KR102417046B1 - Device and method for providing recommended words for character input from user - Google Patents

Abstract

Provided are a device and a method for providing a recommended word for a character input by a user. A method for a device to provide a recommended word includes an operation of receiving a text input from a user, an operation of recommending at least one pseudo-morpheme including the received text input by analyzing the text input, and at least one an operation of recommending at least one word including the selected pseudo morpheme in response to an input of selecting one of the pseudo morphemes, and an operation of displaying the selected word in response to an input of selecting one of the at least one word do.

Description

DEVICE AND METHOD FOR PROVIDING RECOMMENDED WORDS FOR CHARACTER INPUT FROM USER}

The present disclosure relates to a device and method for providing a recommended word, and more particularly, to a device and method for providing a recommended word for a character input from a user.

In addition, the present disclosure relates to an artificial intelligence (AI) system for simulating functions such as cognition and judgment of a human brain by using a machine learning algorithm and an application thereof.

As multimedia technology and network technology develop, a user can receive various services using a device. In particular, various methods for providing a recommended word based on a word being input by a user have been provided. However, in order to provide a recommended word, there is a problem in that the amount of data that needs to be stored in the database is large, and it is difficult to accurately recommend a word intended by the user. Accordingly, there is a need for a method capable of more accurately predicting and recommending a word intended by a user while reducing the amount of data to be stored in the database.

Also, in recent years, artificial intelligence systems that implement human-level intelligence have been used in various fields. Unlike the existing rule-based smart system, an artificial intelligence system is a system in which a machine learns, judges, and becomes smarter by itself. The more the artificial intelligence system is used, the better the recognition rate and the more accurate understanding of user preferences, and the existing rule-based smart systems are gradually being replaced by deep learning-based artificial intelligence systems.

Artificial intelligence technology consists of machine learning (eg, deep learning) and element technologies using machine learning.

Machine learning is an algorithm technology that categorizes/learns the characteristics of input data by itself, and element technology uses machine learning algorithms such as deep learning to simulate functions such as cognition and judgment of the human brain. It consists of technical fields such as understanding, reasoning/prediction, knowledge expression, and motion control.

The various fields where artificial intelligence technology is applied are as follows. Linguistic understanding is a technology for recognizing and applying/processing human language/text, and includes natural language processing, machine translation, dialogue system, question and answer, and speech recognition/synthesis. Visual understanding is a technology for recognizing and processing objects like human vision, and includes object recognition, object tracking, image search, human recognition, scene understanding, spatial understanding, image improvement, and the like. Inferential prediction is a technology for logically reasoning and predicting by judging information, and includes knowledge/probability-based reasoning, optimization prediction, preference-based planning, and recommendation. Knowledge expression is a technology that automatically processes human experience information into knowledge data, and includes knowledge construction (data generation/classification) and knowledge management (data utilization). Motion control is a technology for controlling autonomous driving of a vehicle and movement of a robot, and includes motion control (navigation, collision, driving), manipulation control (action control), and the like.

Some embodiments may provide a device and method for providing a recommended word, which can reduce the amount of data to be stored in a database by providing the recommended word in units of pseudo morphemes including characters input by the user.

In addition, some embodiments provide a recommended word in units of a pseudo morpheme smaller than a unit of a word, and provide a recommended word in units of words based on a pseudo morpheme selected from among the recommended pseudo morphemes, so that the user can select a recommended word that matches the intention. It is possible to provide a device and method for providing more effectively.

As a technical means for achieving the above-described technical problem, a first aspect of the present disclosure includes a memory in which at least one program is stored, a user input unit for receiving a user input for inputting characters in a character input window, and at least one program. and at least one processor configured to provide a recommended word for the inputted character by executing, wherein the at least one program analyzes the inputted character to obtain at least one pseudo-morpheme including the inputted character. , recommending at least one word including a pseudo morpheme selected in response to an input of selecting one from among at least one pseudo morpheme, and in response to an input of selecting one from among at least one pseudo morpheme, A device may be provided, including an operation of displaying a word in a text input window.

In addition, a second aspect of the present disclosure provides an operation of receiving a text input from a user, an operation of recommending at least one pseudo-morpheme including the received text input by analyzing the text input, and at least one an operation of recommending at least one word including the selected pseudo morpheme in response to an input of selecting one of the pseudo morphemes, and an operation of displaying the selected word in response to an input of selecting one of the at least one word , it is possible to provide a method for the device to provide a recommended word for a character input by the user.

In addition, a third aspect of the present disclosure may provide a computer-readable recording medium in which a program for executing the method of the second aspect is recorded on a computer.

1 is a diagram illustrating an example in which a device provides a recommended word for a character input by a user, according to some embodiments.
2 is a flowchart illustrating a method for a device to provide a recommended word, according to some embodiments.
3 is a flowchart illustrating a method for a device to determine at least one pseudo morpheme to be recommended, according to some embodiments.
4 is a diagram for describing a method of determining, by a device, a recommended word, according to some embodiments.
5A and 5B are diagrams illustrating an example in which a device recommends at least one pseudo morpheme, according to some embodiments.
6A and 6B are diagrams illustrating an example in which a device recommends at least one pseudo morpheme, according to some embodiments.
7 is a flowchart illustrating a method for a device to display a space character, according to some embodiments.
8 is a diagram illustrating an example in which a device displays a blank character and recommends at least one pseudo morpheme to be positioned next to the displayed blank character, according to some embodiments.
9 is a diagram illustrating an example of a neural network according to some embodiments.
10 is a diagram illustrating an example in which displayed pseudo morphemes are distinguished to indicate whether there is a word to be recommended by the device according to some embodiments.
11 is a diagram illustrating an example in which a device recommends a pseudo morpheme in consideration of a user's frequency of use for each of the pseudo morphemes, according to some embodiments.
12 is a diagram illustrating an example in which a device recommends a pseudo morpheme in consideration of the type and order of at least one previously displayed word, according to some embodiments.
13 is a diagram illustrating an example in which a device recommends at least one word in consideration of a character input characteristic of a user, according to some embodiments.
14 is a block diagram of a device in accordance with some embodiments.
15 is a block diagram of a device according to another embodiment.
16 is a block diagram of a processor in accordance with some embodiments.
17 is a block diagram of a data learning unit according to some exemplary embodiments.
18 is a block diagram of a data recognition unit according to some exemplary embodiments.
19 is a diagram illustrating an example of learning and recognizing data by interworking with a device and a server according to some embodiments.
20 is a flow diagram of a device using a data recognition model in accordance with some embodiments.
21 to 23 are flowcharts of a network system using a data analysis model in accordance with some embodiments.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, this means that other components may be further included rather than excluding other components unless otherwise stated.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating an example in which a device provides a recommended word based on a character input by a user, according to an exemplary embodiment.

Referring to FIG. 1 , the device 1000 may receive a user input for inputting text in a text input window, and may provide a recommended word based on the input text.

When the device 1000 provides the recommended word in word unit based on the input text, the device 1000 recommends the word including the text input by the user based on the frequency of use of each word stored in the database. can do. In this case, since the frequency of use is calculated for each word unit, for example, "love", "with love", and "only love" are different words, so even if all three words contain the noun "love" , "with love", "with love", and "only with love" can be calculated to have a low frequency of use. Accordingly, even when the user frequently uses an expression including “love”, a word including “love” may not be preferentially recommended. For example, "apple" is 1000 times, "love" is 800 times, "apple" is 30 times, "love" is 700 times, "apple" is 20 times, "love only" is 500 times. When used once, users use "love" more frequently than "apple", but depending on the frequency of use calculated in word units, "apple", "love", and "only love" may be recommended in that order. have. Accordingly, when the device 1000 provides a recommended word for a character input from a user, a method for providing the recommended word in units smaller than a word is needed.

The device 1000 according to some exemplary embodiments may provide various recommended words having different meanings to the user by first recommending the characters input by the user in units of pseudo morphemes. A pseudo morpheme may mean the smallest unit of speech having a meaning. For example, the pseudo morpheme may include a noun of a minimum unit having a meaning, and according to an embodiment, a compound noun having a meaning may also be classified as one pseudo morpheme. For example, "love", "apple", and "person" are nouns with different meanings, and each may be a pseudo-morpheme. Also, a word may be used in a meaning including a pseudo morpheme and a word word.

When an input for selecting one from among at least one recommended pseudo morpheme is received, the device 1000 according to some embodiments recommends at least one word including the selected pseudo morpheme, and thus a recommended word matching the user's intention can be provided more easily.

In addition, the device 1000 according to some embodiments may determine at least one pseudo morpheme to be recommended and at least one word to be recommended by using various context information regarding character input, and may determine a position where a blank character is displayed. can judge

The context information may include information on text input by the user in the past, information on at least one word previously displayed in the text input window, information on voice data input by the user in the past, and the like. Information on characters input by the user in the past includes information on pseudo morphemes and word words frequently used by the user in the past, and information on pseudo morphemes selected by the user from among pseudo morphemes recommended by the device 1000 . can do. The previously displayed information on the at least one word may include information related to the part-of-speech and sentence component of the at least one pre-displayed word in the text input window. For example, the information on at least one previously displayed word may include information indicating whether the previously displayed word in the text input window is a word corresponding to a subject or a word corresponding to a subject and a verb, but is limited thereto. doesn't happen

The device 1000 is a smart phone, a tablet PC, a PC, a smart TV, a mobile phone, a personal digital assistant (PDA), a laptop, a media player, a micro server, a global positioning system (GPS) device, an e-book terminal, a digital broadcasting terminal, It can be, but is not limited to, navigation, kiosks, MP3 players, digital cameras, consumer electronics and other mobile or non-mobile computing devices. Also, the device 1000 may be a wearable device such as a watch, glasses, a hair band, and a ring having a communication function and a data processing function. However, the present invention is not limited thereto, and the device 1000 may include any type of device capable of receiving a user's voice input and providing an answer message to the user.

