JP7845704B2 - Character conversion support system and method for supporting character conversion of the character conversion support system - Google Patents

Description

This invention relates to a character conversion support system and a character conversion support method for a character conversion support system, comprising a client device capable of connecting to a specific domain via a predetermined communication medium, a client application launched on the client device, and an application server launched on a server device, all of which cooperate with each other.

Traditionally, character information processing in computer systems has involved a Front-End Processor (FEP) that, under the control of the operating system, converts input key codes into intended character information while referencing a dictionary library. In particular, the FEP plays a crucial role in character processing within Japanese input systems.

The FEP (Front-End Processor) is responsible for receiving input from the keyboard and converting it into Japanese characters such as hiragana and kanji. This process includes steps such as converting the entered Roman letters to hiragana and then presenting candidate kanji or words for the user to select.
In addition to the above, FEP functions include suggesting and selecting candidate words, and supporting multiple input methods (e.g., Roman alphabet input, kana input).

On the other hand, FEPs (Front-End Processors) have a problem where unintended character conversions occur. This is particularly noticeable with technical terms, new words, and loanwords.

Patent Document 1, described below, discloses that, "To enable input and display of diverse characters even when limited computing resources are available, the web browser of client 30 accesses the web server of dictionary center 20 to receive a character input service. The web browser inputs the reading of the character, the dictionary center's web server converts this into a character code, and further converts it into a URL of an image file on the font center's web server 10. As a result, an HTML document containing the URL of the character's image file as the source is sent to the browser, the browser retrieves the image file from the font center's web server 10 based on the URL, and displays the character using the retrieved image."

Furthermore, Patent Document 2 discloses that, "To convert audio files into text and identify speakers using a learning-type server via the Internet, the speech recognition system 1 consists of a sound collection unit 2 that inputs audio, a processing unit 3 that generates an audio file F from this audio, transmits this audio file F to a character conversion server and a speaker identification server, and receives a text file W and the speaker identification results, and a monitor unit 4 that displays the processing results of the processing unit 3. The audio data V, separated by silent periods between speaker H's statements, is sent as the audio file F to a character conversion server 5 and a speaker identification server 6, which are cloud services that perform data analysis and interpretation through self-learning based on data collected from numerous users via the Internet. The received text file W and the speaker identification results are then associated with the audio file F and displayed in chronological order on the monitor unit 4."

Japanese Patent Publication No. 2001-350747 Japanese Patent Publication No. 2020-060735

In Japanese character conversion systems, it is widely known that obtaining a perfect sentence from input hiragana, audio, or gestures alone is difficult when converting hiragana to kanji. Appropriate expressions cannot be generated without considering a wide range of surrounding context and conversational situations.

In particular, in languages like Japanese, which have many homophones and polysemous words based on ideographic characters, the input sentence "senshu ga senshu o naoshimashita" (the bow of the ship repaired the bow) can be interpreted as either "the player healed the player" or "the player treated the player," depending on the context. Furthermore, in actual use, input errors such as typos are often mixed in with the initial input.

For example, if "senshu ga senshu o naoshimashita" (the captain repaired the captain) is entered as "senshu ga senshu o naoshimashita" (the bow repaired the bow), or if it's a typo and entered as "senshu ga senshu o naoshimashita" (the ship's bow repaired last week), it might be entered as "sanshuu no o naoshimashita" (last week repaired the ship's owner). As you can see, even slight differences in input can significantly alter the interpretation (the resulting string).

In particular, when various users connect to a specific domain from client devices and input text information on the provided UI screen, there are issues such as the text not being converted to the intended string, or the text conversion process taking a considerable amount of time when long texts are entered.

The present invention solves the above problems and provides a character conversion support system and a character conversion support method for the character conversion support system. By employing conversion candidate strings obtained from an inference engine that infers conversion strings provided by a server device connected to the network, the functional processing of the client device's front-end processor can enable the same user operating the client device to obtain and output highly accurate conversion string candidates for input character information, even when the attributes of the input device, the attributes of the application being launched, the attributes of the service being launched, and the input method and style of character information change, in various character processing environments.

The character conversion support system of the present invention, which achieves the above objective, has the following configuration.

