Disambiguation of contact information using historical data

1. A system comprising:

one or more computers; and

a computer-readable medium coupled to the one or more computers having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising:

receiving an audio signal,

generating, for each of two or more items of contact information, an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information, and further based on a recency of one or more past interactions between the user and the contact associated with the item of contact information, wherein generating the affinity score includes associating a first weight value to an electronic message in which the contact information is included in a primary recipient field, and a second, different weight value to an electronic message in which the contact information is included in a carbon copy recipient field,

inferring, for each of the items of contact information, a probability that the user intends to initiate a communication using the item of contact information based on the affinity score generated for the item of contact information,

generating a communication initiation grammar that includes one or more transitions associated with each of the items of contact information, wherein, for each of the items of contact information, the one or more transitions associated with the item of contact information are weighted according to the probability inferred for the item of contact information,

performing speech recognition on the audio signal using the communication initiation grammar, to select a particular item of contact information, and

initiating the communication using the particular item of contact information.

2. The system of claim 1 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has accepted a result of a voice dial speech recognition that has dialed the contact; and

incrementing the affinity score based on the frequency with which the user has accepted the result of the voice dial speech recognition.

3. The system of claim 1 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has manually dialed the contact; and

incrementing the affinity score based on the frequency with which the user has manually dialed the contact.

4. The system of claim 1 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has sent a Short Message Service (SMS) message to the contact or has received an SMS message from the contact; and

incrementing the affinity score based on the frequency with which the user has sent an SMS message to the contact or has received an SMS message from the contact.

5. The system of claim 1 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has sent an e-mail message to the contact or has received an e-mail message from the contact; and

incrementing the affinity score based on the frequency with which the user has sent an e-mail message to the contact or has received an e-mail message from the contact.

6. The system of claim 1 , wherein generating an affinity score based on a recency of one or more past interactions between the user and the contact associated with the item of contact information further comprises:

determining that the contact has been added to an address book of the user within a predetermined period of time; and

incrementing the affinity score based on determining that the contact has been added to the address book of the user within a predetermined period of time.

7. The system of claim 1 , wherein:

the operations further comprise:

receiving data that references a current context associated with the user,

receiving, for each of two or more past interactions between the user and the contact, data that references a context associated with the past interaction,

comparing the current context associated with the user with the contexts associated with the past interactions, and

selecting a subset of the two or more past interactions based on comparing the current context associated with the user with the contexts associated with the past interactions; and

the affinity scores are generated using the subset of the past interactions only.

8. The system of claim 1 , wherein generating, for each of two or more items of contact information, an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

generating a first partial affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information using a first communications channel;

generating a second partial affinity score based on a frequency with which the user has previously communicated with the contact associated with the item of contact information using a second communication channel;

weighting the first partial affinity score and the second partial affinity score; and

combining the weighted, first partial affinity score with the weighted, partial affinity score to generate the affinity score.

9. The system of claim 1 , wherein generating, for each of two or more items of contact information, an affinity score based on a recency of one or more past interactions between the user and the contact associated with the item of contact information further comprises:

applying an exponential decay function to data which references when the one or more past interactions between the user and the contact associated with the item of contact information occurred.

10. The system of claim 1 , wherein generating, for each of two or more items of contact information, an affinity score based on a recency of one or more past interactions between the user and the contact associated with the item of contact information further comprises:

selecting a subset of the one or more past interactions that have occurred within a predetermined period of time.

11. A computer-implemented method comprising:

receiving an audio signal;

generating, for each of two or more items of contact information, an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information, and further based on a recency of one or more past interactions between the user and the contact associated with the item of contact information, wherein generating the affinity score includes associating a first weight value to an electronic message in which the contact information is included in a primary recipient field, and a second, different weight value to an electronic message in which the contact information is included in a carbon copy recipient field;

inferring, for each of the items of contact information, a probability that the user intends to initiate a communication using the item of contact information based on the affinity score generated for the item of contact information;

generating a communication initiation grammar that includes one or more transitions associated with each of the items of contact information, wherein, for each of the items of contact information, the one or more transitions associated with the item of contact information are weighted according to the probability inferred for the item of contact information;

performing speech recognition on the audio signal using the communication initiation grammar, to select a particular item of contact information; and

initiating the communication using the particular item of contact information.

12. A computer storage medium encoded with a computer program, the program comprising instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

receiving an audio signal;

generating, for each of two or more items of contact information, an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information, and further based on a recency of one or more past interactions between the user and the contact associated with the item of contact information, wherein generating the affinity score includes associating a first weight value to an electronic message in which the contact information is included in a primary recipient field, and a second, different weight value to an electronic message in which the contact information is included in a carbon copy recipient field;

inferring, for each of the items of contact information, a probability that the user intends to initiate a communication using the item of contact information based on the affinity score generated for the item of contact information;

generating a communication initiation grammar that includes one or more transitions associated with each of the items of contact information, wherein, for each of the items of contact information, the one or more transitions associated with the item of contact information are weighted according to the probability inferred for the item of contact information;

performing speech recognition on the audio signal using the communication initiation grammar, to select a particular item of contact information; and

initiating the communication using the particular item of contact information.

13. The computer storage medium of claim 12 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has accepted a result of a voice dial speech recognition that has dialed the contact; and

incrementing the affinity score based on the frequency with which the user has accepted the result of the voice dial speech recognition.

14. The computer storage medium of claim 12 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has manually dialed the contact; and

incrementing the affinity score based on the frequency with which the user has manually dialed the contact.

15. The computer storage medium of claim 12 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has sent a Short Message Service (SMS) message to the contact or has received an SMS message from the contact; and

incrementing the affinity score based on the frequency with which the user has sent an SMS message to the contact or has received an SMS message from the contact.

16. The computer storage medium of claim 12 , wherein generating an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

determining a frequency with which the user has sent an e-mail message to the contact or has received an e-mail message from the contact; and

incrementing the affinity score based on the frequency with which the user has sent an e-mail message to the contact or has received an e-mail message from the contact.

17. The computer storage medium of claim 12 , wherein generating an affinity score based on a recency of one or more past interactions between the user and the contact associated with the item of contact information further comprises:

determining that the contact has been added to an address book of the user within a predetermined period of time; and

incrementing the affinity score based on determining that the contact has been added to the address book of the user within a predetermined period of time.

18. The computer storage medium of claim 12 , wherein:

the operations further comprise:

receiving data that references a current context associated with the user,

receiving, for each of two or more past interactions between the user and the contact, data that references a context associated with the past interaction,

comparing the current context associated with the user with the contexts associated with the past interactions, and

selecting a subset of the two or more past interactions based on comparing the current context associated with the user with the contexts associated with the past interactions; and

the affinity scores are generated using the subset of the past interactions only.

19. The computer storage medium of claim 12 , wherein generating, for each of two or more items of contact information, an affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information further comprises:

generating a first partial affinity score based on a frequency with which a user has previously communicated with a contact associated with the item of contact information using a first communications channel;

generating a second partial affinity score based on a frequency with which the user has previously communicated with the contact associated with the item of contact information using a second communication channel;

weighting the first partial affinity score and the second partial affinity score; and

combining the weighted, first partial affinity score with the weighted, partial affinity score to generate the affinity score.

20. The computer storage medium of claim 12 , wherein generating, for each of two or more items of contact information, an affinity score based on a recency of one or more past interactions between the user and the contact associated with the item of contact information further comprises:

applying an exponential decay function to data which references when the one or more past interactions between the user and the contact associated with the item of contact information occurred.