Word prediction with multiple overlapping contexts
Systems and processes for word prediction using multiple contexts are provided. For example, a plurality of words are received. A first word context including a first plurality of received words, and a second word context corresponding to the first plurality of received words and a second plurality of received words, are obtained. A first current word probability is determined based on a first language model using the first word context. A second current word probability is determined based on a second language model using the second word context. A third current word probability is determined based on the second language model using the first word context. A fourth current word probability is determined based on the first current word probability, the second current word probability, and the third current word probability. An output is provided, to a user, including a current word prediction based on the fourth current word probability.
1. An electronic device, comprising:
one or more processors;
a memory; and
one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for:
receiving, from a user, a plurality of words;
obtaining, from the plurality of received words, a first word context including a first plurality of received words;
obtaining, from the plurality of received words, a second word context corresponding to the first plurality of received words and a second plurality of received words;
determining, based on a first language model, a first current word probability using the first word context;
determining, based on a second language model, a second current word probability using the second word context;
determining, based on the second language model, a third current word probability using the first word context;
obtaining a fourth current word probability based on the first current word probability, the second current word probability, and the third current word probability; and
providing, to the user, an output including a current word prediction based on the fourth current word probability.
2. The device of claim 1 , wherein the second word context includes a plurality of word portions corresponding to the first plurality of received words and the second plurality of received words.
3. The device of claim 2 , wherein receiving, from a user, a plurality of words comprises:
receiving a first word and a second word, wherein
the second word was received prior to the first word, and
the first word and the second word each include at least one word portion.
4. The device of claim 1 , wherein receiving, from a user, a plurality of words comprises:
receiving the second plurality of words prior to receiving the first plurality of words.
5. The device of claim 1 , wherein the first plurality of received words includes between one and five words.
6. The device of claim 1 , wherein a combination of the first plurality of received words and the second plurality of received words includes at least eighty words.
7. The device of claim 1 , the one or more programs including instructions for:
in accordance with determining a word probability using the first language model, determining a word probability based on at least one word included in the first word context.
8. The device of claim 1 , the one or more programs including instructions for:
in accordance with determining a word probability using the second language model, determining a word probability based on at least one word portion in the second word context.
9. The device of claim 1 , the one or more programs including instructions for:
obtaining, based on the first language model, a fifth current word probability based on the second word context.
10. The device of claim 9 , the one or more programs including instructions for:
identifying a coefficient based on a combination of the second current word probability and the fifth current word probability; and
obtaining the fourth current word probability based on the identified coefficient.
11. The device of claim 1 , the one or more programs including instructions for:
weighting, using a first weight, the first current word probability based on a third word context, wherein
the second word context includes a first plurality of word portions and a second plurality of word portions obtained prior to the first plurality of word portions, and wherein
the third word context includes the second plurality of word portions.
12. The device of claim 1 , the one or more programs including instructions for:
obtaining a weight factor based on a fraction including the second current word probability over the third current word probability; and
weighting, using the weight factor, the first current word probability.
13. The device of claim 1 , the one or more programs including instructions for:
obtaining, based on the first word context and the current word prediction, a weight factor.
14. The device of claim 1 , the one or more programs including instructions for:
determining, based on the first language model, a first plurality of current word probabilities using the first word context;
determining, based on a second language model, a second plurality of current word probabilities using the second word context;
determining, based on the second language model, a third plurality of current word probabilities using the first word context; and
obtaining a fourth plurality of current word probabilities based on the first plurality of current word probabilities, the second plurality of current word probabilities, and the third plurality of current word probabilities.
15. The device of claim 14 , the one or more programs including instructions for:
for each current word probability of the fourth plurality of current word probabilities:
providing, to the user, an output including a respective current word prediction based on a respective current word probability of the fourth plurality of current word probabilities.
16. The device of claim 1 , the one or more programs including instructions for:
receiving, from the user, the plurality of words as a textual input to a messaging application, wherein the first word context and the second word context are associated with the messaging application.
17. The device of claim 1 , the one or more programs including instructions for:
displaying, as the provided output to the user, an affordance including a word corresponding to the current word prediction; and
in response to receiving a user selection of the affordance, providing the word corresponding to the current word prediction as a textual input to a messaging application.
18. A computer-implemented method, comprising:
at an electronic device with one or more processors and memory:
receiving, from a user, a plurality of words;
obtaining, from the plurality of received words, a first word context including a first plurality of received words;
obtaining, from the plurality of received words, a second word context corresponding to the first plurality of received words and a second plurality of received words;
determining, based on a first language model, a first current word probability using the first word context;
determining, based on a second language model, a second current word probability using the second word context;
determining, based on the second language model, a third current word probability using the first word context;
obtaining a fourth current word probability based on the first current word probability, the second current word probability, and the third current word probability; and
providing, to the user, an output including a current word prediction based on the fourth current word probability.
19. The method of claim 18 , wherein the second word context includes a plurality of word portions corresponding to the first plurality of received words and the second plurality of received words.
20. The method of claim 19 , wherein receiving, from a user, a plurality of words comprises:
receiving a first word and a second word, wherein
the second word was received prior to the first word, and
the first word and the second word each include at least one word portion.
21. The method of claim 18 , wherein receiving, from a user, a plurality of words comprises:
receiving the second plurality of words prior to receiving the first plurality of words.
22. The method of claim 18 , wherein the first plurality of received words includes between one and five words.
23. The method of claim 18 , wherein a combination of the first plurality of received words and the second plurality of received words includes at least eighty words.