The character conversion support system according to the present invention is a character conversion support system in which a client device having a front-end processor and an input unit for string processing and a server device that acquires strings from the front- end processor communicate via a predetermined communication medium to coordinate string conversion processing, wherein the client device has a configuration in which the front-end processor acquires strings and user information input by the input unit, and sequentially transfers these to the server device each time characters are input from the input unit, and each time the transfer occurs, the server device identifies the context based on the input history including the user information and the strings sequentially input from the input unit, re-evaluates the context by rewinding to the beginning of previously input sentences, performs conversion processing in real time using one of a plurality of inference engines using a large-scale language model based on the re-evaluation result, and returns a converted string candidate generated reflecting the identified context and the strings sequentially input from the input unit as a response to the client device, and the client device performs processing to display the returned converted string candidate .

According to the present invention, by employing conversion candidate strings obtained from an inference engine that infers conversion strings provided by a server device connected to the network, the client device's front-end processor can obtain and output highly accurate conversion string candidates for the input character information, even when the attributes of the input device, the attributes of the application being launched, the attributes of the service being launched, and the input method and style of the character information change, the same user operating the client device can do so in various character processing environments.

The drawings illustrate specific embodiments of the present invention, including not only essential components of the invention but also optional and preferred embodiments.
A block diagram illustrating the configuration of the character conversion support system shown in this embodiment. A block diagram illustrating the software functions of the character conversion support system shown in Figure 1, assuming the hardware resources available. A block diagram illustrating the hardware configuration of the server device shown in Figure 1. A block diagram illustrating the hardware resources and software functions of the client device shown in Figure 1. A flowchart showing an example of a character conversion support method for the character conversion support system of this embodiment. A flowchart showing an example of a character conversion support method for the character conversion support system of this embodiment.

Next, the best mode for carrying out the present invention will be described with reference to the drawings.

<System Configuration Description>
[First Embodiment]
Figure 1 is a block diagram illustrating the configuration of the character conversion support system shown in this embodiment. This example describes a character conversion support system in which client devices CL1 to CLN, which can connect to a specific domain via a predetermined communication medium, a client application 100 launched on client devices CL1 to CLN, and an application server 11 launched on server device 1 cooperate to support character conversion processing.

In Figure 1, 1 is a server device that communicates with client devices CL1 to CLN via network 21 using a predetermined protocol.

In server device 1, 11 is an application server that communicates with client applications on client devices CL1 to CLN via the network 21. 12 is a Domain Knowledge Server Selection Unit (DKS) that receives user information, service information (uniquely assigned service number (Universal Unique Identifier)), application information, document properties, etc., from the front-end processor (FEP) 102, which performs character conversion processing based on a predetermined language, using information input by the input unit 103 of client devices CL1 to CLN. The DKS 12 then executes a process to select one candidate from the knowledge server group 13 for the second inference processing in the character conversion process.

50 is a Domain Name Server (DNS) that provides IP addresses corresponding to domain names entered from the input sections 103 of client devices CL1 to CLN.

Client devices CL1 to CLN connect to specific URLs 1 to URLN via network 22, routed according to the IP address (xxx.xxx.xxx) specified by DNS 50. These specific URLs 1 to URLN cover a wide range of genres, and it has been observed that when engaging in conversations or chats appropriate to each URL, specific terminology is likely to be selected, and there are certain trends in the strings used in mutual conversations.

In this embodiment, the FEP 102 requests a second inference process from the server device 1 in the background to infer character information, based on the domain name entered by the user of client devices CL1 to CLN from the input unit 103.

Figure 2 is a block diagram illustrating the software functions of the character conversion support system shown in Figure 1, assuming the hardware resources of the system. Components identical to those in Figure 1 are denoted by the same reference numerals.

In Figure 2, the server device 1 comprises multiple second inference units 13-1 to 13-N, each including multiple knowledge servers 13-1-1 to 13-1-N and profiles 13-2-1 to 13-2-N.

In client devices CL1 to CLN, the switching unit 101 performs a process to switch between executing the first inference process locally during character conversion or executing the first inference process on the server device 1 side during character conversion, based on the information (including the domain name) input by the input unit 103.

The FEP 102 comprises a first inference unit 102-1 and a storage unit 102-2 that holds user profile information (including a user dictionary). The storage unit 102-2 is configured to hold user information to be transferred to the server device 1. In this embodiment, the user information includes the properties of the document used by the user, the attributes of the service used by the user, and the attributes of the application used by the user, as characteristics of the character processing used by the user. Furthermore, the user information held in the storage unit 102-2 is maintained in a way that allows it to be updated in accordance with the user's input environment.

