KR102068083B1 - Method and apparatus for user's inputting speed based input method error correction - Google Patents

Abstract

Provided are a method and apparatus for correcting an error of an input mode based on a user input speed. The method comprises the steps of: obtaining an input speed of a current separate key which is triggered, and adjusting a weight when an input code corresponding to the current separate key is corrected as input codes corresponding to other keys according to the input speed; performing error correction on the input code corresponding to the current separate key based on the weight, thereby obtaining at least one input code after the error correction; and according to the at least one input code after the error correction and corresponding to each key in one-time integral input, determining at least one candidate word corresponding to the integral input. According to the present invention, the input mode is assessed in a time dimension by taking the input speed as a factor for correcting the input code through an error correction model, so the error correction accuracy of the error correction model is improved and user input quality is enhanced.

Description

Input method error correction method and device based on user input speed {METHOD AND APPARATUS FOR USER'S INPUTTING SPEED BASED INPUT METHOD ERROR CORRECTION}

The present invention relates to the field of input method technology, and more particularly, to an input method error correction technology based on user input speed.

When a user inputs a text from a user device, error correction is mainly performed in the following two conventional methods. That is, 1) error correction is performed through a model of text error correction, and 2) through the start point information at the time of input, it is determined how the error point is to be corrected for the input code based on the positional relationship from each button. Here, the text error correction model is not only relatively hard, but also secures the quality of word output through a relatively strict screening strategy. In addition, it neglects a lot of environment and scene information at the time of user input, and also causes miscalculations due to differences in factors such as screen size, resolution, user's age, hand size, usage habits of one hand and two hands, As a result, the quality of word output through error correction is not high.

Therefore, one of the technical problems to be solved by those skilled in the art is to provide a more efficient, more accurate and smooth input error correction scheme.

An object of the present invention is to provide an input method error correction method and apparatus based on a user input speed.

In one aspect of the invention, there is provided an input method error correction method based on user input speed, the error correction method,

a. Acquiring an input speed of a triggered current independent button;

b. Adjusting a weighting value when correcting an input code corresponding to the current independent button with an input code corresponding to the other button according to the input speed;

c. Performing error correction on an input code corresponding to the current independent button based on the adjusted weighting to obtain at least one error corrected input code; And

d. Determining at least one candidate word corresponding to the entire input, according to the error-corrected at least one input code corresponding to each button of one full input.

Preferably, the error correction method,

e. Acquiring an average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code corresponding to the one entire input, and then determining the at least one candidate word corresponding to the entire input. And adjusting the ordering arrangement.

Preferably, step (e) is

e1. Calculating the average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code;

e2. Determining a corresponding error correction model based on the average input speed; And

e3. And adjusting the order arrangement of the at least one candidate word corresponding to the entire input according to the error correction model.

Preferably, the step of adjusting the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other button according to the input speed,

Increasing weighting when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;

Reducing a weight of itself when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;

Increasing a weight of itself when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; And

Reducing a weight when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; It includes at least one of.

Preferably, the determination of the difficulty of the button touch error,

Level class determination according to the input speed of the independent button;

Rating weighting according to the input speed of the independent button;

Determining a rating mapping to an input speed of the independent button and a fitting function according to the current input environment; And

A rating weighting decision mapping to an input speed of the independent button and a fitting function according to the current input environment; It includes at least one of.

Preferably, the error correction model,

An error correction level model independently based on the average input speed;

An error correction rating model independently based on the average input speed;

An error correction level model based on the average input speed and the remaining environmental parameter fittings; And

An error correction rating model based on the average input speed and the remaining environmental parameter fittings; It includes at least one of.

Preferably, step (e) is

Performing a noise removing process on a sequence consisting of a sequence into which each input code of the entire input codes is input, to obtain a noise canceled sequence;

Determining an input time of the entire input code according to the noise canceled sequential sequence; And

And obtaining the average input speed of the entire input code based on the input time of the entire input code, and adjusting the order arrangement of the at least one candidate word corresponding to the entire input.

Preferably, the noise canceling process,

If the sequence into which one input code is input is greater than a predetermined upper limit threshold or less than a predetermined lower limit threshold, removing the sequence;

Removing the sequence when the proportion of the sequence into which the one input code is input and the average sequence value input at this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And

Removing the sequence if the mapping value of the fitting function to the input environment at which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold; It includes at least one of.

Preferably, the error correction method,

Determining a user group in which the user is located according to a feature of the user triggering the button; And

Optimizing the error correction method according to the user group information.

Preferably, the error correction method,

Acquiring user's input habit record information; And

Optimizing the error correction method according to the input habit record information.

According to another aspect of the present invention, a user input speed based input method error correction device further provided, the error correction device,

An acquisition device configured to acquire an input speed of a triggered current independent button;

An adjusting device configured to adjust weighting when correcting an input code corresponding to the current independent button with an input code corresponding to the other button according to the input speed;

An acquisition device configured to perform error correction on an input code corresponding to the current independent button based on the adjusted weighting to obtain at least one error corrected input code; And

And a determination device configured to determine at least one candidate word corresponding to the entire input, according to the error-corrected at least one input code corresponding to each button of one full input.

Preferably, the error correction device,

Acquiring an average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code corresponding to the one entire input, and then determining the at least one candidate word corresponding to the entire input. And an ordering device configured to adjust the ordering.

Preferably, the ordering device,

A calculating unit, configured to calculate the average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code;

A first determination unit configured to determine a corresponding error correction model based on the average input speed; And

And a first adjustment unit configured to adjust an ordering arrangement of the at least one candidate word corresponding to the entire input according to the error correction model.

Preferably, according to the input speed, adjusting the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other button,

Increasing weighting when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;

Reducing the weight of itself when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;

Increasing the weight of itself when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; And

Reducing weighting when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; It includes at least one of.

Preferably, the determination of the difficulty of the button touch error,

Level class determination according to the input speed of the independent button;

Rating weighting according to the input speed of the independent button;

Determining a rating mapping to an input speed of the independent button and a fitting function according to the current input environment; And

A rating weighting decision mapping to an input speed of the independent button and a fitting function according to the current input environment; It includes at least one of.

Preferably, the error correction model,

An error correction level model independently based on the average input speed;

An error correction rating model independently based on the average input speed;

An error correction level model based on the average input speed and the remaining environmental parameter fittings; And

An error correction rating model based on the average input speed and the remaining environmental parameter fittings; It includes at least one of.

Preferably, the ordering device,

A noise canceling unit configured to perform a noise canceling process on a sequence consisting of a sequence into which each input code of the entire input codes is input, to obtain a noise canceled sequence;

A second confirmation unit, configured to confirm an input time of the entire input code according to the noise canceled sequential sequence; And

A second adjustment unit, configured to obtain the overall input speed of the full input code based on an input time of the full input code, and adjust an ordering arrangement of the at least one candidate word corresponding to the full input; do.

Preferably, the noise canceling process,

If the sequence into which the one input code is input is greater than the predetermined sequential upper limit threshold or less than the predetermined sequential lower limit threshold, removing the sequence;

Removing the sequence when the proportion of the sequence into which the one input code is input and the average sequence value input at this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And

Removing the sequence if the mapping value of the fitting function with the input environment at which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold; It includes at least one of.

Preferably, the error correction device,

Determine the user group in which the user is located according to the characteristics of the user triggering the button,

It is further configured to optimize the error correction device according to the user group information.

Preferably, the error correction device,

Obtain the user's input habit record information,

It is further configured to optimize the error correction device according to the input habit record information.

Preferably, the present invention further provides a computer readable storage medium, wherein the computer code is stored in the computer readable storage medium, and when the computer code is executed, any method described above is executed.

Preferably, the present invention further provides a computer program product, wherein any of the methods described above are executed when the computer program product is executed by a computer device.

Preferably, the present invention further provides a computer device, wherein the computer device,

One or multiple processors; And

A storage device configured to store one or multiple computer programs; Including;

When the one or more computer programs are executed by the one or more processors, let the one or more processors implement any of the methods described above.

Compared with the existing technology, the present invention has the following advantages.

The present invention analyzes big data and mines the relationship between user input speed and error correction in terms of time, and helps an error correction model to perform accurate error correction. The present invention continuously acquires an input speed when a user inputs each independent button, and converts an input code corresponding to the current independent button into an input code corresponding to the other button according to the touch error rate corresponding to the input speed. By switching correspondingly to adjust the weighting when correcting, and evaluating the input method in terms of time using the input speed as one element of correcting the input code through the error correction model, the user Further correcting the input code determined through this spatial dimension, such as the starting point of the triggered keypad, improves the accuracy of error correction with the error correction model and improves the quality of the user input.

Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments, which proceeds with reference to the accompanying drawings.
1 is an exemplary configuration diagram of an input method error correction apparatus based on a user input speed according to an aspect of the present invention.
2A illustrates a relationship between the touch error rate and an input time of an independent button.
2B is an example of an input display.
3 is an exemplary flowchart of an input method error correction method based on a user input speed according to an aspect of the present invention.
In the accompanying drawings, the same or similar reference numerals represent the same or similar components.

