endgame this would give an advantage to the owner of the pawn, but in the middle game, before most of the pieces have been exchanged, it actually becomes a weakness, difficult to defend and liable to be captured.
The best strategy against a minority attack is generally not to try and prevent the advance of the pawns (which usually ends up being counterproductive and can end up creating further weaknesses) but to stir up some action in another area of the board, for instance by starting a kingside attack.