In addition, the device 1000 includes information on characters input by the user in the past, information on at least one word previously displayed in the text input window, and information on voice data input by the user in the past. In order to use the context information, it may communicate with a server and other devices through a predetermined network. In this case, the network includes a local area network (LAN), a wide area network (WAN), a value added network (VAN), a mobile radio communication network, a satellite communication network, and their mutual It is a data communication network in a comprehensive sense that includes a combination and enables each network constituent entity to communicate smoothly with each other, and may include a wired Internet, a wireless Internet, and a mobile wireless communication network. Wireless communication is, for example, wireless LAN (Wi-Fi), Bluetooth, Bluetooth low energy, Zigbee, WFD (Wi-Fi Direct), UWB (ultra wideband), infrared communication (IrDA, infrared Data Association) ), NFC (Near Field Communication), etc. may be there, but is not limited thereto.

2 is a flowchart illustrating a method for the device 1000 to provide a recommended word, according to some embodiments.

In operation S200 , the device 1000 receives a user input for inputting text in the text input window. In this case, the text input window of the device 1000 may include a window used to input a text message or SNS message, and the input text may include at least one alphabet or at least one syllable. .

In operation S210 , the device 1000 recommends at least one pseudo-morpheme including the input character by analyzing the input character. The device 1000 according to some embodiments may recommend at least one pseudo morpheme including the input text by using context information regarding text input. For example, the device 1000 may include statistical data on a typical frequency of use for each pseudo morpheme, information about characters input by the user in the past, and a user from among the pseudo morphemes recommended by the device 1000 . At least one pseudo morpheme may be recommended in consideration of information on the selected pseudo morpheme, information on voice data input by the user in the past, and the like.

In operation S220 , the device 1000 recommends at least one word including the selected pseudo morpheme in response to an input for selecting one from among the recommended at least one pseudo morpheme.

The user may select a pseudo morpheme that the user intends to input in the text input window from among the at least one recommended pseudo morpheme. As an input for selecting a pseudo morpheme is received, the device 1000 may recommend at least one word including the selected pseudo morpheme. In this case, the device 1000 may recommend at least one word by using context information regarding text input. For example, the device 1000 may store information on text input by the user in the past, information on at least one word previously displayed in the text input window, and information on voice data input by the user in the past. Taking this into account, at least one word word may be recommended. When the selected pseudo morpheme is a noun, at least one word including the pseudo morpheme may be a form in which a noun is combined with a proposition, but is not limited thereto.

Also, the device 1000 may recommend at least one word including the selected pseudo morpheme based on the character input characteristic of the user. For example, the device 1000 may recommend at least one word including the selected pseudo morpheme in consideration of an expression frequently used by the user.

Also, the device 1000 may recommend at least one other word including the selected word in response to an input for selecting one from among the at least one recommended word. For example, the device 1000 may store information on text input by the user in the past, information on at least one word previously displayed in the text input window, and information on voice data input by the user in the past. Taking this into account, at least one other word may be recommended. In this case, the at least one other word including the selected word may include a word in which another proposition is added to the selected word, but is not limited thereto.

In operation S230 , the device 1000 displays the selected word in the text input window in response to an input for selecting one from among at least one word.

Also, the device 1000 according to some embodiments may predict whether a blank character will be positioned after the selected word, and display the selected word and blank character together in the text input window based on the predicted result.

In the above, it has been described that the determination of the pseudo morpheme to be recommended, the determination of the word to be recommended, and the determination of the position where the blank character is displayed are performed based on separate learning, but is not limited thereto. At least two or more of the determination of the pseudo morpheme to be recommended, the determination of the word to be recommended, and the determination of the position where the blank character is displayed may be performed by one learning model.

In addition, the device 1000 according to some embodiments may use a deep neural network to determine at least one pseudo morpheme to be recommended, at least one word to be recommended, and a reference for a position where a blank character is displayed. can learn

3 is a flowchart illustrating a method for a device to determine at least one pseudo morpheme to be recommended, according to some embodiments.

In operation S300 , the device 1000 may receive a user input for inputting text in the text input window. For example, the device 1000 may receive a user input for inputting a text in a window for inputting a text message or an SNS message. In this case, the input character may include at least one alphabet or at least one syllable.

In operation S310 , the device 1000 may collect information about characters input by the user in the past. For example, the device 1000 may determine a pseudo morpheme to be recommended in consideration of information on pseudo morphemes and words frequently used by the user in the past.

In operation S320 , the device 1000 may collect general statistical data regarding the frequency of use of each pseudo morpheme stored in the database. The device 1000 may determine a pseudo morpheme with a high frequency of use as a pseudo morpheme to be recommended in consideration of general statistical data regarding the frequency of use of each pseudo morpheme.

In operation S330 , the device 1000 may collect information about voice data input by the user in the past. The device 1000 may interpret the user's voice data input through the microphone using various natural language interpretation techniques, and may determine a recommended word in consideration of information on the analyzed voice data.

In FIG. 3 , for convenience, the device 1000 sequentially collects information on characters input by the user, general statistical data on the frequency of use of each pseudo morpheme, and information on voice data input by the user in the past. However, the device 1000 may learn a method of collecting the above-described information in parallel and determining a pseudo morpheme to be recommended by using the collected information.

In operation S340 , the device 1000 may recommend at least one pseudo morpheme including the input character.

4 is a diagram for describing a method of determining, by a device, a recommended word, according to some embodiments.

The device 1000 according to some embodiments may recommend at least one pseudo morpheme including the input character by analyzing the character input by the user. The device 1000 may extract at least one pseudo morpheme including an input character from a database including the pseudo morphemes. The device 1000 may determine at least one pseudo morpheme to be recommended from among at least one pseudo morpheme extracted from the database in consideration of various context information regarding text input. For example, referring to FIG. 4 , the device 1000 includes information 410 on characters input by the user in the past, general statistical data 411 on the frequency of use of each pseudo morpheme, and information on characters input by the user in the past. At least one pseudo morpheme to be recommended may be determined in consideration of the information 412 on one piece of voice data.

The device 1000 includes information 410 on characters input by the user in the past, general statistical data 411 on the frequency of use of each pseudo morpheme, information 412 on voice data input by the user in the past, and the like. It is possible to learn a criterion for determining a pseudo morpheme to be recommended, a criterion for determining at least one word to be recommended, and a criterion for determining a position at which a blank character is displayed. For example, the determination of which pseudo morpheme to recommend, which word to recommend, and at which position the blank character should be displayed, is performed by a deep neural network including a Recurrent Neural Network (RNN) ( Deep Neural Network) may be performed based on learning according to the technology.

5A and 5B are diagrams illustrating examples in which a device recommends a pseudo morpheme according to some embodiments.

Referring to FIG. 5A , for example, when the user inputs the character “four” into the character input window 500 of the device 1000, the device 1000 displays “four” as a pseudo morpheme including “four”. Love", "apology", and "the boss" may be recommended 501 . In this case, the recommended at least one pseudo morpheme may be a result output when a character input from the user is input to a recurrent neural network (RNN). For example, the device 1000 considers information on characters input by the user in the past, general statistical data on the frequency of use of each pseudo morpheme, information on voice data input by the user in the past, and the like. , at least one pseudo morpheme including "four" may be recommended. Which pseudo morpheme is to be recommended among various pseudo morphemes including "four" may be determined based on learning according to a preset criterion.

The device 1000 according to some embodiments may display at least one recommended pseudo morpheme outside the text input window 500 . The user may select one pseudo morpheme intended by the user from among the at least one recommended pseudo morpheme. For example, as shown in FIG. 5A , the user may select 502 “person” from among the three recommended pseudo morphemes.

The device 1000 according to some embodiments may recommend at least one word including the selected pseudo morpheme in response to a user input for selecting one from among at least one pseudo morpheme. In this case, the recommended at least one word may be a result output when the selected pseudo morpheme is input to a recurrent neural network (RNN). For example, the device 1000 may store information on characters input by the user in the past, general statistical data for each word including the selected pseudo morpheme, and information on voice data input by the user in the past. In consideration of this, at least one word including the selected pseudo morpheme may be recommended. When the pseudo morpheme is a noun having a predetermined meaning, a word including the pseudo morpheme may have a form in which a proposition is added to the noun.

For example, referring to FIG. 5A , in response to an input of selecting “person”, the device 1000 may recommend at least one word including “person”. In this case, the recommended word may be a form in which a proposition is added to “person” such as “a person”, “a person”, and “with a person”, but is not limited thereto.

Also, the device 1000 according to some embodiments may recommend at least one other word including the selected word in response to an input for selecting one from among the at least one recommended word. For example, two or more prepositions may be connected to a noun to form one word. In the case of Hangul, one word formed by connecting a noun (eg, "love") and a proposition (eg, "only") is added to another connecting ending (eg, "only love"). , "great") may be additionally connected to form another word (eg, "as much as love"). Accordingly, when an input for selecting one word from among at least one word recommended by the device 1000 is received, the device 1000 may recommend at least one other word including the selected word. For example, referring to FIG. 5B , the user may select ( 512 ) “with a person” from among the words “a person”, “a person”, and “with a person” recommended by the device 1000 . can In this case, the device 1000 may recommend ( 521 ) “with a person” and “with a person” as at least one other word including “with a person”. The device 1000 may display the recommended “with a person” and “with a person” outside the text input window 500 .

The device 1000 according to an embodiment first recommends at least one pseudo morpheme including a character input into the text input window 500 , and at least one pseudo morpheme including a selected pseudo morpheme from among the recommended at least one pseudo morpheme. I can recommend the word of . Accordingly, the device 1000 may reduce the amount of data to be stored in the database compared to the case where the recommended word is provided in units of words. In addition, the device 1000 may more effectively provide the recommended word meeting the user's intention by providing the recommended word in units of pseudo morphemes and then in units of words.

6A and 6B are diagrams illustrating an example in which a device recommends at least one pseudo morpheme, according to some embodiments.