24. The method of claim 18 , comprising:
in accordance with determining a word probability using the first language model, determining a word probability based on at least one word included in the first word context.
25. The method of claim 18 , comprising:
in accordance with determining a word probability using the second language model, determining a word probability based on at least one word portion in the second word context.
26. The method of claim 18 , comprising:
obtaining, based on the first language model, a fifth current word probability based on the second word context.
27. The method of claim 26 , comprising:
identifying a coefficient based on a combination of the second current word probability and the fifth current word probability; and
obtaining the fourth current word probability based on the identified coefficient.
28. The method of claim 18 , comprising:
weighting, using a first weight, the first current word probability based on a third word context, wherein
the second word context includes a first plurality of word portions and a second plurality of word portions obtained prior to the first plurality of word portions, and wherein
the third word context includes the second plurality of word portions.
29. The method of claim 18 , comprising:
obtaining a weight factor based on a fraction including the second current word probability over the third current word probability; and
weighting, using the weight factor, the first current word probability.
30. The method of claim 18 , comprising:
obtaining, based on the first word context and the current word prediction, a weight factor.
31. The method of claim 18 , comprising:
determining, based on the first language model, a first plurality of current word probabilities using the first word context;
determining, based on a second language model, a second plurality of current word probabilities using the second word context;
determining, based on the second language model, a third plurality of current word probabilities using the first word context; and
obtaining a fourth plurality of current word probabilities based on the first plurality of current word probabilities, the second plurality of current word probabilities, and the third plurality of current word probabilities.
32. The method of claim 31 , comprising:
for each current word probability of the fourth plurality of current word probabilities:
providing, to the user, an output including a respective current word prediction based on a respective current word probability of the fourth plurality of current word probabilities.
33. The method of claim 18 , comprising:
receiving, from the user, the plurality of words as a textual input to a messaging application, wherein the first word context and the second word context are associated with the messaging application.
34. The method of claim 18 , comprising:
displaying, as the provided output to the user, an affordance including a word corresponding to the current word prediction; and
in response to receiving a user selection of the affordance, providing the word corresponding to the current word prediction as a textual input to a messaging application.
35. A non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by one or more processors of a first electronic device, cause the first electronic device to:
receive, from a user, a plurality of words;
obtain, from the plurality of received words, a first word context including a first plurality of received words;
obtain, from the plurality of received words, a second word context corresponding to the first plurality of received words and a second plurality of received words;
determine, based on a first language model, a first current word probability using the first word context;
determine, based on a second language model, a second current word probability using the second word context;
determine, based on the second language model, a third current word probability using the first word context;
obtain a fourth current word probability based on the first current word probability, the second current word probability, and the third current word probability; and
provide, to the user, an output including a current word prediction based on the fourth current word probability.
36. The computer-readable storage medium of claim 35 , wherein the second word context includes a plurality of word portions corresponding to the first plurality of received words and the second plurality of received words.
37. The computer-readable storage medium of claim 36 , wherein receiving, from a user, a plurality of words comprises:
receiving a first word and a second word, wherein
the second word was received prior to the first word, and
the first word and the second word each include at least one word portion.
38. The computer-readable storage medium of claim 35 , wherein receiving, from a user, a plurality of words comprises:
receiving the second plurality of words prior to receiving the first plurality of words.
39. The computer-readable storage medium of claim 35 , wherein the first plurality of received words includes between one and five words.
40. The computer-readable storage medium of claim 35 , wherein a combination of the first plurality of received words and the second plurality of received words includes at least eighty words.
41. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
in accordance with determining a word probability using the first language model, determine a word probability based on at least one word included in the first word context.
42. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
in accordance with determining a word probability using the second language model, determine a word probability based on at least one word portion in the second word context.
43. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
obtain, based on the first language model, a fifth current word probability based on the second word context.
44. The computer-readable storage medium of claim 43 , wherein the instructions cause the first electronic device to:
identify a coefficient based on a combination of the second current word probability and the fifth current word probability; and
obtain the fourth current word probability based on the identified coefficient.
45. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
weight, using a first weight, the first current word probability based on a third word context, wherein
the second word context includes a first plurality of word portions and a second plurality of word portions obtained prior to the first plurality of word portions, and wherein
the third word context includes the second plurality of word portions.
46. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
obtain a weight factor based on a fraction including the second current word probability over the third current word probability; and
weight, using the weight factor, the first current word probability.
47. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
obtain, based on the first word context and the current word prediction, a weight factor.
48. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
determine, based on the first language model, a first plurality of current word probabilities using the first word context;
determine, based on a second language model, a second plurality of current word probabilities using the second word context;
determine, based on the second language model, a third plurality of current word probabilities using the first word context; and
obtain a fourth plurality of current word probabilities based on the first plurality of current word probabilities, the second plurality of current word probabilities, and the third plurality of current word probabilities.
49. The computer-readable storage medium of claim 48 , wherein the instructions cause the first electronic device to:
for each current word probability of the fourth plurality of current word probabilities:
provide, to the user, an output including a respective current word prediction based on a respective current word probability of the fourth plurality of current word probabilities.
50. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
receive, from the user, the plurality of words as a textual input to a messaging application, wherein the first word context and the second word context are associated with the messaging application.
51. The computer-readable storage medium of claim 35 , wherein the instructions cause the first electronic device to:
display, as the provided output to the user, an affordance including a word corresponding to the current word prediction; and
in response to receiving a user selection of the affordance, provide the word corresponding to the current word prediction as a textual input to a messaging application.