Input-method error correction method and input-method error correction device

Abstract

The invention provides an input-method error correction method and an input-method error correction device. The method includes that equipment displays a character inputted by a user in an error correction window in real time; the equipment detects that the user clicks the character, positioned in an area, in the error correction window for the first time to know that the user wants to correct the character, searches a preset number of keys, nearest to the character, on a keyboard, and generates a corrected-character option list in which characters of the keys are sorted from the near to the distant according to distances between the characters and the character to be corrected by the user; the equipment selects the first character in the corrected-character option list and substitutes the selected character for the character, to be corrected by the user, in the error correction window. By the input-method error correction method and the input-method error correction device, input-method error correction efficiency is improved.

Description

technical field [0001] The present application relates to the technical field of intelligent input, in particular to an input method error correction method and device. Background technique [0002] At present, there are mainly the following types of related technologies for error correction of input characters on smart mobile devices: [0003] 1. Intelligent error correction algorithm based on keyboard key distance and click position: According to the user input key or click position, adjacent input characters are obtained, and validity analysis is performed to generate error correction options. [0004] 2. Search for error correction candidates based on preset thesaurus and error-tolerant mechanism: Edit the input character string according to the preset thesaurus of the input method to generate error correction candidates. The editing operation includes inserting, deleting, replacing or exchanging characters order. [0005] 3. Context recognition technology: select one ...

AI Technical Summary

Problems solved by technology

[0007] 1. For the error correction algorithm based on the distance of the keyboard keys and the click position, when the user clicks by mistake, although it can be corrected to a certain extent by inputting characters, it will inevitably increase the number of candidate words and increase the user's judgment and choice. time, which affects the input speed, especially when there are many errors in the input of long sentences, it is impossible to prepare for accurate options

[0008] 2. For the error correction candidate search based on the preset thesaurus and error tolerance mechanism, the error correction options are generated by inserting, deleting, replacing and exchanging the character order of the input characters, although the accuracy and correction of the input method itself are increased. Error capability, but because it is an automatic recognition that is out of the user's thinking, there will still be insufficient accuracy. There is no way to achieve 100% accuracy, especially when none of the candidate words is what the user wants, the user still needs to delete the error character, to re-enter, complex and cumbersome

[0009] 3. For the context recognition technology, it can recognize the characters adjacent to the characters clicked by the user, but if the error range of the user input is larger than the adjacent characters or the user wants to change a word, there will be problems, especially in long sentences. It takes a long time to delete the previously entered characters

[0010] With the vigorous development of wearable devices, traditional error correction methods also have obvious deficiencies. Due to the small screen of wearable devices (such as smart watches), the probability of users clicking by mistake is very high, which means that through error correction Algorithms and context recognition will produce more error correction options, and finally the candidate words provided to users will be greatly expanded. Users need to confirm and choose from many candidate words, which makes the operation difficult and the user experience is not good; at the same time, when users input The deviation is large, and the error correction algorithm cannot be corrected effectively. At this time, the user needs to modify it manually. However, in view of the small screen, the accuracy of the user's click is extremely low. Therefore, if the input string is modified and edited by touching the keyboard with a small button , itself has the problem of difficult operation and inaccurate input, and cannot effectively solve the problem of quickly modifying and editing the input string

Images