Prior to a more detailed discussion of example embodiments, reference is made to some example embodiments described by the process or method shown in the flowchart. Although each operation is described as a sequential process in the flowchart, many of the operations can be performed in parallel, in parallel, or simultaneously. In addition, the order of each operation may be rearranged. When these operations are completed, the process may end, but additional steps not shown in the accompanying drawings may be further included. The processing may correspond to a method, a function, a procedure, a subroutine, a subprogram, or the like.

A “computer device” (also referred to as a “computer”), as referred to in this context, refers to a smart electronic device that runs a predefined program or instruction to perform a predefined process, such as numerical computing and / or logic computing, which is a processor. And a storage device, wherein the processor executes a predetermined process by executing instructions stored in the storage device in advance, or executes a predetermined process with hardware such as an ASIC, FPGA, or DSP, or both. It can be implemented in combination. Computer devices include, but are not limited to, servers, personal computers, laptop computers, tablet PCs, smart phones, and the like.

The computer device includes a user device and a network device. Here, the user device includes, but is not limited to, a computer, a smart phone, a PDA, and the like; The network device includes, but is not limited to, a single network server, a server group composed of a plurality of network servers, or a cloud composed of a large number of computing devices or network servers based on cloud computing. Is one type of distributed computing, which is a super virtual computing device composed of a set of disjoint groups of computing devices. Here, the computer device may operate independently to implement the present invention, or the present invention may be implemented through interaction with other computer devices in the network by introducing a network. Here, the network in which the computer device is located includes, but is not limited to, the Internet, WAN, MAN, LAN, VPN network, and the like.

The user device, the network device and the network and the like are merely examples, and other existing or future computer devices or networks should also be included within the protection scope of the present invention when applicable to the present invention, and are incorporated herein by reference. I want to explain.

The methods to be discussed below, some of which are shown in flow charts, may be implemented in hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented in software, firmware, middleware or microcode, the program code or code segments for performing the necessary tasks may be stored in a machine or computing device readable medium (eg, storage medium). One or more processors can perform the necessary tasks.

The details of the specific structures and functions disclosed herein are merely representative and are used for the purpose of describing exemplary embodiments of the invention. However, the present invention may be specifically implemented through various alternative forms, and should not be construed as limited only to the embodiments mentioned herein.

Although the terms "first", "second", and the like available herein can be used to describe each unit, it should be understood that such units are not limited to these terms. This term is only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as a second unit, and similarly, the second unit may be referred to as a first unit. The term "and / or" as used herein includes any or all combinations of one or a plurality of related items listed.

When one unit is described as being "connected" or "coupled" to another unit, it should be understood that it may be directly connected or coupled to the other unit, or there may be an intermediate unit. Conversely, if one unit is described as being "directly connected" or "directly coupled" to another unit, no intermediate unit is present. Other words to describe the relationship between units (for example, "between" and "directly between ..." and "adjacent to ..." and "directly adjacent to ..." ) Should be interpreted in a similar manner.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. Unless the context clearly indicates otherwise, the singular forms “a”, “an” and “an” are intended to include the plural forms as well. The terms "comprises" and / or "includes" as used herein only define the presence of the stated feature, integer, step, operation, unit and / or part, and one or more other features, integers, steps, It is to be understood that this does not exclude the presence or addition of operations, units, components and / or combinations thereof.

In addition, in some alternative implementation manners, it is to be understood that the functions / acts mentioned may occur in a different order than that shown in the accompanying drawings. For example, depending on the function / operation involved, the two figures shown sequentially may actually be executed simultaneously simultaneously or sometimes in the reverse order.

1 is an exemplary configuration diagram of an input method error correction apparatus based on a user input speed according to an aspect of the present invention. The error correction device 1 includes an acquisition device 101, an adjustment device 102, an acquisition device 103, and a determination device 104.

Here, the error correction apparatus 1 is located at, for example, a user device, and the user device includes, but is not limited to, a personal computer, a laptop computer, a tablet PC, a smart phone, a PDA, and the like.

The user device is only an example, and existing or future user device should be included within the protection scope of the present invention, if applicable to the present invention, and should be included in the present application. I will understand. The user device described herein may refer to a smart electronic device that runs a predetermined program or instruction to perform a predetermined processing process, such as numerical computing and / or logic computing, which may include a processor and a storage device. The predetermined processing may be executed by executing instructions stored in the storage device in advance, or the predetermined processing may be performed by hardware such as an ASIC, FPGA, or DSP, or a combination of the above may be implemented.

Here, the acquisition apparatus 101 acquires the input speed of the triggered current independent button. Specifically, when the user proceeds with input, the user presses any one button of the keypad of the user device, wherein the keypad includes a physical keypad or a virtual keypad on the screen of the user device. When pressed or touched, the acquiring device 101 may interact with the corresponding user device, for example, through one or more calls to an application program interface (API) provided by the input application. Acquiring the input speed of the current independent button collected by the corresponding user collected, the acquisition apparatus 101 according to the input speed,

Those skilled in the art should understand that the touch error rate corresponding to the input speed is merely an example, and in fact, that the mapping relationship corresponding to the input speed and the touch error rate is a touch error value calculated by analyzing big data. do.

The adjusting device 102 adjusts the weighting when correcting the input code corresponding to the current independent button to the input code corresponding to the other button according to the input speed. Specifically, through a research experiment, the touch error rate corresponding to the input time of the independent button is statisticized by setting the time when each button is independently input as a grouped container. Here, the input time of the independent button is a time interval in which two adjacent independent buttons are triggered. 2A illustrates a relationship between the touch error rate and an input time of an independent button. The input speed of the independent button can be calculated according to the input time of the independent button, and the input speed is 1 / input time, and as shown in FIG. 2A, the slower the input speed of the independent button is, ie, the input of the independent button. As time elapses, the touch error rate gradually decreases, i.e., the faster the user input speed, the easier the typo, and thus the adjusting device 102 generates an input code corresponding to the current independent button according to the input speed. The weighting at the time of correcting with an input code corresponding to a button is adjusted, and the weighting is applied to an existing error correction model to assist the existing error correction model to perform error correction, thereby improving the accuracy of the user input. For example, when the input speed of any independent button is relatively fast, the adjusting device 102 weights (w) to correct the input code corresponding to the current independent button to the input code corresponding to the other button according to the input speed. Increase); Alternatively, if the input speed of any independent button is relatively slow, the adjusting device 102 adjusts the weight w to correct the input code corresponding to the current independent button to the input code corresponding to the other button according to the input speed. Decreases. Here, the specific value of the weight (w) can be set through a predetermined rule, for example, the total weight (tw) is set to 1, w is any number between 0-1, and the specific value of w The figures are fixed according to the rules of the prescribed.

Here, adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other buttons according to the input speed may include: 1) an input code corresponding to a button prone to a touch error. Increase the weighting when correcting with the input code corresponding to the guitar button; 2) reduce the weighting itself when correcting the input code corresponding to the button which is prone to touch error with the input code corresponding to the other button; 3) increase the weight of itself when correcting the input code corresponding to the button which is hard to generate touch error with the input code corresponding to the other button; 4) reducing the weighting when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; but is not limited thereto. Specifically, when the input speed is relatively high, for example, when the input speed of the current independent button exceeds a predetermined speed threshold value v _f , the current independent button is a button that is prone to a touch error. To increase the weighting to correct the touch button, more likely to touch the button with the other buttons, and reduce the weight of the current independent button itself. Here, the weight is recorded as a first weight, and the first weight refers to the weight of the current independent button itself and the weight of correcting the current independent button with other buttons. For example, with the QWERTY full keypad, for example, if the acquired user current independent button is “r” and the current independent button “r” at this time is a button that is prone to a touch error, that is, the current If the input speed of the independent button of is relatively large, the button that the user wants to trigger decreases the first weight of the "r" itself, and the buttons "e" and "f around the button" r "that the user wants to trigger "," increase the first weight of "t", respectively. Here, the button to be triggered by the user is the user's purpose button, the current independent button is a trigger button acquired by the user device, and the current independent button may not be the user's purpose button. According to the first weight of the button, the current independent button itself also has a corresponding first weight so as to determine the user's destination button by combining other parameters. On the other hand, when the input speed of the current independent button is lower than the predetermined speed threshold value v _s , that is, when the input speed of the current independent button is relatively small, the current independent button may cause a touch error. It is a difficult button and increases the first weight of the button itself, which is difficult to cause touch error, and reduces the first weight when correcting the input code corresponding to the button which is less likely to occur touch error to the input code corresponding to the other button. do. For example, if the acquired user current independent button is "i", and the current independent button "i" at this time is a button which is hard to generate a touch error, the button which the user wants to trigger is "i" itself. The first weight is increased, and the first weights of the buttons “u”, “k”, and “o” around the button “i” that the user intends to trigger are respectively decreased.