Unlike Hangul, English is usually written with a space between words. Accordingly, in the case of English, a noun, an adjective, an adverb, and a verb prototype of a unit that cannot be further divided may be one pseudo morpheme. For example, if the pseudo-morpheme is a verb prototype (eg, "say"), one word containing the pseudo-morpheme is a modified form of the verb (eg, "said", "says", " saying", etc.), but is not limited thereto.

Referring to FIG. 6A , for example, when the user inputs “wa” into the character input window 600 of the device 1000 , the device 1000 is configured as at least one pseudo morpheme including “wa”. , “wait”, “want”, and “water” may be recommended 601 .

The device 1000 may display at least one recommended pseudo morpheme outside the text input window. The user may select one pseudo morpheme intended by the user from among the at least one recommended pseudo morpheme. In response to an input of selecting one pseudo morpheme, the device 1000 may recommend at least one word including the selected pseudo morpheme. For example, as shown in FIG. 6A , the user may select “wait” from among the three recommended pseudo morphemes “wait”, “want”, and “water”. The device 1000 may recommend ( 611 ) at least one word including “wait” in response to an input of selecting “wait”. In this case, as shown in FIG. 6A , the recommended word may be a modified form of “wait,” which is the original verb, such as “wait,” “waiting,” and “waited,” but is not limited thereto. For example, when the pseudo morpheme is a noun (eg "child"), the recommended word includes nouns derived from the pseudo morpheme (eg, "child", "children", "childhood") can do.

Also, in response to an input of selecting one of the at least one recommended word, the device 1000 may recommend at least one other word including the selected word. For example, referring to FIG. 6B , when the user inputs “mea” into the text input window, the device 1000 selects “meaning”, “meager”, and “meadow” as pseudo morphemes including “mea”. can recommend

Also, in response to an input of selecting “meaning” from among the recommended pseudo morphemes, the device 1000 may recommend 631 “meaningless” and “meaningful” as words including “meaning”. And, as the user selects “meaningless” from among the recommended words, the device 1000 may further recommend 641 “meaninglessness” and “meaninglessly” including “meaningless”. Accordingly, the device 1000 may recommend at least one pseudo morpheme based on a character input by the user, thereby increasing the possibility of providing a pseudo morpheme that meets the user's intention.

7 is a flowchart illustrating a method of displaying a space character by the device 1000 according to some exemplary embodiments.

The device 1000 according to some embodiments may predict whether a blank character will be positioned next to a word selected from among at least one recommended word based on context information regarding text input. The device 1000 may, for example, store information on characters input by the user in the past, statistical data indicating the general frequency of use for each pseudo morpheme, and voice data input by the user in the past. It is possible to predict whether a blank character will be positioned next to a word selected from among at least one recommended word by using the information about the word.

In operation S700 , the device 1000 may receive a user input for inputting text. In this case, the input character may include at least one alphabet or at least one syllable.

In operation S710 , the device 1000 may recommend at least one pseudo morpheme including the input character. Operations S700 and S710 correspond to operations S200 and S210 of FIG. 2 , and thus detailed descriptions thereof will be omitted.

In operation S720 , the device 1000 may recommend at least one word including the selected pseudo morpheme in response to an input for selecting one from among the at least one recommended pseudo morpheme. The user may select a pseudo morpheme intended by the user from among at least one pseudo morpheme recommended by the device 1000 . Accordingly, the device 1000 may recommend at least one word including the pseudo morpheme selected by the user. In this case, when the pseudo morpheme is a noun, at least one word including the pseudo morpheme may have a form in which a proposition is connected to the noun.

In operation S730, the device 1000 may receive a user input for selecting one from among at least one recommended word. The user may select a word that the user intends to input into the text input window of the device 1000 from among at least one recommended word.

In operation S740 , the device 1000 may predict whether a blank character will be positioned after the selected word in response to an input for selecting one from among at least one recommended word. For example, when an input for selecting one from among at least one recommended word is received and a space character is highly likely to appear after the selected word, the device 1000 may predict that a space character will be positioned after the selected word. have. For example, when the word "to the Olympics" is selected by the user from among the at least one recommended word, the user's intended expression is another word including "at the Olympics", such as "at the Olympics" or "at the Olympics". it can be idiotic However, if the word "apology" is selected by the user, there may be a higher probability that a space character is positioned than other words including "apology". Accordingly, the device 1000 may predict whether or not a blank character will be positioned after the selected word based on information on text data input by the user in the past, statistical data on words with high frequency of use in general, and the like. can In this case, the prediction regarding whether a blank character will be positioned after the selected word may be performed based on learning according to a deep neural network technique.

In operation S750 , the device 1000 may display the selected word and blank character in the text input window based on the prediction result. For example, when it is determined that a blank character is highly likely to be positioned after the selected word, the device 1000 may sequentially display the selected word and the blank character in the text input window.

Also, as the blank character is displayed in the character input window, the device 1000 according to some embodiments may select at least one pseudo morpheme to be positioned after the blank character based on at least one word displayed in the character input window. can recommend

8 is a diagram illustrating an example of displaying a blank character and recommending at least one pseudo morpheme to be positioned after the displayed blank character, according to some embodiments.

The device 1000 according to some embodiments recommends at least one pseudo morpheme including a character input by the user, and includes the selected pseudo morpheme in response to a user input for selecting one from among the recommended at least one pseudo morpheme At least one word can be recommended. In this case, when the user selects one of at least one recommended word, the device 1000 may predict whether a blank character is positioned after the selected word. When the device 1000 predicts that a blank character will be located after the selected word, the device 1000 may sequentially display the selected word and the blank character on the character input window. Also, as the blank character is displayed on the character input window, the device 1000 may recommend at least one pseudo morpheme to be positioned after the blank character.

For example, referring to FIG. 8 , upon receiving a user input for inputting “saree” in the text input window, the device 1000 recommends “love”, “person”, and “consideration 801” (801 ). )can do. And, as a user input for selecting “love” is received, the device 1000 may recommend “love”, “love”, and “with love” as at least one word including “love”. have. In this case, when a user input for selecting “I love you” is received, the device 1000 may predict whether a blank character will be positioned after “I love you”. The device 1000 determines whether there is another word including the word "love", the frequency of use of another word including "love", the user's character input characteristics, and a space character after "love" It is possible to collect information about statistical data about the case. The device 1000 may learn a position to display a blank character based on the collected information. For example, referring to FIG. 8 , when the device 1000 predicts that a blank character will be located after “I love you”, the device 1000 sequentially displays “I love you” and a blank character in the character input window. Display 811 may be performed.

The device 1000 according to some embodiments may recommend at least one pseudo morpheme to be positioned after the blank character as the blank character is displayed on the character input window. Referring to FIG. 8 , when a blank character is displayed on the character input window, the device 1000 indicates that at least one pseudo morpheme to be positioned next to “love”, such as “sick”, “happy”, “excited”, etc. can be recommended (821).

9 is a diagram illustrating an example of a neural network according to some embodiments.

The device 1000 according to some embodiments determines, using a neural network, a word (W t ) to be recommended next to a previously recommended word (W _{t -1} ), and determines whether a blank character is to be displayed can decide whether

Referring to FIG. 9 , when a first word (W _t-1 ) 910 previously recommended by the device 1000 is input to a neural network 900 , learning according to a neural network technology Based on , the probability distribution 920 for the second word W _t to be positioned next to the first word W _t-1 and the probability to the blank character to be positioned next to the second word W _t . A distribution 921 may be output. For example, the probability distribution 920 for the second word W _t is located next to the first word W t- ₁ for each word including the first word W _t-1 . It may mean a probability distribution indicating the probability of doing so. Also, the probability distribution 921 regarding the blank character to be positioned next to the second word W _t may mean a distribution indicating the probability that the blank character will be positioned after the second word W _t . Based on the output result of the neural network 900 , the device 1000 places a word having a high probability distribution among words including the first word W t _{- 1 after the first word W t-1} It can be determined as the second word (W _t ) to be performed.

The neural network 900 may update a learning result whenever one word is input. For example, the probability distribution 920 for the second word W _t of the neural network 900 and the probability distribution 921 for the blank character may be used again to update the learning result of the neural network 900 . .

10 is a diagram illustrating an example in which displayed pseudo morphemes are distinguished to indicate whether there is a word to be recommended by the device according to some embodiments.

When recommending at least one pseudo morpheme including a character input by the user, the device 1000 according to some embodiments may distinguish the recommended pseudo morpheme to indicate whether there is a word to recommend including the pseudo morpheme. can be displayed as For example, referring to FIG. 10 , the device 1000 responds to a user input of inputting the character “four”, as pseudo morphemes including “sa”, “love”, “apple”, and “ You can recommend "the boss." In this case, the device 1000 may allow the user to easily identify whether there is at least one recommended word including each pseudo morpheme for "love", "apple", and "boss". For example, as shown in FIG. 10 , the device 1000 may distinguishly display a predetermined figure (eg, a triangle) 1002 next to a pseudo morpheme. Alternatively, according to an embodiment, the device 1000 may distinguishly display the pseudo morpheme by highlighting the pseudo morpheme using a predetermined color, but is not limited thereto.

11 is a diagram illustrating an example in which a device recommends a pseudo morpheme in consideration of a user's frequency of use for each pseudo morpheme, according to some embodiments.

When recommending at least one pseudo morpheme including a character input by the user, the device 1000 according to some exemplary embodiments may make the recommendation in consideration of the user's frequency of use of each of the pseudo morphemes. The device 1000 includes not only information on the general frequency of use of each pseudo morpheme, but also statistical data about the typical frequency of use of characters input by the user in the past, and at least one pseudo morpheme recommended by the device 1000 . At least one pseudo morpheme may be recommended in consideration of information on the pseudo morpheme selected by the user, information on voice data input by the user in the past, and the like. For example, the device 1000 may provide information about characters input by the user in the past, information on a pseudo morpheme selected by the user in the past from among at least one pseudo morpheme recommended by the device 1000 , and information about a pseudo morpheme selected by the user in the past. It is possible to collect information about the voice data entered in the . The device 1000 may learn a pseudo morpheme frequently used by the user by using the collected information, and may set a relatively larger weight for the pseudo morpheme frequently used by the user based on the learned result. The device 1000 may update information on the user's frequency of use for each of the pseudo morphemes when a new character is input from the user or an input for selecting one of the recommended pseudo morphemes is received. The device 1000 may recommend at least one pseudo morpheme in consideration of a weight set for the pseudo morpheme.

