During known-plaintext attacks, the attacker has an access to the ciphertext and its corresponding plaintext. His goal is to guess the secret key (or a number of secret keys) or to develop an algorithm which would allow him to decrypt any further messages.
This gives the attacker much bigger possibilities to break the cipher than just by performing ciphertext only attacks. However, he is no able to actively provide customized data or secret keys which would be processed by the cipher.
Known-plaintext attacks are most effective when they are used against the simplest kinds of ciphers. For example, applying them against simple substitution ciphers allows the attacker to break them almost immediately.
Known-plaintext attacks were commonly used for attacking the ciphers used during the Second World War. The most notably example would be perhaps the attempts made by the British while attacking German Enigma ciphers. The English intelligence targeted some common phrases, commonly appearing in encrypted German messages, like weather forecasts or geographical names.
The simple XOR cipher, used in the early days of computers, can be also broken easily by knowing only some parts of plaintext and corresponding encrypted messages.
Modern ciphers are generally resistant against purely known-plaintext attacks. One of the unfortunate exceptions was the old encryption method using in PKZIP application. Having just one copy of encrypted file, together with its original version, it was possible to completely recover the secret key.
In most cases however, the attacker should use more sophisticated types of cryptographic attacks in order to break a well-designed modern cipher.