Preferably, the process of determining the difficulty of the button touch error includes, but is not limited to:

1) Level level determination according to the input speed of the independent button, for example, the level of the input speed of the independent button according to the size range of the input speed is divided into a high level and a low level, the input speed of the independent button is a high level When positioned at, the independent button acquires a second weight corresponding to a high level, and the independent button is a button that is less likely to cause a touch error or a button that is prone to a touch error based on the second weight. Can be determined. When the input speed of the independent button is located at a low level, the independent button acquires a second weight corresponding to the low level, and is the independent button a button that is hard to generate a touch error based on the second weight? Or you can determine if the button is prone to touch errors. Here, the level classification that proceeds with respect to the input speed of the independent button according to the size range of the input speed may exist in various ways.

2) Determination of the weighting weight according to the input speed of the independent button, specifically, when the user triggers one independent button, the acquisition apparatus 101 acquires the input speed of the triggered current independent button, Depending on the input speed, the corresponding independent button is scored. For example, by setting a function having an input speed as a variable, calculating a score corresponding to each different input speed, obtaining a second weighting corresponding to the input speed of the independent button according to the score, and obtaining the second weight. According to the weight, it is determined whether the independent button is a button that is prone to a touch error or a button that is hard to generate a touch error.

3) determining a rating mapping to the input speed of the independent button and the fitting function according to the current input environment, specifically generating a fitting function according to these two parameters such as the input speed of the independent button and the current input environment, It is determined comprehensively at which level the independent button is located, and it is determined whether the current independent button is a button that is prone to a touch error or a button where a touch error is less likely to occur according to the level at which the independent button is located. Here, the current input environment is, for example, whether there is enough light, whether the current input device is a computer, a tablet or a mobile phone, whether the user device is in a stationary state or a mobile state, and how fast is the moving speed. .

4) A rating weighting decision mapping to the input speed of the independent button and the fitting function according to the current input environment, and specifically, the independent button is rated according to the input speed of the independent button and the current input environment. For example, by setting a fitting function that takes an input speed and a current input environment as variables, a score corresponding to each of the independent buttons is calculated under different input speeds and different input environments, and the input speed of the independent buttons according to the scores. A second weight corresponding to is obtained, and it is determined whether the independent button is a button that is prone to a touch error or a button that is hard to generate a touch error according to the second weighting.

The method of adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other button according to the input speed is merely an example, and the input speed to be displayed in other existing or future devices. According to the present invention, the method of adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other button should also be included within the protection scope of the present invention. It should be understood by those skilled in the art that it should be incorporated by reference.

The obtaining device 103 performs error correction on the input code corresponding to the current independent button based on the adjusted weighting, and acquires at least one error corrected input code. Specifically, the obtaining device 103 modifies the corresponding error correction model of the input method corresponding to the error correction device 1 based on the weighting, where the weighting is the first weighting. Using the input speed of the current independent button triggered as one parameter of the error correction model, when the error correction model proceeds with error correction, the evaluation and calculation are carried out in terms of time. In connection with an already triggered recording button that is not input to at least one top screen belonging to the first full input, error correction is performed on the input code corresponding to the current independent button, and at least one error corrected input code is entered. Acquire. Here, the one full input is configured by adding an already triggered recording button not input to the upper screen to the current independent button, that is, all input codes not input to the upper screen, one independent button each time. After further inputting, all changes occur in candidate words corresponding to already triggered recording buttons that are not entered in the upper screen. For example, as shown in the input process shown in FIG. 2B, FIG. 2B is an example of an input display, where one full input points to “shoifashi” and the user does not enter on the top screen and continues to “zainali”. If you enter, the whole input at this time is shoifashizainali ".

For example, when the weight when the input code corresponding to the current independent button adjusted by the adjusting device 102 is corrected to the input code corresponding to the other button is relatively large, the error correction device 1 The ratio of correcting the input code corresponding to the independent button with the other input code is relatively large, and the probability of correcting the current independent button with the guitar button with the error correction model corresponding to the input method is improved. Selecting a suitable existing error correction model according to the adjusted weighting, and using the adjusted weighting to the existing error correction model, performing error correction with other parameters in the existing error correction model, that is, time Proceed with error correction at the dimension. As an important consideration factor of the existing error correction model, the input speed of the current independent button is introduced into the existing error correction model, so that error correction is correctly performed on the input code, and the acquisition device 103 uses the error correction model for error. Obtain a corrected at least one input code. Here, the error corrected at least one input code may still be an input code corresponding to the current independent button, and the input code corresponding to the current independent button may be corrected to the input code corresponding to the other button. For example, if the input interface is a touch-type virtual QWERTY full keypad interface, the error correcting apparatus 1 may be configured to display all the keys on the keypad according to the starting point position L of the keypad triggered when the user proceeds with the input. Calculate the distances (s1, s2 ... sn) from the button center point to the corresponding starting point position (L), and calculate the probability that the current independent buttons triggered by the user are each button according to the distance; And then acquire an input code corresponding to the current independent button triggered by the user according to the probability or by attaching other parameters is "s" and the input speed of "s" is 9 characters / second, 101) acquires that the input speed of the input code “s” is 9 characters / second, and sets the one input speed threshold to 5 characters / second, that is, when the input speed of the input code is larger than the corresponding input speed threshold, that is, The input speed is lower In the case of the level, the weight of correcting the input code corresponding to the current independent button to the input code corresponding to the other button is increased, and the input code corresponding to the current independent button is set to "s". Reducing the weighting to be corrected, i.e., if the user's destination button is still a weighting of " s " itself, the adjusting device 102 determines that, according to the input speed, the input speed is greater than an input speed threshold, In this case, the weight for correcting “s” to “w” is increased to w1, the weight for correcting “s” to “a” is increased to w2, the weight to correct to “d” is increased to w3, the weight correcting to "x" is increased to w4, the weight of the user's destination button is still "s" itself to be reduced to w5, and the acquiring device 103 has weights w1, w2, w3, w4, w5 and Current Independent Button Based on the previous record button, the current Independent button is changed to Other. It is judged to determine the probability. For example, based on the calculation, the probability of correcting "s" to "w", "a", "d", and "x", respectively, is P1, P2, P3, and P4, respectively. Is a probability P5, and the user's purpose button is determined to be "a" in accordance with P1, P2, P3, P4 and P5. The probability of correcting the current independent button with the other button is comprehensively determined based on the recording button which is not input to the upper screen before the current independent button. In connection with the error correction method of the existing error correction model, the error correction is jointly performed on the input code corresponding to the current independent button. If " p " and " s " are both Virgins, the weighting to correct the letter " s " to mica " a " adjacent to " s "

The method of deciding how to adjust the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other button according to the above-described input speed threshold value is merely an example, and other existing or The method of determining how to adjust the weighting when the input code corresponding to the current independent button to be corrected to the input code corresponding to the other button to be shown in the future is also included within the protection scope of the present invention. It should be understood by those skilled in the art that the present invention should be incorporated herein by reference. Here, the acquiring device 103 acquires at least one error corrected input code, registers it in the system background, does not display it on the input interface, that is, what is displayed on the input interface still corresponds to the user triggered button. Is the input code.

Here, the big data is analyzed and the relationship between the speed of user input and the error correction is mined in terms of time, and an error correction model is used to perform accurate error correction. The error correction model is obtained by continuously acquiring the input speed at the time of inputting the button and adjusting the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other buttons according to the input speed. As an element of correcting the input code, the input method is evaluated in terms of time using the input speed, and the input code determined through this spatial dimension such as the starting point of the keypad triggered by the user is evaluated. To improve the accuracy of error correction with the error correction model, It improves quality.

(두려워할 파),

(긁을 파),

(긁을 파),

(머리띠 파)......를 확정한다. 나아가, 확정 장치(104)는 여전히 오류 정정되기 전의 적어도 하나의 입력 코드에 따라, 대응되는 적어도 하나의 후보 단어를 확정하며, 오류 정정되기 전의 적어도 하나의 입력 코드에 대응되는 적어도 하나의 후보 단어의 우선 순위는 오류 정정된 적어도 하나의 입력 코드에 대응되는 적어도 하나의 후보 단어보다 낮다. 예를 들어, 오류 정정되기 전의 입력 코드 “ps”에 대응되는 적어도 하나의 후보 단어“

(평술)", "

(평시)", "

(촬영)" 등의 우선 순위는 “

(두려워할 파),

(긁을 파),

(긁을 파),

(머리띠 파)......”보다 낮다. The determination device 104 determines the at least one candidate word corresponding to the entire input according to the error-corrected at least one input code corresponding to each button of one entire input. Specifically, after acquiring at least one error corrected input code by the acquiring device 103 sequentially, and acquiring an error corrected input code corresponding to each button of one full input with the acquiring device 103, The determination device determines the at least one candidate word corresponding to the entire input. For example, following the above example, after correcting the letter "s" to mica "a" adjacent to "s" on the keypad, the letter "a" and the previous letter "p" are combined into "pa" and The determining device 104 may include at least one candidate word corresponding to “pa”:

(Waves to be afraid),

(Scraping),

(Scraping),

Confirm the (headband wave) ...... Furthermore, the determining device 104 still confirms the corresponding at least one candidate word according to the at least one input code before the error correction, and determines the at least one candidate word corresponding to the at least one input code before the error correction. The priority is lower than at least one candidate word corresponding to the error corrected at least one input code. For example, at least one candidate word corresponding to the input code “ps” before error correction.