For example, referring to FIG. 11 , Table 1 1101 shows a list of pseudo morphemes that user A frequently uses. For example, when the user A inputs the character “wa” into the character input window 1110 of the device 1000 , the device 1000 determines that the user A frequently Pseudo morphemes that are frequently used can be recommended preferentially. For example, when user A frequently uses "want", "wait", and "walk" among pseudo morphemes including "wa" in the order, the device 1000 may display "want", "wait" , and “walk” may be preferentially recommended 1120 . Accordingly, the device 1000 may effectively recommend a pseudo morpheme that meets the intention of the user A.

12 is a diagram illustrating an example in which a device recommends a pseudo morpheme in consideration of the type and order of at least one previously displayed word, according to some embodiments.

In a language expression including at least two word phrases, the type of the word to be positioned next may vary depending on the kind of the word positioned before it. In this case, the type of word may include a part-of-speech including a noun, an adjective, a verb, or a type of a sentence component including a subject, a verb, an object, and a complement, but is not limited thereto. For example, the probability that a verb is located after a word corresponding to a subject is relatively high, and a probability that a noun is located after an adjective is relatively high. Accordingly, the device 1000 may recommend at least one pseudo morpheme in consideration of the type and order of at least one word previously displayed in the text input window, thereby recommending the pseudo morpheme and word appropriate to the context.

For example, referring to FIG. 12 , the word previously displayed in the character input window 1201 of the device 1000 is “He”, and as a part of the word to be positioned after “He”, the user uses the character “wa” If you input , the probability of a verb appearing after "He" corresponding to the subject is relatively high. Accordingly, the device 1000 may recommend 1202 verbs “wait”, “want”, and “watch” among pseudo morphemes including “wa”.

As another example, when a word previously displayed in the character input window 1203 of the device 1000 is “Clean” and the user inputs the character “wa” as a part of a word to be positioned after “Clean”, the adjective The probability of a noun appearing after "Clean" belonging to Accordingly, the device 1000 may recommend ( 1204 ) “water,” “wave,” and “way,” which are nouns, among pseudo morphemes including “wa”.

13 is a diagram illustrating an example in which a device recommends at least one word in consideration of a character input characteristic of a user, according to some embodiments.

Character input characteristics of the user may refer to expressions frequently used by the user. For example, the user may more frequently use the expression "--did" or "-did" instead of the expression "--did". The device 1000 may include information about text input by the user in the past, information on a word selected by the user in the past among at least one word recommended by the device 1000 , and voice data input in the past from the user. You can collect information about The device 1000 may determine the character input characteristics of the user by using the collected information. For example, the device 1000 may determine the character input characteristics of the user by performing learning according to the neural network technology. In recognizing the user's character input characteristics, what kind of information will be used may be determined according to learning based on a preset standard.

For example, referring to FIG. 13 , when the user inputs the character “expression” into the character input window 1301 of the device 1000, the device 1000 displays “expression” as a pseudo morpheme including “expression”. Diet”, “food”, and “meal” may be recommended ( 1310 ). The recommended “meal”, “food”, and “meal” may be displayed outside the text input window 1301 of the device 1000 . When a user input for selecting “meal” from among the displayed three pseudo morphemes is received, the device 1000 may recommend at least one word including “meal”. In this case, the device 1000 may recommend at least one word including “meal” in consideration of the user's character input characteristics. For example, when the user frequently uses the expression “--I did” rather than the expression “--I did”, the device 1000 may preferentially recommend “I ate” to “I ate” ( 1330 ). can

14 and 15 are block diagrams illustrating a configuration of a device according to some embodiments.

14 , the device 1000 according to some embodiments may include a user input unit 1100 , a display unit 1210 , a processor 1300 , and a memory 1700 . However, not all of the components shown in FIG. 14 are essential components of the device 1000 . The device 1000 may be implemented by more components than the components shown in FIG. 14 , or the device 1000 may be implemented by fewer components than the components shown in FIG. 14 .

For example, as shown in FIG. 15 , the device 1000 according to some embodiments includes a sensing unit 1400 in addition to the user input unit 1100 , the display unit 1210 , the processor 1300 , and the memory 1700 . ), an A/V input unit 1600 , and a communication unit 1500 may be further included.

The user input unit 1100 means a means for a user to input data for controlling the device 1000 . For example, the user input unit 1100 includes a key pad, a dome switch, and a touch pad (contact capacitive method, pressure resistance film method, infrared sensing method, surface ultrasonic conduction method, integral type). There may be a tension measurement method, a piezo effect method, etc.), a jog wheel, a jog switch, and the like, but is not limited thereto.

The user input unit 1100 according to some embodiments may receive a user input for inputting a character into the character input window of the device 1000 . The user input unit 1100 may receive a user input for selecting one from among at least one recommended pseudo morpheme or a user input for selecting one from at least one recommended word. Also, the user input unit 1100 may receive a user input for inputting a blank character in the character input window of the device 1000 , but is not limited to the above-described example.

The output unit 1200 may output an audio signal, a video signal, or a vibration signal, and the output unit 1200 may include a display unit 1210 , a sound output unit 1220 , and a vibration motor 1230 . have.

The display unit 1210 displays and outputs information processed by the device 1000 . For example, the display unit 1210 according to some embodiments may display a user interface for providing a recommended word based on a character input by the user. Also, the display unit 1210 may display at least one recommended pseudo morpheme based on a character input by the user and at least one recommended word when one of the recommended at least one pseudo morpheme is selected.

The sound output unit 1220 outputs audio data received from the communication unit 1500 or stored in the memory 1700 . Also, the sound output unit 1220 outputs a sound signal related to a function (eg, a call signal reception sound, a message reception sound, and a notification sound) performed by the device 1000 .

The processor 1300 typically controls the overall operation of the device 1000 . For example, the processor 1300 executes programs stored in the memory 1700 , and thus the user input unit 1100 , the output unit 1200 , the sensing unit 1400 , the communication unit 1500 , and the A/V input unit 1600 . ) can be controlled in general. Also, the processor 1300 may execute the programs stored in the memory 1700 to perform the functions of the device 1000 described in FIGS. 1 to 14 .

The processor 1300 according to some embodiments recommends at least one pseudo-morpheme including the input character by analyzing the input character. The processor 1300 may recommend at least one pseudo morpheme including the input text by using context information about the input text and context information of the user. For example, the processor 1300 may be configured to generate information about a typical frequency of use of each pseudo morpheme, information about characters input by the user in the past, and a pseudo morpheme selected by the user from among the pseudo morphemes recommended by the device 1000 . At least one pseudo morpheme may be recommended in consideration of information on the pseudo morpheme, information on voice data input by the user in the past, and the like.

The processor 1300 according to some embodiments recommends at least one word including the selected pseudo morpheme in response to an input for selecting one from among the at least one recommended pseudo morpheme. The user may select a pseudo morpheme that the user intends to input in the text input window from among at least one recommended pseudo morpheme. As an input for selecting a pseudo morpheme is received, the processor 1300 may recommend at least one word including the selected pseudo morpheme. In this case, when the selected pseudo morpheme is a noun having a meaning, at least one word including the pseudo morpheme may be a form in which a noun is combined with a proposition, but is not limited thereto.

Also, the processor 1300 may recommend at least one word including the selected pseudo morpheme based on the character input characteristic of the user. For example, the device 1000 may recommend at least one word including the selected pseudo morpheme in consideration of an expression frequently used by the user.

Also, the processor 1300 may recommend at least one other word including the selected word in response to an input for selecting one from among the at least one recommended word. In this case, the at least one other word including the selected word may include a word in which another proposition is added to the selected word, but is not limited thereto.

In addition, the processor 1300 may use a data recognition model stored in the memory 1700 or the server 2000 to provide a recommended word, which will be described in more detail with reference to FIGS. 16 to 19 .

In addition, the processor 1300 can efficiently learn a criterion for determining a pseudo morpheme to be recommended by using the data recognition model stored in the memory 1700 or the server 2000, and according to the learned result, the user's intention It is possible to provide a user with a recommended word matching the .

The sensing unit 1400 may detect a state of the device 1000 or a state around the device 1000 , and transmit the sensed information to the processor 1300 . The sensing unit 1400 may be used to generate some of the context information indicating the surrounding situation of the user or the device 1000 .

The sensing unit 1400 includes a magnetic sensor 1410 , an acceleration sensor 1420 , a temperature/humidity sensor 1430 , an infrared sensor 1440 , a gyroscope sensor 1450 , and a position sensor. (eg, GPS) 1460 , a barometric pressure sensor 1470 , a proximity sensor 1480 , and at least one of an illuminance sensor 1490 , but is not limited thereto. Since a function of each sensor can be intuitively inferred from a person skilled in the art from the name, a detailed description thereof will be omitted.

The communication unit 1500 may include one or more components that allow the device 1000 to communicate with another apparatus (not shown) and the server 2000 . The other device (not shown) may be a computing device such as the device 1000 or a sensing device, but is not limited thereto. For example, the communication unit 1500 may include a short-distance communication unit 1510 , a mobile communication unit 1520 , and a broadcast receiving unit 1530 .

Short-range wireless communication unit 1510, Bluetooth communication unit, BLE (Bluetooth Low Energy) communication unit, short-range wireless communication unit (Near Field Communication unit), Wi-Fi (WLAN) communication unit, Zigbee (Zigbee) communication unit, infrared ( It may include an IrDA, infrared Data Association) communication unit, a Wi-Fi Direct (WFD) communication unit, an ultra wideband (UWB) communication unit, an Ant+ communication unit, and the like, but is not limited thereto.