(Flat) ","

(Normal) ","

(Shooting) ", etc.

(Waves to be afraid),

(Scraping),

(Scraping),

(Headband wave) ...... ”is lower than.

Here, the input process is continuous, and if the user constantly inputs and does not input the upper screen in the process of inputting the input code, the input code included in the current one full input is constantly increasing and newly After the added input code is error corrected, the write input code which is not input to the upper screen is continually attached so that a dynamic change occurs in at least one candidate word corresponding to the entire input.

Preferably, the error correction device 1 further includes a sequence arrangement device (not shown).

The ordering device obtains an average input speed of a button corresponding to each input code among the entire input codes according to the input time of the entire input code corresponding to the entire input, and the at least one corresponding to the entire input. Adjust the ordering arrangement of candidate words. Specifically, when a user's one full input is completed, the ordering device is determined according to the input time of the full input code corresponding to the one full input acquired and the quantity of the input codes during the one full input process. An average input speed of a button corresponding to each input code of one full input code is calculated, and then the order arrangement of the at least one candidate words corresponding to the entire inputs is adjusted according to the average input speed.

Preferably, the ordering arrangement includes a calculating unit (not shown), a first definitive unit (not shown), and a first adjusting unit (not shown).

Here, the calculating unit calculates the average input speed of the button corresponding to each input code of the entire input codes according to the input time of the entire input code. Specifically, the calculating unit calculates the average input speed of the button corresponding to each input code of the entire input codes according to the input time of the entire input code and the quantity of input codes included in the entire input code.

(수발실),

(발표식),

(수법은)...... 등 "shoufashi"과 높은 관련도를 구비하는 일 유형의 후보 단어를 후보 단어 목록의 앞에 위치시키고, 오류 정정 처리를 진행하지 않은 초기 전체 입력 코드 "shoifashi"에 대응되는 후보 단어:

,

......등을 후보 단어 목록의 뒤에 위치시킨다. 마찬가지로, 상기 평균 입력 속도가 로우 레벨에 위치하면, 순서 배열을 교환한다. 상기 오류 정정 모델이 독립적으로 상기 평균 입력 속도를 기반으로 하는 오류 정정 평점 모델을 채용할 경우, 상기 오류 정정 평점의 점수는 상기 평균 입력 속도의 연속적인 선형 함수 또는 비 선형 함수에 관한 것으로, 상기 평균 입력 속도의 변화에 따라 상이하며, 제1 조정 유닛은 상기 입력 속도의 크기에 따라 대응되는 오류 정정 점수를 획득하고, 이어서 상기 점수에 따라 상기 전체 입력 중의 각 입력 코드가 취득 장치(101)가 제일 처음으로 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하여, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다. 상기 오류 정정 모델이 상기 평균 입력 속도 및 나머지 환경 파라미터 피팅을 기반으로 하는 오류 정정 레벨 모델을 채용할 경우, 상기 오류 정정 레벨은 평균 입력 속도 및 나머지 환경 파라미터 피팅에 따라 대응되는 오류 정정 레벨을 확정하기 위한 것으로, 상기 오류 정정 레벨에 따라 상기 전체 입력 중의 각 입력 코드가 취득 장치(101)가 제일 처음으로 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하여, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다. 상기 오류 정정 모델이 상기 평균 입력 속도 및 나머지 환경 파라미터 피팅을 기반으로 하는 오류 정정 평점 모델을 채용할 경우, 상기 오류 정정 평점의 점수는 상기 평균 입력 속도와 나머지 환경 파라미터 피팅의 연속적인 선형 함수 또는 비 선형 함수에 관한 것으로, 상기 평균 입력 속도의 변화 또는 나머지 환경 파라미터의 변화에 따라 상이하며, 이어서 상기 점수에 따라 상기 전체 입력 중의 각 입력 코드가 취득 장치(101)가 제일 처음으로 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하여, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다.The first determination unit determines a corresponding error correction model based on the average input speed, wherein the error correction model comprises: 1) an error correction level model independently based on the average input speed; 2) an error correction rating model independently based on the average input speed; 3) an error correction level model based on the average input speed and the remaining environmental parameter fittings; And 4) an error correction rating model based on the average input speed and the remaining environmental parameter fittings. The first adjustment unit is configured to adjust the ordered arrangement of the at least one candidate word corresponding to the entire input according to the error correction model. Specifically, when the error correction model independently adopts an error correction level model based on the average input speed, for example, a level may be independently high level, middle level, and low level based on the average input speed. When the average input speed is located at a low level, the candidate words corresponding to the input code, which is the first input code acquired by the acquisition apparatus 101, are placed in front of each input code of the entire inputs. Each input code in the entire input increases the weight which is the first input code acquired by the acquiring device 101, and decreases the weight that is the input code after each input code in the entire input is error corrected, respectively, When the average input speed is at the high level, the input after each input code of the total inputs is error corrected. The weight of an input code after each input code of the entire inputs is each error-corrected is increased so that a candidate word corresponding to the input code, which is a code, is placed in front of each other. Reduce weighting, which is the first acquired input code. For example, if the user enters “shoifashi” and the average input speed is at a high level, the probability that an error is present during this entire input is relatively high. In this case, the error corrected input code " candidate words for "shoufashi":

(Weather room),

(Presentation),

One method of having a high degree of relevance to "shoufashi", etc., is placed in front of the candidate word list, and the initial full input code "shoifashi" has not been processed for error correction. Matching candidate words:

,

Place ...... and so on after the list of candidate words. Likewise, if the average input speed is at a low level, the ordered arrangement is exchanged. When the error correction model independently adopts an error correction rating model based on the average input speed, the score of the error correction rating relates to a continuous linear function or a non-linear function of the average input speed, wherein the average The first adjustment unit obtains a corresponding error correction score according to the magnitude of the input speed, and then, according to the score, each input code of the entire input is obtained by the acquiring device 101 first. The weight which is the first input code and the weight which is the input code after each input code in the entire inputs are each error-corrected are adjusted to determine the order sequence in the candidate word list corresponding to both. When the error correction model adopts an error correction level model based on the average input speed and the remaining environmental parameter fitting, the error correction level determines the corresponding error correction level according to the average input speed and the remaining environmental parameter fitting. Weighting each input code of the entire inputs according to the error correction level as an input code first acquired by the acquisition apparatus 101 and an input code after each input code of the entire inputs is error corrected, respectively. Is adjusted to determine a sequence of candidate words corresponding to both in the candidate word list. If the error correction model employs an error correction rating model based on the average input speed and the remaining environmental parameter fitting, the score of the error correction rating is a continuous linear function or ratio of the average input speed and the remaining environmental parameter fitting. A linear function, which is different depending on a change in the average input speed or a change in the remaining environmental parameters, and then, according to the score, each input code in the total input is an input code acquired first by the acquiring device 101. And weighting, which is an input code after each input code of the entire inputs has been error corrected, respectively, so that candidate words corresponding to both are arranged in the candidate word list.

Preferably, the ordering device further comprises a noise canceling unit (not shown), a second deciding unit (not shown), and a second adjusting unit (not shown).

The noise removing unit performs a noise removing process on a sequence consisting of a sequence into which each input code of the entire input code is input, to obtain a noise canceled sequence. Specifically, in one full input of a user, a noise canceling process is performed on one sequence consisting of a sequence in which each input code is input, for example, after the user inputs one input code, If the input process stops due to any change, the sequence corresponding to the input code far exceeds the predetermined upper limit threshold of the input when normally inputted, and the sequence corresponding to the input code is removed from the sequence. Preferably, the noise canceling process comprises: 1) removing the sequence if a sequence into which one input code is input is greater than a predetermined upper limit threshold or less than a predetermined lower limit threshold; 2) removing the sequence if the proportion of the sequence into which the one input code is input and the average sequence value entered this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And 3) removing the sequence if the mapping value of the fitting function to the input environment at which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold. It includes at least one of.

The second confirmation unit confirms the input time of the entire input code according to the noise canceled sequential sequence. Specifically, the input time of the entire input code is equal to the sum of all sequences in the noise canceled sequence, and the quantity of the entire input code should be subtracted from the quantity of sequences removed through the noise removal process.

The second adjustment unit is an ordered arrangement of the at least one candidate word corresponding to the full input by obtaining the full input speed of the full input code based on the input time of the full input code. Specifically, the second adjustment unit obtains the average input speed of the entire input code based on the input time of the entire input code and the quantity of the input code after correspondingly subtracting the quantity of the noise canceling input code. Adjust the ordering arrangement of the at least one candidate word corresponding to the entire input.