The mobile communication unit 1520 transmits/receives a radio signal to and from at least one of a base station, an external terminal, and a server on a mobile communication network. Here, the wireless signal may include various types of data according to transmission and reception of a voice call signal, a video call signal, or a text/multimedia message.

The broadcast receiver 1530 receives a broadcast signal and/or broadcast-related information from the outside through a broadcast channel. The broadcast channel may include a satellite channel and a terrestrial channel. According to an implementation example, the device 1000 may not include the broadcast receiver 1530 .

Also, the communication unit 1500 may transmit and receive information necessary for requesting a response message to the user's voice input and executing an operation related to the response message, with another device (not shown) and the server 2000 .

The A/V (Audio/Video) input unit 1600 is for inputting an audio signal or a video signal, and may include a camera 1610 , a microphone 1620 , and the like. The camera 1610 may obtain an image frame such as a still image or a moving image through an image sensor in a video call mode or a shooting mode. The image captured through the image sensor may be processed through the processor 1300 or a separate image processing unit (not shown). The image captured by the camera 1610 may be used as context information of the user.

The microphone 1620 receives an external sound signal and processes it as electrical voice data. For example, the microphone 1620 may receive an acoustic signal from an external device or a user. The microphone 1620 may receive a user's voice input. The microphone 1620 may use various noise removal algorithms for removing noise generated in the process of receiving an external sound signal.

The memory 1700 may store a program for processing and controlling the processor 1300 , and may store data input to or output from the device 1000 .

The memory 1700 may include a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (eg, SD or XD memory), and a RAM. (RAM, Random Access Memory) SRAM (Static Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic disk , may include at least one type of storage medium among optical disks.

Programs stored in the memory 1700 may be classified into a plurality of modules according to their functions, for example, may be classified into a UI module 1710 , a touch screen module 1720 , a notification module 1730 , and the like. .

The UI module 1710 may provide a specialized UI, GUI, or the like that interworks with the device 1000 for each application. The touch screen module 1720 may detect a touch gesture on the user's touch screen and transmit information about the touch gesture to the processor 1300 . The touch screen module 1720 according to some embodiments may recognize and analyze a touch code. The touch screen module 1720 may be configured as separate hardware including a controller.

The notification module 1730 may generate a signal for notifying the occurrence of an event in the device 1000 . Examples of the event generated in the device 1000 include call signal reception, message reception, key signal input, schedule notification, and the like. The notification module 1730 may output a notification signal in the form of a video signal through the display unit 1210 , may output a notification signal in the form of an audio signal through the sound output unit 1220 , and the vibration motor 1230 . It is also possible to output a notification signal in the form of a vibration signal through

16 is a block diagram of a processor 1300 in accordance with some embodiments.

Referring to FIG. 16 , the processor 1300 according to some embodiments may include a data learning unit 1310 and a data recognition unit 1320 .

The data learning unit 1310 may learn a criterion for determining a recommended word for a character input by the user. For example, the data learning unit 1310 may learn a criterion for determining a pseudo morpheme to be recommended and a word to be recommended according to a character input by the user. Also, the data learning unit 1310 may learn a criterion for determining whether to display a blank character after the recommended word. Also, the data learning unit 1310 may learn a criterion regarding which data to use to determine a pseudo morpheme to be recommended and a word to be recommended, and to determine whether to display a blank character. The data learning unit 1310 may learn a criterion for determining a recommended word for a character input from a user by acquiring learning data to be used for learning, and applying the acquired learning data to a data recognition model to be described later. .

Specifically, the data learning unit 1310 is training data for learning the criteria, from among information on characters or voice data input by the user in the past, and at least one pseudo morpheme recommended by the device 1000 . Information on the pseudo morpheme selected by the user, information on the word selected by the user from among at least one word recommended by the device 1000 , and statistical data on the frequency of common user use of each of the pseudo morphemes or words is available.

The data learning unit 1310 may learn a data recognition model based on the learning data through unsupervised learning. For example, the data learning unit 1310 may train the data recognition model so that a relatively higher probability or a greater weight is set for a pseudo morpheme or word that is frequently used by the user.

Alternatively, the data learning unit 1310 calculates the probability or weight of the position of the blank character after the specific word for each specific word by using the learning data (eg, statistical data on the case where the blank character is positioned after the specific word). You can train a data recognition model to be set up.

Alternatively, the data learning unit 1310 may learn or update the data recognition model by using the user's feedback according to the recognition result. For example, the data learning unit 1310 may train the data recognition model to set a greater weight for the pseudo morpheme or word selected by the user from among at least the pseudo morpheme or word recommended by the device 1000 .

The data recognition unit 1320 may determine a recommended word for a character input from the user based on the data. For example, the data recognizer 1320 may determine at least one pseudo morpheme including a character input by the user by using a learning result regarding the pseudo morpheme to be recommended. Also, the data recognition unit 1320 may determine at least one word including the pseudo morpheme selected by the user by using the learning result on the word to be recommended. According to another example, the data recognition unit 1320 may determine whether to display the blank character after the word selected by the user, using the learning result regarding the position of the blank character. The data recognition unit 1320 may determine at least one pseudo morpheme to be recommended and at least one word to be recommended by using the learned data recognition model, and determine whether to display a blank character. The data recognition unit 1320 may acquire predetermined data according to a preset criterion by learning, and use the acquired data as an input value to use a data recognition model. Also, the data recognition unit 1320 may determine, based on predetermined data, at least one pseudo morpheme to be recommended and at least one word to be recommended, and determine whether to display a blank character by using the data recognition unit 1320 . In addition, a result value output by the data recognition model using the obtained data as an input value may be used to update the data recognition model.

Specifically, the data recognition unit 1320 may obtain a character input from the user as recognition data. The data recognition unit 1320 may obtain at least one pseudo morpheme by applying the obtained character to the data recognition model. For example, the data recognition unit 1320 may obtain a predetermined number of pseudo morphemes according to the order in which the user's use probability is high or the weight is high by applying the obtained character to the data recognition model.

Alternatively, the data recognition unit 1320 may acquire a pseudo morpheme input by a user or selected from among a plurality of pseudo morphemes as recognition data. The data recognition unit 1320 may determine at least one word by applying the acquired pseudo morpheme to the data recognition model. For example, the data recognizer 1320 may apply the obtained pseudo morpheme to the data recognition model to obtain a predetermined number of pseudo morphemes according to an order in which a user's use probability is high or an order in which a weight is high.

Alternatively, the data recognition unit 1320 may acquire a word input by the user or selected by the user from among a plurality of words as recognition data. The data recognition unit 1320 may determine at least one other word by applying the obtained word to the data recognition model. For example, the data recognition unit 1320 may obtain a certain number of words according to the order in which the user's use probability is high or the weight is high by applying the obtained word words to the data recognition model.

Alternatively, the data recognition unit 1320 may obtain information on whether to display a blank character by applying the obtained word to the data recognition model. For example, the data recognition unit 1320 applies the obtained word to the data recognition model to determine whether to display the blank character according to whether the probability or weight of the position of the blank character after the specific word is above or below a certain standard. information can be obtained whether or not

Alternatively, the data recognition unit 1320 applies the obtained word to the data recognition model to obtain information on whether to display a blank character, and when the information is information indicating displaying a blank character, the blank At least one pseudo morpheme or word to be positioned after the character may be acquired. In this case, the data recognition unit 1320 may acquire a certain number of pseudo morphemes or words according to the order in which the user's use probability is high or the weight is high. At least a portion of the data learning unit 1310 and at least a portion of the data recognition unit 1320 may be manufactured in the form of at least one hardware chip and mounted in an electronic device. For example, at least one of the data learning unit 1310 and the data recognition unit 1320 may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or a conventional general-purpose processor (eg, CPU) Alternatively, it may be manufactured as a part of an application processor) or a graphics-only processor (eg, GPU) and mounted in the various electronic devices described above. In this case, the dedicated hardware chip for artificial intelligence is a dedicated processor specialized in probability calculation, and has higher parallel processing performance than conventional general-purpose processors, so that it can quickly process computational tasks in artificial intelligence fields, such as machine learning.

The data learning unit 1310 and the data recognition unit 1320 may be mounted on one electronic device, or may be mounted on separate electronic devices, respectively. For example, one of the data learning unit 1310 and the data recognition unit 1320 may be included in the device 1000 , and the other may be included in the server. In addition, the data learning unit 1310 and the data recognition unit 1320 may provide the model information constructed by the data learning unit 1310 to the data recognition unit 1320 through wired or wireless communication, and the data recognition unit ( Data input to 1320 may be provided to the data learning unit 1310 as additional learning data.

Meanwhile, at least a portion of the data learning unit 1310 and at least a portion of the data recognition unit 1320 may be implemented as a software module. When at least one of the data learning unit 1310 and the data recognition unit 1320 is implemented as a software module (or a program module including instructions), the software module is a computer-readable, non-transitory readable It may be stored in a recording medium (non-transitory computer readable media). Also, in this case, at least one software module may be provided by an operating system (OS) or may be provided by a predetermined application. Alternatively, a part of the at least one software module may be provided by an operating system (OS), and the other part may be provided by a predetermined application.

17 is a block diagram of a data learning unit 1310 according to some exemplary embodiments.

Referring to FIG. 17 , the data learning unit 1310 according to some exemplary embodiments includes a data acquiring unit 1310-1, a preprocessing unit 1310-2, a training data selection unit 1310-3, and a model learning unit 1310. -4) and a model evaluation unit 1310-5. According to various embodiments, the data learning unit 1310 essentially includes a data acquiring unit 1310 - 1 and a model learning unit 1310 - 4 , a preprocessing unit 1310 - 2 , and a training data selection unit 1310 - 3) and at least one of the model evaluator 1310-5 may be optionally further included. The data obtaining unit 1310-1 obtains learning data necessary to determine a recommended word for a character input from the user. can do. The data acquisition unit 1310-1 may acquire learning data necessary for learning to determine a pseudo morpheme to be recommended and a word to be recommended, and to determine whether to display a blank character after the recommended word.