Preferably, the error correction device 1 further includes a first optimization device (not shown). The first optimization device determines a user group in which the user is located according to a feature of the user triggering the button; The error correction apparatus 1 is optimized according to the user group information. Specifically, the error correction apparatus 1 creates an image for the feature according to each feature of the user through a predetermined time of learning, and a user group corresponding to the image through big data, for example, a young male enjoying fast input, Mining middle-aged women, etc. who enjoy serious input, but each feature of the user includes, but is not limited to, the overall input speed of the user, the input speed of each specific button of the user, the words he or she made, the user's frequent input Do not. The first optimization device optimizes the error correction device 1 by proceeding with the threshold adjustment of each term with respect to the error correction device 1 through a feedback mechanism.

The error correction device 1 further includes a second optimization device (not shown). The second optimization device acquires user input habit record information; Optimize the error correction device according to the input habit record information. Specifically, the error correction apparatus 1 determines the user's input habits and methods, such as one-handed or two-handed input habits, a horizontal screen or a vertical screen input habit, etc. according to the user's input starting point situation through a predetermined time learning. However, the input start point situation of the user includes, but is not limited to, the degree of dispersion of the input start point of the user, a speed of inputting a specific button, a touch error rate of inputting a specific button, and the like. The second optimization device acquires such input habit record information of the user, and adjusts the threshold value of each term in the error correction device 1 in accordance with its input habit and method to optimize the error correction device 1.

3 is an exemplary flowchart of an input method error correction method based on a user input speed according to an aspect of the present invention.

Here, in step S301, the input speed of the current independent button triggered by the error correction apparatus 1 is acquired. Specifically, when the user proceeds with input, the user presses any one button of the keypad of the user device, wherein the keypad includes a physical keypad or a virtual keypad on the screen of the user device. In the case of pressing or touching, in step S301, the user corrects the interaction with the user device to the error correction device 1, for example, once or multiple times for the application program interface (API) provided by the input application. Acquire the input speed of the current independent button triggered by the corresponding user collected by the sensor of the device, and in step S301, according to the input speed to the error correction device 1,

Those skilled in the art should understand that the touch error rate corresponding to the input speed is merely an example, and in fact, that the mapping relationship corresponding to the input speed and the touch error rate is a touch error value calculated by analyzing big data. do.

In step S302, the error correction apparatus 1 adjusts the weight when correcting the input code corresponding to the current independent button with the input code corresponding to the other button according to the input speed. Specifically, through a research experiment, the touch error rate corresponding to the input time of the independent button is statisticized by setting the time when each button is independently input as a grouped container. Here, the input time of the independent button is a time interval in which two adjacent independent buttons are triggered. 2A illustrates a relationship between the touch error rate and an input time of an independent button. The input speed of the independent button can be calculated according to the input time of the independent button, and the input speed is 1 / input time, and as shown in FIG. 2A, the slower the input speed of the independent button is, ie, the input of the independent button. As the time is longer, the touch error rate gradually decreases, i.e., the faster the user input speed is, the easier the typo is. Therefore, in step S302, the current independent button according to the input speed to the error correction device 1 is corrected. Adjust the weighting when correcting the input code corresponding to the input code corresponding to the other button, and apply the weighting to the existing error correction model to assist the existing error correction model to perform error correction. Improve accuracy. For example, if the input speed of any independent button is relatively fast, in step S302, the error correcting apparatus 1 inputs an input code corresponding to the current independent button according to the input speed corresponding to the other button. Increase the weight w that corrects with the code; Alternatively, if the input speed of any independent button is relatively low, in step S302, the error correction device 1 converts the input code corresponding to the current independent button into the input code corresponding to the other button according to the input speed. Reduce the weight w to correct. Here, the specific value of the weight (w) can be set through a predetermined rule, for example, the total weight (tw) is set to 1, w is any number between 0-1, and the specific value of w The figures are fixed according to the rules of the prescribed.

Here, the step of adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other buttons according to the input speed may include: 1) an input corresponding to a button that is prone to a touch error. Increasing the weighting when correcting the code to an input code corresponding to the guitar button; 2) reducing the weight of itself when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button; 3) increasing the weight of itself when correcting the input code corresponding to the button which is less likely to cause a touch error to be input code corresponding to the other button; And 4) reducing weighting when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button. Specifically, when the input speed is relatively high, for example, when the input speed of the current independent button exceeds a predetermined speed threshold value v _f , the current independent button is a button that is prone to a touch error. To increase the weighting to correct the touch button, more likely to touch the button with the other buttons, and reduce the weight of the current independent button itself. Here, the weight is recorded as a first weight, and the first weight refers to the weight of the current independent button itself and the weight of correcting the current independent button with other buttons. For example, with the QWERTY full keypad, for example, if the acquired user current independent button is “r” and the current independent button “r” at this time is a button that is prone to a touch error, that is, the current If the input speed of the independent button of is relatively large, the button that the user wants to trigger decreases the first weight of the "r" itself, and the buttons "e" and "f around the button" r "that the user wants to trigger "," increase the first weight of "t", respectively. Here, the button to be triggered by the user is the user's purpose button, the current independent button is a trigger button acquired by the user device, and the current independent button may not be the user's purpose button. According to the first weight of the button, the current independent button itself also has a corresponding first weight so as to determine the user's destination button by combining other parameters. On the other hand, when the input speed of the current independent button is lower than the predetermined speed threshold value v _s , that is, when the input speed of the current independent button is relatively small, the current independent button may cause a touch error. It is a difficult button and increases the first weight of the button itself, which is difficult to cause touch error, and reduces the first weight when correcting the input code corresponding to the button which is less likely to occur touch error to the input code corresponding to the other button. do. For example, if the acquired user current independent button is "i", and the current independent button "i" at this time is a button which is hard to generate a touch error, the button which the user wants to trigger is "i" itself. The first weight is increased, and the first weights of the buttons “u”, “k”, and “o” around the button “i” that the user intends to trigger are respectively decreased.

Preferably, the process of determining the difficulty of the button touch error includes, but is not limited to:

1) Level level determination according to the input speed of the independent button, for example, the level of the input speed of the independent button is divided into high level and low level according to the size range of the input speed, the input speed of the independent button is a high level When located at, the independent button acquires a second weight corresponding to a high level, and the independent button is a button that is less likely to cause a touch error based on the second weight, or is a button that is prone to a touch error. Can be determined. When the input speed of the independent button is located at a low level, the independent button acquires a second weight corresponding to the low level, and is the independent button a button that is hard to generate a touch error based on the second weight? Or you can determine if the button is prone to touch errors. Here, the level classification that proceeds with respect to the input speed of the independent button according to the size range of the input speed may exist in various ways.

2) Determining a rating weighting according to the input speed of the independent button, specifically, when the user triggers one independent button, in step S301, the input speed of the current independent button triggered by the error correction device 1 is determined. It acquires and progresses a rating with respect to the said independent button according to the input speed of the said independent button. For example, by setting a function having an input speed as a variable, calculating a score corresponding to each different input speed, obtaining a second weighting corresponding to the input speed of the independent button according to the score, and obtaining the second weight. According to the weight, it is determined whether the independent button is a button that is prone to a touch error or a button that is hard to generate a touch error.

3) a rating determination for mapping to the fitting speed according to the input speed and the current input environment of the independent button, specifically, to generate the fitting function according to these two parameters such as the input speed and the current input environment of the independent button, It is determined comprehensively which level of the independent button is located, and determines whether the current independent button is a button that is prone to a touch error or a button which is difficult to generate a touch error according to the level at which the independent button is located. Here, the current input environment is, for example, whether there is enough light, whether the current input device is a computer, a tablet or a mobile phone, whether the user device is in a stationary state or a mobile state, and how fast is the moving speed. .

4) A rating weighting decision mapping to the input speed of the independent button and the fitting function according to the current input environment, and specifically, the independent button is rated according to the input speed of the independent button and the current input environment. For example, by setting a fitting function that takes an input speed and a current input environment as variables, a score corresponding to each of the independent buttons is calculated under different input speeds and different input environments, and the input speed of the independent buttons according to the scores. A second weight corresponding to is obtained, and it is determined whether the independent button is a button that is prone to a touch error or a button that is hard to generate a touch error according to the second weighting.

The method of adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other button according to the input speed is merely an example, and the input speed to be displayed in other existing or future devices. According to the present invention, the method of adjusting the weighting when the input code corresponding to the current independent button is corrected to the input code corresponding to the other button should also be included within the protection scope of the present invention. It should be understood by those skilled in the art that it should be incorporated by reference.

In step S303, the error correction apparatus 1 performs error correction on the input code corresponding to the current independent button based on the adjusted weighting, to obtain at least one error corrected input code. Specifically, in step S303, the error correction apparatus 1 modifies the corresponding error correction model of the input method corresponding to the error correction apparatus 1 based on the weighting, where the weighting is the first weighting. . Using the input speed of the current independent button triggered as one parameter of the error correction model, when the error correction model proceeds with error correction, the evaluation and calculation are carried out in terms of time. In connection with an already triggered recording button that is not input to at least one top screen belonging to the first full input, error correction is performed on the input code corresponding to the current independent button, and at least one error corrected input code is entered. Acquire. Here, the one full input is configured by adding an already triggered recording button not inputted to the upper screen to the current independent button, that is, all input codes not inputted to the upper screen, one independent at a time After further input of the button, changes also occur in all candidate words corresponding to already triggered recording buttons that are not entered on the top screen. For example, as shown in the input process shown in FIG. 2B, one full input points to “shoifashi” and the user does not enter the top screen, but instead enters “zainali”, one full input at this time. Input is shoifashizainali ".

For example, in step S302, if the weight when correcting the input code corresponding to the current independent button adjusted by the error correction apparatus 1 to the input code corresponding to the other button is relatively large, error correction The ratio of correcting the input code corresponding to the current independent button with the other input code by the device 1 is relatively large, and the probability of correcting the current independent button with the other button with the error correction model corresponding to the input method is relatively high. Is improved. Selecting a suitable existing error correction model according to the adjusted weighting, and using the adjusted weighting to the existing error correction model, performing error correction with other parameters in the existing error correction model, that is, time Proceed with error correction at the dimension. As an important consideration factor of the existing error correction model, the input speed of the current independent button is introduced into the existing error correction model, and the error correction is correctly performed on the input code. In step S303, the error correction device 1 Obtain at least one input code that has been error corrected with the error correction model. Here, the error corrected at least one input code may still be an input code corresponding to the current independent button, and the input code corresponding to the current independent button may be corrected to the input code corresponding to the other button. For example, if the input interface is a touch-type virtual QWERTY full keypad interface, the error correcting apparatus 1 may be configured to display all the keys on the keypad according to the starting point position L of the keypad triggered when the user proceeds with the input. Compute the distance (s1, s2 ... sn) from the button center point to the corresponding starting point position (L), and calculate the probability that each of the current independent buttons triggered by the user are each button according to the distance. Next, the input code corresponding to the current independent button triggered by the user according to the probability or by attaching other parameters is “s”, and the input speed of “s” is 9 characters / sec. ), If the input speed of the input code "s" is 9 characters / second with the error correction device 1, and the daily input speed threshold is set to 5 characters / second, the input speed of the input code is the corresponding input speed threshold. If greater than the value, When the input speed is at a high level, the weight of correcting the input code corresponding to the current independent button to the input code corresponding to the other button is increased, and the input code corresponding to the current independent button is increased. Reducing the weighting correction to itself "s", that is, if the user's destination button is still a weighting of "s" itself, in step S302, according to the input speed to the error correction device 1, the input It is determined that the speed is greater than the input speed threshold, in which case, the weight for correcting "s" to "w" is increased to w1, the weight for correcting "s" to "a" is increased to w2, and " the weight correcting to d "is increased to w3, the weight correcting to" x "is increased to w4, the weight of the user's destination button is still" s "itself to decrease to w5, and in step S303, Weighting (w1, w2, w3, w4, w5) to the correction device 1 and before the current independent button Based on the lock button, it determines the probability of correcting a current independent of the button to other button synthetically. For example, based on the calculation, the probability of correcting "s" to "w", "a", "d", and "x", respectively, is P1, P2, P3, and P4, respectively. Is a probability P5, and the user's purpose button is determined to be "a" in accordance with P1, P2, P3, P4 and P5. The probability of correcting the current independent button with the other button is comprehensively determined based on the recording button which is not input to the upper screen before the current independent button. In connection with the error correction method of the existing error correction model, the error correction is jointly performed on the input code corresponding to the current independent button. If " p " and " s " are both Virgins, the weighting to correct the letter " s " to mica " a " adjacent to " s "

The method of deciding how to adjust the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other button according to the input speed threshold is merely an example, and other existing or future. The method of determining how to adjust the weighting when correcting the input code corresponding to the current independent button to the input code corresponding to the other button to be shown also should be included within the protection scope of the present invention when applicable to the present invention. It should be understood by those skilled in the art that the present disclosure should be incorporated herein by reference. Here, in step S303, the error correcting apparatus 1 obtains at least one error corrected input code, registers it in the system background, does not display it on the input interface, that is, what is displayed on the input interface is still present. Input code corresponding to the button triggered by the user.

Here, the big data is analyzed and the relationship between the speed of user input and the error correction is mined in terms of time, and the error correction model is used to perform accurate error correction. The error correction model is obtained by continuously acquiring the input speed at the time of inputting the button and adjusting the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other buttons according to the input speed. As an element of correcting the input code, the input method is evaluated in terms of time using the input speed, and the input code determined through this spatial dimension such as the starting point of the keypad triggered by the user is evaluated. To improve the accuracy of error correction with the error correction model, It improves quality.

(두려워할 파),

(긁을 파),

(긁을 파),

(머리띠 파)......를 확정한다. 나아가, 단계(S304)에서, 오류 정정 장치(1)로 여전히 오류 정정되기 전의 적어도 하나의 입력 코드에 따라, 대응되는 적어도 하나의 후보 단어를 확정하며, 오류 정정되기 전의 적어도 하나의 입력 코드에 대응되는 적어도 하나의 후보 단어의 우선 순위는 오류 정정된 적어도 하나의 입력 코드에 대응되는 적어도 하나의 후보 단어보다 낮다. 예를 들어, 오류 정정되기 전의 입력 코드 “ps”에 대응되는 적어도 하나의 후보 단어 “

(평술)", "

(평시)", "

(촬영)" 등의 우선 순위는 “

(두려워할 파),

(긁을 파),

(긁을 파),

(머리띠 파)......”보다 낮다.In step S304, the error correcting apparatus 1 determines at least one candidate word corresponding to the entire input, according to the error-corrected at least one input code corresponding to each button of one full input. Specifically, in step S303, at least one input code error corrected by the error correction device 1 is sequentially obtained, and in step S303, each of the entire inputs to the error correction device 1 once. After acquiring an error corrected input code corresponding to the button, in step S304, the error correction apparatus 1 determines at least one candidate word corresponding to the entire input. For example, following the above example, after correcting the letter "s" to mica "a" adjacent to "s" on the keypad, the letter "a" and the previous letter "p" are combined into "pa" and , At step S304, the error correction apparatus 1 at least one candidate word corresponding to "pa":

(Waves to be afraid),

(Scraping),

(Scraping),

Confirm the (headband wave) ...... Further, in step S304, according to the at least one input code before the error correction device 1 is still error corrected, the corresponding at least one candidate word is determined and corresponds to the at least one input code before the error correction. The priority of the at least one candidate word is lower than at least one candidate word corresponding to the at least one input code that has been error corrected. For example, at least one candidate word corresponding to the input code “ps” before error correction.

(Flat) ","

(Normal) ","

(Shooting) ", etc.

(Waves to be afraid),

(Scraping),

(Scraping),

(Headband wave) ...... ”is lower than.

Here, the input process is continuous, and if the user constantly inputs and does not input the upper screen in the process of inputting the input code, the input code included in the current one full input is constantly increasing and newly After the added input code is error corrected, the write input code which is not input to the upper screen is continually attached so that a dynamic change occurs in at least one candidate word corresponding to the entire input.

Preferably, the error correction method further includes step S305 (not shown).

In step S305, the error correcting apparatus 1 obtains an average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code corresponding to the one entire input, Adjust the ordering arrangement of the at least one candidate word corresponding to the entire input. Specifically, when one user's one entire input is completed, in step S305, the input time of the entire input code corresponding to the one entire input acquired by the error correcting apparatus 1 and the one entire input process The average input speed of the button corresponding to each input code in the one entire input code is calculated according to the quantity of input codes in the corresponding one, and then the at least one candidate word corresponding to the entire input according to the average input speed is calculated. Adjust the ordering arrangement.

Preferably, the step S305 further includes a substep S3051 (not shown), a substep S3052 (not shown), and a substep S3053 (not shown).

Here, in the sub step S3051, the error correction apparatus 1 calculates the average input speed of the button corresponding to each input code of the entire input codes according to the input time of the entire input code. Specifically, in the substep S3051, the error correcting apparatus 1 corresponds to each input code of the entire input codes according to the input time of the entire input code and the quantity of input codes included in the entire input code. The average input speed of the button is calculated.

In a sub-step S3052, the error correction device 1 determines a corresponding error correction model based on the average input speed, wherein the error correction model is 1) independently based on the average input speed. Error correction level model; 2) an error correction rating model independently based on the average input speed; 3) an error correction level model based on the average input speed and the remaining environmental parameter fittings; And 4) an error correction rating model based on the average input speed and the remaining environmental parameter fittings.