The data acquisition unit 1310-1 may acquire, for example, a predetermined user voice and predetermined context information. Also, the data acquisition unit 1310-1 may acquire information on the pseudo morpheme or word selected by the user from among at least one pseudo morpheme or word recommended by the device 1000 .

The preprocessor 1310-2 determines a pseudo morpheme to be recommended and a word to recommend, and uses the obtained data for learning to determine whether to display a blank character next to the recommended word. Data can be preprocessed. The pre-processing unit 1310-2 uses the acquired data so that the model learning unit 1310-4, which will be described later, uses the acquired data for learning for determining the user's intention, providing related information, or recommending an alternative operation. It can be processed in a preset format.

The learning data selection unit 1310 - 3 may select learning data necessary for learning from among the pre-processed data. The selected training data may be provided to the model learning unit 1310 - 4 . The learning data selection unit 1310-3 determines a pseudo morpheme to be recommended and a word to be recommended, and learns from the preprocessed data according to a preset criterion for determining whether to display a blank character after the recommended word. You can select the training data you need. Also, the training data selection unit 1310-3 may select training data according to a preset criterion by learning by the model learning unit 1310-4, which will be described later.

The model learning unit 1310-4 determines a pseudo morpheme to be recommended and a word to be recommended based on the training data acquired through the data acquisition unit 1310 - 1 or selected by the training data selection unit 1310 - 3 , A criterion for determining whether to display a space character after a suggested word can be learned. In addition, the model learning unit 1310 - 4 determines a pseudo morpheme to be recommended and a word to be recommended, and a criterion for which training data should be used to determine whether to display a blank character after the recommended word. can also learn

In addition, the model learning unit 1310-4 determines a pseudo morpheme to be recommended and a word to be recommended, and uses the training data to apply a data recognition model used to determine whether to display a blank character after the recommended word. can learn In this case, the data recognition model may be a pre-built model. For example, the data recognition model may be a model built in advance by receiving basic training data (eg, sample data, etc.).

The data recognition model may be constructed in consideration of the field of application of the data recognition model, the purpose of learning, or the computer performance of the device. Data recognition models can be designed to simulate human brain structures on a computer. The data recognition model may include a plurality of network nodes having weights that simulate neurons of a human neural network. A plurality of network nodes may each form a connection relationship to simulate a synaptic activity of a neuron that transmits and receives a signal through a synapse. For example, a data recognition model may include a neural network model, or a deep learning model developed from a neural network model. In a deep learning model, a plurality of network nodes may exchange data according to a convolutional connection relationship while being located at different depths (or layers). For example, Deep Neural Network (DNN), Recurrent Neural (RNN) Network) and a model such as a Bidirectional Recurrent Deep Neural Network (BRDNN) may be used as the data recognition model, but is not limited thereto.

According to various embodiments, when a plurality of pre-built data recognition models exist, the model learning unit 1310 - 4 uses a data recognition model that has a high correlation between the input training data and the basic learning data as a data recognition model to learn. can decide In this case, the basic learning data may be pre-classified for each type of data, and the data recognition model may be previously built for each type of data. For example, the basic training data is pre-classified by various criteria such as the region where the training data is generated, the time when the training data is generated, the size of the training data, the genre of the training data, the creator of the training data, the type of object in the training data, etc. may have been

Also, the model learning unit 1310 - 4 may train the data recognition model using, for example, a learning algorithm including error back-propagation or gradient descent. .

Also, the model learning unit 1310 - 4 may train the data recognition model through, for example, supervised learning using at least a portion of the learning data as a criterion for determination.

In addition, the model learning unit 1310 - 4, for example, by learning the type of data required for determining the user's intention, providing related information, and recommending an alternative operation by using the training data without any guidance, by itself, The data recognition model can be trained through unsupervised learning to discover criteria for determining the intention of , providing related information, and recommending an alternative action. For example, the model learning unit 1310 - 4 may train the data recognition model in consideration of a user's frequency of use of a word or pseudo morpheme. Alternatively, the model learning unit 1310 - 4 may train the data recognition model by using feedback information that is information on a word or pseudo morpheme selected by the user.

In addition, the model learning unit 1310 - 4 is, for example, through reinforcement learning using feedback on whether a result of determining the user's intention according to learning, providing related information, and recommending an alternative operation is correct. , it is possible to train a data recognition model.

Also, when the data recognition model is learned, the model learning unit 1310 - 4 may store the learned data recognition model. In this case, the model learning unit 1310 - 4 may store the learned data recognition model in the memory of the electronic device (eg, the device 1000 ) including the data recognition unit 1320 . Alternatively, the model learning unit 1310 - 4 may store the learned data recognition model in a memory (eg, the device 1000 ) of the electronic device including the data recognition unit 1320 to be described later. Alternatively, the model learning unit 1310 - 4 may store the learned data recognition model in a memory of a server connected to the electronic apparatus (eg, the device 1000 ) through a wired or wireless network.

In this case, the memory in which the learned data recognition model is stored may also store, for example, commands or data related to at least one other component of the electronic device. The memory may also store software and/or programs. A program may include, for example, a kernel, middleware, an application programming interface (API) and/or an application program (or "application"), and the like.

The model evaluation unit 1310-5 may input evaluation data into the data recognition model, and when the recognition result output from the evaluation data does not satisfy a predetermined criterion, the model learning unit 1310-4 may be trained again. have. In this case, the evaluation data may be preset data for evaluating the data recognition model.

For example, the model evaluator 1310-5 sets a predetermined criterion when the number or ratio of evaluation data for which the recognition result is not accurate among the recognition results of the learned data recognition model for the evaluation data exceeds a preset threshold. It can be evaluated as unsatisfactory. For example, when the predetermined criterion is defined as a ratio of 2%, when the learned data recognition model outputs an erroneous recognition result for more than 20 evaluation data out of a total of 1000 evaluation data, the model evaluation unit 1310-5 can be evaluated that the trained data recognition model is not suitable.

On the other hand, when there are a plurality of learned data recognition models, the model evaluation unit 1310 - 5 evaluates whether each learned data recognition model satisfies a predetermined criterion, and recognizes a model that satisfies the predetermined criterion as final data It can be determined as a model. In this case, when there are a plurality of models that satisfy the predetermined criteria, the model evaluator 1310 - 5 may determine any one or a predetermined number of models preset in the order of the highest evaluation score as the final data recognition model.

On the other hand, in the data learning unit 1310, the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310 At least one of -5) may be manufactured in the form of at least one hardware chip and mounted in an electronic device. For example, at least one of the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310-5 One may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as part of an existing general-purpose processor (eg, CPU or application processor) or graphics-only processor (eg, GPU) as described above. It may be mounted on various electronic devices.

In addition, the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310-5 are one electronic unit. It may be mounted on the device, or may be respectively mounted on separate electronic devices. For example, some of the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310-5 may be included in the device 1000, and the remaining part may be included in the server.

In addition, at least one of the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310-5 is It may be implemented as a software module. At least one of the data acquisition unit 1310-1, the preprocessor 1310-2, the training data selection unit 1310-3, the model learning unit 1310-4, and the model evaluation unit 1310-5 is a software module When implemented as (or, a program module including instructions), the software module may be stored in a computer-readable non-transitory computer readable medium. Also, in this case, at least one software module may be provided by an operating system (OS) or may be provided by a predetermined application. Alternatively, a part of the at least one software module may be provided by an operating system (OS), and the other part may be provided by a predetermined application.

The processor 1300 (or control unit) may use various data recognition models, and through the data recognition model, it is possible to determine the user's intention in various ways, provide related information, and efficiently learn criteria for recommending alternative actions. can

18 is a block diagram of a data recognition unit 1320 according to some exemplary embodiments.

Referring to FIG. 18 , the data recognition unit 1320 according to some embodiments includes a data acquisition unit 1320-1, a pre-processing unit 1320-2, a recognition data selection unit 1320-3, and a recognition result providing unit ( 1320-4) and a model update unit 1320-5. In various embodiments, the data recognition unit 1320 essentially includes a data acquisition unit 1320-1 and a recognition result providing unit 1320-4, a pre-processing unit 1320-2, and a recognition data selection unit 1320 -3) and at least one of the model updater 1320-5 may be selectively further included.

The data acquisition unit 1320-1 may acquire data necessary to determine a pseudo morpheme to be recommended and a word to be recommended, and to determine whether to display a blank character. For example, the data acquisition unit 1320 - 1 may acquire a character, pseudo morpheme, or word input from a user as recognition data. Alternatively, the data acquisition unit 1320-1 may acquire a character, pseudo morpheme, or word selected by the user.

The pre-processing unit 1320 - 2 determines a pseudo morpheme to be recommended and a recommended word, and receives the recognition data obtained from the data obtaining unit 1320 - 1 so that the obtained recognition data can be used to determine whether to display a blank character. The acquired recognition data may be pre-processed. The preprocessing unit 1320-2 determines a pseudo morpheme and a recommended word to be recommended by the recognition result providing unit 1320-4, which will be described later, and uses the acquired data to determine whether to display a blank character. , the obtained data can be processed into a preset format. For example, the data acquisition unit 132 - 1 may acquire a character input to the device 1000 . Also, the data acquisition unit 132 - 1 may acquire context information related to the user or the device 1000 . Also, for example, the context information may be information generated by the device 1000 or information received from an external device.

The recognition data selection unit 1320 - 3 may determine a pseudo morpheme to be recommended and a recommended word from among the preprocessed recognition data, and may select recognition data necessary to determine whether to display a blank character. The selected recognition data may be provided to the recognition result providing unit 1320 - 4 . The recognition data selection unit 1320-3 is configured to select some or all of the preprocessed recognition data according to a preset criterion for determining a pseudo morpheme to be recommended and a word to be recommended, and determining whether to display a blank character. can Also, the recognition data selection unit 1320 - 3 may select recognition data according to a preset criterion by learning by the model learning unit 1310 - 4 to be described later.

The recognition result providing unit 1320 - 4 may apply the selected recognition data to the data recognition model to determine a pseudo morpheme to recommend and a recommended word, and determine whether to display a blank character. The recognition result providing unit 1320 - 4 may provide a recognition result according to the purpose of data recognition. The recognition result providing unit 1320-4 may apply the recognition data acquired through the data acquiring unit 1320-1 or the recognition data selected by the recognition data selection unit 1320-3 as an input value to the data recognition model. have. The recognition result providing unit 1320 - 4 may obtain information on whether to display at least one pseudo morpheme, at least one word, or a blank character as a result of applying the data recognition model.

The model updater 1320 - 5 may update the data recognition model based on the evaluation of the recognition result provided by the recognition result providing unit 1320 - 4 . For example, the model updating unit 1320-5 provides the recognition result provided by the recognition result providing unit 1320-4 to the model learning unit 1310-4, so that the model learning unit 1310-4 is The data recognition model can be updated.

Specifically, the evaluation of the recognition result may be user feedback information according to whether a pseudo morpheme recommended by the device 1000 to the user, at least one word, or a blank character is displayed. For example, the pseudo morpheme or word selected by the user from among at least one pseudo morpheme or at least one word recommended by the device 1000 may be used as user feedback information as training data for updating the data recognition model. Alternatively, when the device 1000 displays a blank character, the user input for inputting the next character by accommodating the blank character or removing the blank character is used as learning data for updating the data recognition model as user feedback information can be Alternatively, when the device 1000 does not display a blank character, a user input for inputting a blank character may be used as user feedback information as learning data for updating the data recognition model.

On the other hand, in the data recognition unit 1320, the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result providing unit 1320-4, and the model update unit ( 1320-5) may be manufactured in the form of at least one hardware chip and mounted in an electronic device. For example, among the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result provision unit 1320-4, and the model update unit 1320-5 At least one may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as part of an existing general-purpose processor (eg, CPU or application processor) or graphics-only processor (eg, GPU). It may be mounted on one various electronic devices.

In addition, the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result providing unit 1320-4, and the model update unit 1320-5 are one It may be mounted on an electronic device, or may be respectively mounted on separate electronic devices. For example, among the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result provision unit 1320-4, and the model update unit 1320-5 Some may be included in the device 1000 , and some may be included in the server.

In addition, at least one of the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result providing unit 1320-4, and the model update unit 1320-5 may be implemented as a software module. At least one of the data acquisition unit 1320-1, the preprocessor 1320-2, the recognition data selection unit 1320-3, the recognition result providing unit 1320-4, and the model update unit 1320-5 is software. When implemented as a module (or a program module including instructions), the software module may be stored in a computer-readable non-transitory computer readable medium. Also, in this case, at least one software module may be provided by an operating system (OS) or may be provided by a predetermined application. Alternatively, a part of the at least one software module may be provided by an operating system (OS), and the other part may be provided by a predetermined application.

In addition, the device 1000 may provide a service meeting the user's intention to the user by using the data recognition model to which the learned result is applied.

19 is a diagram illustrating an example of learning and recognizing data by interworking with the device 1000 and the server 2000 according to some embodiments.

Referring to FIG. 19 , the server 2000 may learn a criterion for determining a recommended word for a character inputted by the user, and the device 1000 may learn a criterion for determining a recommendation word for a character inputted by the user, and the device 1000 may be inputted from the user based on the learning result by the server 2000 . It is possible to determine a recommended word for the written character.

In this case, the model learning unit 2340 of the server 2000 may perform the function of the data learning unit 1310 illustrated in FIG. 16 . The model learning unit 2340 of the server 2000 may determine a pseudo morpheme to be recommended and a recommended word, and learn a criterion regarding which data to use to determine whether to display a blank character. In addition, the model learning unit 2340 of the server may learn a criterion regarding how to determine a pseudo morpheme to be recommended and a word to be recommended, and a determination as to whether to display a blank character by using the data. The model learning unit 2340 acquires data to be used for learning, and applies the acquired data to a data recognition model to be described later, thereby determining a pseudo morpheme to be recommended and a word to be recommended, and determining whether to display a blank character. standards can be learned.

In addition, the recognition result providing unit 1320 - 4 of the device 1000 applies the data selected by the recognition data selection unit 1320 - 3 to the data recognition model generated by the server 2000 to make recommendations It is possible to determine a morpheme and a recommended word, and determine whether to display a blank character. For example, the recognition result providing unit 1320-4 transmits the data selected by the recognition data selection unit 1320-3 to the server 2000, and the server 2000 causes the recognition data selection unit 1320-3 ) may be applied to the recognition model to determine a pseudo morpheme to be recommended and a word to be recommended, and to determine whether to display a blank character or not. In addition, the recognition result providing unit 1320 - 4 receives from the server 2000 information regarding the determination of the pseudo morpheme to be recommended and the recommended word determined by the server 2000 , and the determination of whether to display the blank character. can do.

Alternatively, the recognition result providing unit 1320 - 4 of the device 1000 receives the recognition model generated by the server 2000 from the server 2000, and uses the received recognition model to recommend a pseudo morpheme to be recommended. You can determine a word and decide whether to display a space character or not. In this case, the recognition result providing unit 1320 - 4 of the device 1000 applies the data selected by the recognition data selection unit 1320 - 3 to the data recognition model received from the server 2000 to recommend a pseudo morpheme and It is possible to determine a word to be recommended, and determine whether to display a blank character.

In addition, the device 1000 and the server 2000 can effectively distribute and perform tasks for learning and data recognition of the data recognition model, and through this, data processing can be efficiently performed in order to provide a service that meets the user's intention. and can effectively protect the user's privacy.

20 is a flow diagram of a device 1000 using a data recognition model in accordance with some embodiments.

The device 1000 may determine whether a user input for inputting text in the text input window is received (S2011).

When a user input is received (S2011-Y), the device 1000 may obtain at least one pseudo morpheme including the inputted character by applying the inputted character to the learned data recognition model (S2013). For example, the device 1000 may acquire at least one pseudo morpheme by applying the input character to the learned data recognition model stored in the memory 1700 of the device 1000 . The memory 1700 may include a memory provided inside the device 1000 or a memory card mounted in the device 1000 . In addition, the data recognition model includes information on text or voice data input by the user in the past, statistical data on typical frequency of use for each of the pseudo morphemes, and at least one pseudo morpheme recommended by the user selected by the user. It may be a model learned using at least one of information about pseudo morphemes.

The device 1000 may recommend the obtained at least one pseudo morpheme to the user (S2015). For example, the device 1000 may display the acquired at least one pseudo morpheme. In this case, the device 1000 may display only some of the at least one pseudo morpheme obtained by considering the priority of each of the pseudo morphemes.

Next, the device 1000 may determine whether a user input for selecting one of at least one pseudo morpheme is received ( S2017 ).

When a user input is received (S2017-Y), the device 1000 may apply the selected pseudo morpheme to the learned data recognition model to obtain at least one word including the selected pseudo morpheme (S2019). For example, the device 1000 may acquire at least one pseudo morpheme by applying the learned data recognition model stored in the memory 1700 of the device 1000 . The memory 1700 may include a memory provided inside the device 1000 or a memory card mounted in the device 1000 . In addition, the data recognition model is based on information on text or voice data input by the user in the past, statistical data on the typical frequency of use for each word, and a word selected by the user from among at least one word recommended by the device. It may be a model learned using at least one of the related information.

The device 1000 may recommend the obtained at least one word to the user ( S2021 ). For example, the device 1000 may display at least one acquired word. In this case, the device 1000 may display only some of the at least one pseudo morpheme obtained by considering the priority of each word.

The device 1000 may display the selected word in the text input window in response to an input for selecting one from among at least one word (S2023).

Meanwhile, in FIG. 20 , data recognition models providing pseudo morphemes and words may be the same data recognition model or different data recognition models.

Also, in FIG. 20 , an embodiment in which the device 1000 acquires both a pseudo morpheme to be recommended to a user and a word to be recommended to the user has been described using the learned data recognition model. Only one of the trained data recognition model set to provide , and the trained data recognition model set to provide the word may be selectively used.

21 to 23 are flowcharts of a network system using a data analysis model in accordance with some embodiments.

21 to 23 , the network system may include first components 2101,2201,2301 and second components 2102, 2202, 2302. Here, the first components 2101,2201,2301 may be the device 1000, and the second components 2102, 2202, 2302 may be the server 2000 in which the data analysis model is stored. Alternatively, the first components 2101,2201,2301 may be general-purpose processors, and the second components 2102, 2202, and 2302 may be artificial intelligence-only processors. Alternatively, the first component 2101,2201,2301 may be at least one application, and the second component 2102, 2202, 2302 may be an operating system (OS). That is, the second component 2102, 2202, 2302 is more integrated, dedicated, has a smaller delay, has superior performance, or uses more resources than the first component 2101,2201,2301. As a component having a data recognition model, many operations required when generating, updating, or applying a data recognition model can be processed more quickly and effectively than the first components 2101,2201,2301.

In this case, an interface for transmitting/receiving data between the first components 2101,2201,2301 and the second components 2102, 2202, and 2302 may be defined.

For example, an API (application program interface) having training data to be applied to the data recognition model as a factor value (or a parameter value or a transfer value) may be defined. An API is defined as a set of subroutines or functions that can be called for any processing of one protocol (eg, the protocol defined in the device 1000) to another protocol (eg, the protocol defined in the server 2000). can be That is, an environment in which an operation of another protocol can be performed in one protocol can be provided through the API.

First, the first component 2101 may receive a user input for inputting a character in the character input window (S2111).

The first component 2101 may transmit information on the input character to the second component 2102 ( S2113 ). For example, the first component 2101 may apply an input character as an argument value of an API function provided for use of the data recognition model. In this case, the API function may transmit the input text to the second component 2102 as recognition data to be applied to the data recognition model. In this case, the information on the input text may include communication data in which the text is changed or supplemented according to a communication format.