(수발실),

(발표식),

(수법은)...... 등 "shoufashi"과 높은 관련도를 구비하는 일 유형의 후보 단어를 후보 단어 목록의 앞에 위치시키고, 오류 정정 처리를 진행하지 않은 초기 전체 입력 코드 "shoifashi"에 대응되는 후보 단어:

,

를 후보 단어 목록의 뒤에 위치시킨다. 마찬가지로, 상기 평균 입력 속도가 하이 레벨에 위치하면, 순서 배열을 교환한다. 상기 오류 정정 모델이 독립적으로 상기 평균 입력 속도를 기반으로 하는 오류 정정 평점 모델을 채용할 경우, 상기 오류 정정 평점의 점수는 상기 평균 입력 속도의 연속적인 선형 함수 또는 비 선형 함수에 관한 것으로, 상기 평균 입력 속도의 변화에 따라 상이하며, 서브 단계(S3053)에서, 오류 정정 장치(1)로 상기 입력 속도의 크기에 따라 대응되는 오류 정정 점수를 획득하고, 이어서 상기 점수에 따라 상기 전체 입력 중의 각 입력 코드가 오류 정정 장치(1)로 제일 처음에 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하며, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다. 상기 오류 정정 모델이 상기 평균 입력 속도 및 나머지 환경 파라미터 피팅을 기반으로 하는 오류 정정 레벨 모델을 채용할 경우, 상기 오류 정정 레벨은 평균 입력 속도 및 나머지 환경 파라미터 피팅에 따라 대응되는 오류 정정 레벨을 확정하기 위한 것으로, 상기 오류 정정 레벨에 따라 상기 전체 입력 중의 각 입력 코드가 오류 정정 장치(1)로 제일 처음에 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하며, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다. 상기 오류 정정 모델이 상기 평균 입력 속도 및 나머지 환경 파라미터 피팅을 기반으로 하는 오류 정정 평점 모델을 채용할 경우, 상기 오류 정정 평점의 점수는 상기 평균 입력 속도와 나머지 환경 파라미터 피팅의 연속적인 선형 함수 또는 비 선형 함수에 관한 것으로, 상기 평균 입력 속도의 변화 또는 나머지 환경 파라미터의 변화에 따라 상이하며, 이어서 상기 점수에 따라 상기 전체 입력 중의 각 입력 코드가 오류 정정 장치(1)로 제일 처음에 취득한 입력 코드인 가중 및 상기 전체 입력 중의 각 입력 코드가 각각 오류 정정된 이후의 입력 코드인 가중을 조정하며, 양자에 각각 대응되는 후보 단어가 후보 단어 목록에서의 순서 배열을 확정한다.In a substep S3053, the error correction apparatus 1 adjusts the order arrangement of the at least one candidate word corresponding to the entire input according to the error correction model. Specifically, when the error correction model independently adopts an error correction level model based on the average input speed, for example, a level may be independently high level, middle level, and low level based on the average input speed. When the average input speed is located at a low level, the candidate words corresponding to the input code, which is the first input code acquired by the error correction apparatus 1, are placed in front of each input code of the entire input, Each input code of the entire inputs increases the weighting which is the input code first obtained in step S301 by the error correcting apparatus 1, and the weighting which is an input code after each of the input codes of the entire inputs is respectively error corrected. And, likewise, when the average input speed is at a high level, each input code of the entire input fails. The weights that are input codes after each input code of the entire inputs are each error-corrected are increased so that candidate words corresponding to the input codes that are predetermined post-codes are located later, and each input code of the entire inputs is error corrected. The apparatus 1 reduces the weight which is the first input code acquired. For example, if the user enters “shoifashi” and the average input speed is at a low level, the probability that an error is present during this entire input is relatively low. In this case, the error corrected input code " candidate words for "shoufashi":

(Weather room),

(Presentation),

One method of having a high degree of relevance to "shoufashi", etc., is placed in front of the candidate word list, and the initial full input code "shoifashi" has not been processed for error correction. Matching candidate words:

,

Is placed after the candidate word list. Similarly, if the average input speed is at a high level, the ordering sequence is exchanged. When the error correction model independently adopts an error correction rating model based on the average input speed, the score of the error correction rating relates to a continuous linear function or a non-linear function of the average input speed, wherein the average It is different depending on the change of the input speed, and in the sub-step S3053, the error correction apparatus 1 obtains a corresponding error correction score according to the magnitude of the input speed, and then inputs each of the entire inputs according to the score. The code adjusts the weight which is the input code first obtained by the error correction apparatus 1 and the weight which is the input code after each input code of the entire inputs is error corrected, respectively, and the candidate words corresponding to both are candidate words. Confirm the ordering arrangement in the list. When the error correction model adopts an error correction level model based on the average input speed and the remaining environmental parameter fitting, the error correction level determines the corresponding error correction level according to the average input speed and the remaining environmental parameter fitting. Weighting which is an input code first obtained by the error correcting apparatus 1 according to the error correction level, and an input code after each input code of the all inputs is error corrected, respectively. Weighting is adjusted, and candidate words respectively corresponding to both determine an ordering arrangement in the candidate word list. If the error correction model employs an error correction rating model based on the average input speed and the remaining environmental parameter fitting, the score of the error correction rating is a continuous linear function or ratio of the average input speed and the remaining environmental parameter fitting. It relates to a linear function, which is different depending on the change of the average input speed or the change of the remaining environmental parameters, and then, according to the score, each input code of the entire input is the first input code acquired by the error correction device 1. Weighting and weighting, which are input codes after each input code of the entire inputs are each error-corrected, are adjusted, and candidate words corresponding to both determine the ordering arrangement in the candidate word list.

Preferably, the step S305 further includes a substep S3054 (not shown), a substep S3055 (not shown), and a substep S3056 (not shown).

In the substep S3054, the error correction apparatus 1 performs a noise removing process on a sequence consisting of a sequence in which each input code of the entire input codes is input, thereby obtaining a noise canceled sequence. Specifically, in one full input of a user, a noise canceling process is performed on one sequence consisting of a sequence in which each input code is input, for example, after the user inputs one input code, If the input process stops due to any change, the sequence corresponding to the input code far exceeds the predetermined upper limit threshold of the input when normally inputted, and the sequence corresponding to the input code is removed from the sequence. Preferably, the noise canceling process comprises: 1) removing the sequence if a sequence into which one input code is input is greater than a predetermined sequential upper limit threshold or less than a predetermined sequential lower limit threshold; 2) removing the sequence when the proportion of the sequence into which the one input code is input and the average sequence value input at this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And 3) removing the sequence if the mapping value of the fitting function with the inputted environment to which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold. It includes at least one of.

In a sub step S3055, the error correction apparatus 1 determines the input time of the entire input code according to the noise-free sequence. Specifically, the input time of the entire input code is equal to the sum of all sequences in the noise canceled sequence, and the quantity of the entire input code should be subtracted from the quantity of sequences removed through the noise removal process.

In a sub-step S3056, the error correcting apparatus 1 obtains the full input speed of the full input code based on the input time of the full input code, and determines the at least one candidate word corresponding to the full input. It is an ordered array. Specifically, in the sub-step S3056, based on the input time of the entire input code by the error correction apparatus 1 and the quantity of the input code after correspondingly subtracting the quantity of the noise canceling input code, Obtaining the average input speed adjusts an ordering arrangement of the at least one candidate word corresponding to the entire input.

Preferably, the method further comprises step S306 (not shown). In step S306, determining the user group in which the user is located according to the characteristics of the user triggering the button with the error correction apparatus 1; The error correction apparatus 1 is optimized according to the user group information. Specifically, according to the characteristics of the user through the learning of a certain time with the error correction device (1), to create a picture for it, a user group corresponding to the image through big data, for example, a young male enjoying fast input, Mining middle-aged women, etc. who enjoy serious input, but each feature of the user includes, but is not limited to, the overall input speed of the user, the input speed of each specific button of the user, the words he or she made, the user's frequent input, etc. Do not. In step S306, the error correction device 1 is optimized by performing threshold adjustment of each term with respect to the error correction device 1 through a feedback mechanism.

The method further includes step S307 (not shown). In step S307, the user input habit record information is acquired by the error correction apparatus 1; The error correction device 1 is optimized in accordance with the input habit record information. In detail, the error correcting apparatus 1 determines a user's input habits and methods, such as one-handed or two-handed input habits, a horizontal screen or a vertical screen input habit, according to the user's input starting point situation through a predetermined time learning. The user's input start point situation includes, but is not limited to, a degree of dispersion of the user's input start point, a speed of inputting a specific button, a touch error rate of inputting a specific button, and the like. In step S307, the error correction apparatus 1 acquires such input habit record information of the user, and adjusts the threshold value of each term in the error correction apparatus 1 according to the input habit and method thereof to correct the error. Optimize the device (1).

The present invention further provides a computer readable storage medium, wherein computer code is stored in the computer readable storage medium, and when the computer code is executed, any of the methods described above are executed.

The present invention further provides a computer program product, wherein any of the methods described above are executed when the computer program product is executed by a computer device.

The present invention further provides a computer device, wherein the computer device,

One or multiple processors; And

A storage device configured to store one or multiple computer programs; Including;

When the one or more computer programs are executed by the one or more processors, let the one or more processors implement any of the methods described above.