The second component 2102 may apply the input character to the learned data recognition model (S2115). As a result of the application, the second component 2102 may acquire at least one pseudo morpheme including the input character (S2117). In addition, the second component 2102 may transmit the acquired information on at least one pseudo morpheme to the first component 2101 ( S2119 ). Here, the information on the pseudo morpheme may include communication data obtained by changing or supplementing the pseudo morpheme into a communication format.

The first component 2101 may recommend the received at least one pseudo morpheme to the user (S2121). For example, the first component 2101 may display the received at least one pseudo morpheme.

The first component 2101 may recommend at least one word including the selected pseudo morpheme in response to an input for selecting one from among the at least one recommended pseudo morpheme ( S2123 ).

The first component 2101 may display the selected word in the text input window in response to a user input for selecting one of the at least one word (S2125).

As another embodiment, in FIG. 22 , the first component 2201 may recommend at least one pseudo morpheme including the input character in response to a user input of inputting a character into the character input window ( S2211 ). . For example, the first component 2201 may analyze the input character and display at least one pseudo morpheme to be recommended.

The first component 2201 may receive a user input for selecting one from among at least one recommended pseudo morpheme ( S2213 ).

The first component 2201 may transmit information on the selected pseudo morpheme to the second component 2202 ( S2215 ). For example, the first component 2201 may apply the selected pseudo morpheme as an argument value of an API function provided for use of the data recognition model. In this case, the API function may transmit the selected pseudo morpheme to the second component 2202 as recognition data to be applied to the data recognition model. In this case, the information about the input text may include communication data in which the pseudo morpheme is changed or supplemented according to a communication format.

The second component 2202 may apply the selected pseudo morpheme to the learned data recognition model ( S2117 ). As a result of the application, the second component 2202 may acquire at least one word including the selected pseudo morpheme ( S2219 ). In addition, the second component 2202 may transmit the acquired information on at least one word to the first component 2201 ( S2221 ). Here, the information on the word may include communication data obtained by changing or supplementing the word into a communication format.

The first component 2201 may recommend the received at least one word to the user (S2223). For example, the first component 2201 may display at least one received word.

The first component 2201 may display the selected word in the text input window in response to a user input for selecting one from among at least one word (S2225).

As another embodiment, in FIG. 23 , the first component 2301 may receive a user input for inputting text in the text input window ( S2311 ). The first component 2301 may transmit information on the input character to the second component 2302 (S2313).

The second component 2302 may obtain at least one pseudo morpheme including the input character by applying the input character to the learned data recognition model (S2315 and S2317).

The second component 2302 may transmit the acquired information on at least one pseudo morpheme to the first component 2301 ( S2319 ).

The first component 2301 may receive a user input for selecting one of the received at least one pseudo morpheme ( S2321 ).

The first component 2301 may transmit information on the selected pseudo morpheme to the second component 2302 ( S2323 ).

The second component 2302 may obtain at least one word including the selected pseudo morpheme by applying the selected pseudo morpheme to the learned data recognition model ( S2325 and S2327 ).

The second component 2302 may transmit the acquired information on at least one word to the first component 2301 ( S2329 ).

The first component 2301 may recommend at least one received word to the user (S2331). In response to a user input of selecting one of the at least one word, the first component 2301 may display the selected word in the text input window ( S2333 ).

According to various embodiments, in FIG. 23 , the data recognition model set to provide at least one pseudo morpheme including the input character and the data recognition model set to provide at least one word including the selected pseudo morpheme include one data It can be implemented as a recognition model.

For example, operations S2315 , S2317 , S2325 , and S2327 included in 2340 of FIG. 23 may be performed through one data recognition model.

Specifically, in FIG. 23 , the second component 2302 may receive information about the input character ( S2313 ). In this case, the second component 2302 may obtain at least one word including the input character by applying the information on the input character to the learned data recognition model.

For example, when receiving a user input of inputting “four” in the text input window, the first component 2301 may transmit the letter “four” to the second component 2302 . The second component 2302 may obtain at least one word such as “love”, “love”, and “with love” by applying the letter “four” to the learned data recognition model. When the second component 2302 transmits the obtained information on at least one word to the first component 2301, the first component 2301 can recommend at least one word related to the input character. have.

Some embodiments of this document may also be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. In addition, computer-readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transport mechanism, and includes any information delivery media.

In addition, the method according to the disclosed embodiments may be provided as a computer program product (computer program product).

The computer program product may include a S/W program, a computer-readable storage medium in which the S/W program is stored, or a product traded between a seller and a buyer.

For example, the computer program product is a product (eg, a downloadable app) in the form of a S/W program distributed electronically through the device 1000 or the manufacturer of the device 1000 or an electronic market (eg, an app store). ) may be included. For electronic distribution, at least a portion of the S/W program may be stored in a storage medium or may be temporarily generated. In this case, the storage medium may be a server of a manufacturer or an electronic market, or a storage medium of a relay server.

Also, in this specification, “unit” or “module” may be a hardware component such as a processor or circuit, and/or a software component executed by a hardware component such as a processor.

The above description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

Claims (20)

a memory in which at least one program is stored;
a user input unit for receiving a user input for inputting text in the text input window; and
at least one processor configured to provide a recommended word for the inputted character by executing the at least one program;
the at least one program,
recommending at least one pseudo-morpheme including the input character by analyzing the input character;
recommending at least one word including the selected pseudo morpheme based on an input of selecting one of the at least one pseudo morpheme and a character input characteristic of the user; and
and at least one command for executing an operation of displaying the selected word in the text input window in response to an input of selecting one from among the at least one word;
The device of claim 1, wherein the character input characteristic is obtained by performing learning according to a neural network technique based on information about voice data input from the user.
According to claim 1, wherein the at least one pseudo morpheme,
A device that includes at least one noun having a different meaning.
The method of claim 1,
The operation of recommending the at least one pseudo morpheme may include recommending a pseudo morpheme based on a result of analyzing the input character and a frequency of use of the user for each of the pseudo morphemes.
The method of claim 1,
The operation of recommending the at least one word may include recommending at least one word including the selected pseudo morpheme based on the text input characteristic of the user.
The method of claim 1,
The operation of recommending the at least one word is,
recommending at least one other word including the selected word in response to an input for selecting one of the at least one recommended word;
further comprising,
The operation of displaying the selected word in the text input window comprises:
displaying the selected other word in the text input window in response to an input of selecting one of the recommended at least one other word;
A device comprising a.
The method of claim 5, wherein the displaying of the selected word comprises:
predicting whether or not a blank character will be positioned next to the selected word based on information on voice data input from the user according to the execution of learning according to a neural network technique; and
recommending the other word including the selected word and other word not including the selected word based on the predicted result;
The device further comprising instructions to execute
The method of claim 1,
The operation of recommending the at least one pseudo morpheme is,
predicting the type of a word to be positioned next based on the type and order of at least one word previously displayed on the text input window; and
recommending the at least one pseudo morpheme including the input character based on the prediction result;
A device further comprising a.
7. The method of claim 6,
Another word that does not include the selected word is determined based on at least one word previously displayed in the text input window, the selected word, and a blank character.
The method of claim 1,
The operation of recommending the at least one pseudo morpheme is,
displaying the at least one pseudo morpheme outside the text input window; and
distinguishingly displaying the displayed at least one pseudo morpheme to indicate whether there is a word to recommend for each of the at least one pseudo morpheme;
A device further comprising a.
The method of claim 1,
The operation of recommending the at least one pseudo morpheme is,
and recommending the at least one pseudo morpheme obtained by applying the input character to a learned data recognition model,
The data recognition model includes information on characters or voice data input by the user in the past, statistical data on a typical frequency of use for each of the pseudo morphemes, and a user among at least one pseudo morpheme recommended by the device. The device, which is learned using at least one of information about the selected pseudo morpheme. In a method for a device to provide a recommended word for a character input from a user,
receiving a user input for inputting text in a text input window;
recommending at least one pseudo-morpheme including the input character by analyzing the input character;
recommending at least one word including the selected pseudo morpheme based on an input of selecting one of the at least one pseudo morpheme and a character input characteristic of the user; and
displaying the selected word in response to an input for selecting one of the at least one word;
including,
The method of claim 1, wherein the character input characteristic is acquired according to learning according to a neural network technique based on information about voice data input from the user.
12. The method of claim 11, wherein the at least one pseudo morpheme comprises:
A method comprising at least one noun having a different meaning.
12. The method of claim 11,
The method of recommending the at least one pseudo morpheme may include recommending a pseudo morpheme based on a result of analyzing the input character and a frequency of use of the user for each of the pseudo morphemes.
12. The method of claim 11,
The operation of recommending the at least one word is,
recommending at least one word including the selected pseudo morpheme based on the character input characteristic of the user;
A method further comprising:
12. The method of claim 11,
The operation of recommending the at least one word is,
recommending at least one other word including the selected word in response to an input for selecting one of the at least one recommended word;
further comprising,
The operation of displaying the selected word in the text input window comprises:
displaying the other selected word in the text input window in response to an input of selecting one of the recommended at least one word;
A method comprising
The method of claim 15, wherein the displaying of the selected word comprises:
predicting whether or not a blank character will be positioned next to the selected word based on information on voice data input from the user according to the execution of learning according to a neural network technique; and
recommending the other word including the selected word and other word not including the selected word based on the predicted result;
A method further comprising:
12. The method of claim 11,
The operation of recommending the at least one pseudo morpheme is,
predicting the type of a word to be positioned next based on the type and order of at least one word previously displayed on the text input window; and
recommending the at least one pseudo morpheme including the input character based on the prediction result;
A method further comprising:
17. The method of claim 16,
Another word that does not include the selected word is determined based on at least one word previously displayed in the text input window, the selected word, and a blank character.
12. The method of claim 11,
The operation of recommending the at least one pseudo morpheme is,
displaying the at least one pseudo morpheme outside the text input window; and
distinguishingly displaying the displayed at least one pseudo morpheme to indicate whether there is a word to recommend for each of the at least one pseudo morpheme;
A method further comprising:
A recording medium storing a computer-readable program including at least one command set to enable a device to perform the operation of claim 11 .

Images