The invention may be practiced in software and / or a combination of software and hardware, for example, each device of the invention may be implemented by employing an application specific integrated circuit (ASIC) or any other similar hardware device. In one embodiment, the software program of the present invention may be executed by a processor to implement the above-described steps or functions. Similarly, the software program (including associated data structures) of the present invention may be stored in a computing device readable recording medium, such as a RAM storage device, a magnetic or optical drive or a floppy disk, and similar devices. In addition, some steps or functions of the present invention may be implemented by employing hardware, for example, may be implemented as a circuit for executing each step or function in cooperation with a processor.

It will be apparent to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms as long as it does not violate the spirit or basic features of the present invention. . Accordingly, in some respects, the embodiments are not to be considered limiting but illustrative, and the scope of the present invention is not limited to the above description, but is defined by the appended claims, and therefore, the scope of the claims. All changes within the meaning and range of equivalent requirements should be included in the present invention. Any reference signs in the claims should not be construed as limiting the related patent claims. It is also to be understood that the term "comprises" does not exclude other units or steps, and the singular does not exclude a plurality. Multiple units or devices mentioned in the system claims may be implemented through software or hardware in one unit or device. Terms such as "first" and "second" refer only to names, and do not indicate any specified order.

Claims (23)

In the input method error correction method based on the user input speed,
(a) acquiring an input speed of a triggered current independent button;
(b) adjusting a weighting value when correcting an input code corresponding to the current independent button with an input code corresponding to the other button according to the input speed;
(c) performing error correction on the input code corresponding to the current independent button based on the adjusted weighting to obtain at least one error corrected input code; And
(d) determining at least one candidate word corresponding to the entire input, according to the error corrected at least one input code corresponding to each button of one full input;
Including,
Step (c) is,
c1. Comprehensively determining a probability of correcting the current independent button with the other button based on the adjusted weighting and the recording button before the current independent button; And
c2. Obtaining at least one error corrected input code based on the probability,
The error corrected at least one input code may still be an input code corresponding to the current independent button, and the input code corresponding to the current independent button may be corrected to an input code corresponding to the other button. Error correction method. The method of claim 1,
(e) acquiring an average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code corresponding to the one entire input, and generating the at least one corresponding to the entire input; And adjusting the arrangement of the candidate words. The method of claim 2,
Step (e),
(e1) calculating the average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code;
(e2) determining a corresponding error correction model based on the average input speed; And
(e3) adjusting an ordering arrangement of the at least one candidate word corresponding to the entire input according to the error correction model. The method according to any one of claims 1 to 3,
According to the input speed, the step of adjusting the weight when correcting the input code corresponding to the current independent button to the input code corresponding to the other button,
Increasing weighting when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;
Reducing a weight of itself when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;
Increasing a weight of itself when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; And
Reducing a weight when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; Error correction method comprising at least one of. The method of claim 4,
Determination of the difficulty of a button touch error,
Level class determination according to the input speed of the independent button;
Rating weighting according to the input speed of the independent button;
Determining a rating mapping to an input speed of the independent button and a fitting function according to the current input environment; And
A rating weighting decision mapping to an input speed of the independent button and a fitting function according to the current input environment; Error correction method comprising at least one of. The method of claim 3,
The error correction model,
An error correction level model independently based on the average input speed;
An error correction rating model independently based on the average input speed;
An error correction level model based on the average input speed and the remaining environmental parameter fittings; And
An error correction rating model based on the average input speed and the remaining environmental parameter fittings; Error correction method comprising at least one of. The method according to any one of claims 2, 3, or 6,
Step (e),
Performing a noise removing process on a sequence consisting of a sequence into which each input code of the entire input codes is input, to obtain a noise canceled sequence;
Determining an input time of the entire input code according to the noise canceled sequential sequence; And
Acquiring the average input speed of the entire input code based on an input time of the entire input code, and adjusting an order arrangement of the at least one candidate words corresponding to the entire input; Error correction method. The method of claim 7, wherein
The noise reduction process is,
If the sequence into which one input code is input is greater than a predetermined upper limit threshold or less than a predetermined lower limit threshold, removing the sequence;
Removing the sequence when the proportion of the sequence into which the one input code is input and the average sequence value input at this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And
Removing the sequence if the mapping value of the fitting function to the input environment at which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold; Error correction method comprising at least one of. The method according to any one of claims 1, 2, 3 or 6,
Determining a user group in which the user is located according to a feature of the user triggering the button; And
Optimizing the error correction method according to the user group information. The method according to any one of claims 1, 2, 3 or 6,
Acquiring user's input habit record information; And
And optimizing the error correction method according to the input habit record information. In the input method error correction device based on the user input speed,
An acquisition device configured to acquire an input speed of a triggered current independent button;
An adjusting device configured to adjust weighting when correcting an input code corresponding to the current independent button with an input code corresponding to the other button according to the input speed;
An acquisition device configured to perform error correction on an input code corresponding to the current independent button based on the adjusted weighting to obtain at least one error corrected input code; And
A confirmation device, configured to determine at least one candidate word corresponding to the entire input, according to the error corrected at least one input code corresponding to each button of one full input;
Including,
The obtaining device,
A device for comprehensively determining a probability of correcting the current independent button with the other button based on the adjusted weighting and the recording button before the current independent button; And
An apparatus for obtaining at least one input code that is error corrected based on the probability,
The error corrected at least one input code may still be an input code corresponding to the current independent button, and the input code corresponding to the current independent button may be corrected to an input code corresponding to the other button. Error correction device. The method of claim 11,
The error correction device,
Acquiring an average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code corresponding to the one entire input, and then determining the at least one candidate word corresponding to the entire input. And an order arranging device configured to adjust the order arranging. The method of claim 12,
The order arrangement device,
A calculating unit, configured to calculate the average input speed of a button corresponding to each input code of the entire input codes according to the input time of the entire input code;
A first determination unit configured to determine a corresponding error correction model based on the average input speed; And
And a first adjustment unit configured to adjust an ordering arrangement of the at least one candidate word corresponding to the entire input according to the error correction model. The method according to any one of claims 11 to 13,
Adjusting the weighting when correcting the input code corresponding to the current independent button with the input code corresponding to the other button according to the input speed,
Increasing weighting when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;
Reducing the weight of itself when correcting an input code corresponding to a button which is prone to a touch error with an input code corresponding to the other button;
Increasing the weight of itself when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; And
Reducing weighting when correcting an input code corresponding to a button which is hard to generate a touch error with an input code corresponding to the other button; Error correction device comprising at least one of. The method of claim 14,
Determination of the difficulty of a button touch error,
Level class determination according to the input speed of the independent button;
Rating weighting according to the input speed of the independent button;
Determining a rating mapping to an input speed of the independent button and a fitting function according to the current input environment; And
A rating weighting decision mapping to an input speed of the independent button and a fitting function according to the current input environment; Error correction device comprising at least one of. The method of claim 13,
The error correction model,
An error correction level model independently based on the average input speed;
An error correction rating model independently based on the average input speed;
An error correction level model based on the average input speed and the remaining environmental parameter fittings; And
An error correction rating model based on the average input speed and the remaining environmental parameter fittings; Error correction device comprising at least one of. The method according to any one of claims 12, 13, or 16,
The order arrangement device,
A noise canceling unit configured to perform a noise canceling process on a sequence consisting of a sequence into which each input code of the entire input codes is input, to obtain a noise canceled sequence;
A second confirmation unit, configured to confirm an input time of the entire input code according to the noise canceled sequential sequence; And
A second adjustment unit, configured to obtain an overall input speed of the entire input code based on an input time of the entire input code, and adjust an ordering arrangement of the at least one candidate word corresponding to the entire input; Error correction apparatus, characterized in that. The method of claim 17,
The noise reduction process is,
If the sequence into which the one input code is input is greater than the predetermined sequential upper limit threshold or less than the predetermined sequential lower limit threshold, removing the sequence;
Removing the sequence when the proportion of the sequence into which the one input code is input and the average sequence value input at this time is greater than the predetermined proportional upper limit threshold or less than the predetermined proportional lower limit threshold; And
Removing the sequence if the mapping value of the fitting function with the input environment at which the one input code is input is greater than the predetermined mapping upper limit threshold or less than the predetermined mapping lower limit threshold; Error correction device comprising at least one of. The method according to any one of claims 11, 12, 13, or 16,
And a first optimization device configured to determine a user group in which the user is located according to a feature of a user who triggers a button, and to optimize the error correction device according to the user group information. The method according to any one of claims 11, 12, 13, or 16,
And a second optimization device configured to acquire input habit record information of a user and optimize the error correction device according to the input habit record information. A computer readable storage medium having computer code stored thereon,
4. The computer readable storage medium of claim 1, wherein the method of any one of claims 1 to 3 is executed when the computer code is executed. In a computer device,
One or multiple processors; And
A storage device configured to store one or multiple computer programs;
Including but not limited to:
4. When the one or more computer programs are executed by the one or more processors, causing the one or more processors to implement the method of any one of claims 1 to 3. delete

Images