Figure 3 is a block diagram illustrating the hardware configuration of the server device 1 shown in Figure 1. Components identical to those in Figure 1 or Figure 2 are denoted by the same reference numerals.

In Figure 3, 31 is the communication unit, which performs functional processing to communicate with client devices CL1 to CLN connected to the network 21 using a predetermined protocol.

32 is the CPU, which loads the application server 11 (various application programs, including the operating system) stored in external memory 33 into RAM 36 and executes it.

34 is the inference engine unit, where the second inference units 13-1 to 13-N are deployed. These second inference units 13-1 to 13-N constantly collect and update large-scale natural language texts and usage examples (corpora) from the social domains the user interacts with, maintaining an evolving degree of matching for character conversion candidates.

36-1 is the DKS unit, which performs a function to transfer conversion string candidates, inferred by one of the knowledge servers 13-1-1 to 13-1-N selected by the CPU 32 from the inference engine unit 34, to the FEP 102 of one of the client devices CL1 to CLN via the network 21. This function is continuously performed while the user operating client devices CL1 to CLN is inputting strings via the input unit 310.

Unit 39, located in the AI Support Department, executes processes that support innovative character conversion in the field of Natural Language Processing (NLP), which involves building large-scale language models (LLMs) using vast amounts of text data and advanced deep learning techniques.
For large-scale language models, T5 (Text-to-Text Transformer Transformer) is used. Numerous derived models based on this architecture have been proposed. T5 is a series of Transformers announced in 2017, and it learns by taking hiragana text as input and outputting text converted to kanji.

Furthermore, the inference engine of the knowledge server group 13, which is equipped with server device 1, works in cooperation with the AI support unit 39 to perform processing that improves the accuracy of conversion string candidates for kana characters acquired sequentially from client devices CL1 to CLN.

Alternatively, the functional processing of the AI support unit 39 may be integrated into the inference engine.

LLM learning uses "supervised learning," where hiragana sentences are used as input and a corresponding set of kanji sentences (corpus) is provided as training data. Specifically, the output of T5 (as described above) is calculated by inputting hiragana sentences and comparing it to the training kanji sentences to determine the loss. Based on this loss, the model parameters are updated according to the gradient of each parameter.

The inference engine has a corpus for each language community. For example, in formal settings, it uses texts from newspapers, magazines, and books, while in casual settings, it uses conversational texts and social media texts.

The corpus is constantly updated. This means the inference engine doesn't learn from the user's history, but from the entire society to which the user belongs. To achieve this, the inference engine continuously collects a wide range of information, automatically extracting the corresponding hiragana sentences from mixed kana and kanji texts through morphological analysis, and then combining them with the original text to create "kana-kanji sentence pairs." By collecting, adding, and updating a large number of these pairs, the learning process is deepened.

Figure 4 is a block diagram illustrating the hardware resources and software functions of client devices CL1 to CLN shown in Figure 1. Components identical to those in Figure 1 are denoted by the same reference numerals. Client devices CL1 to CLN are assumed to represent various data terminals, including PAD devices, personal computers (PCs), and smartphones.

Furthermore, in this embodiment, when the kana character conversion mode is selected, the FEP 102, via the communication unit 304, transfers the key code information and user information (details will be described later) to the DKS 12 of the server device 1 connected to the network 21, along with the key code input from the input unit 310.

The DKS12 selects a second inference unit 13-1 to 13-N to determine a conversion candidate string based on the key code information and user information received from one of the client devices CL1 to CLN. The response unit 36-2 then sends the converted Japanese string data containing kanji characters to the FEP 102 of the requesting client device CL1 to CLN via the communication unit 31.

In this case, the FEP 102 is controlled so that the browser outputs the converted string candidates from the server device 1 to the character input line on the display screen of the display 311, as if they were string candidates converted by the server device 1 itself.

In this process, the server device 1 can freely select one of the second inference units 13-1 to 13-N according to the TPO (Time, Place, and Occasion) of the user operating the client devices CL1 to CLN. Even during conversion, the server device 1 can flexibly switch to another second inference unit performing string conversion.

Furthermore, DKS12 is configured to recognize, for example, whether the input from the user operating client devices CL1 to CLN is polite Japanese, or special phrases used on social networking services (strings not typically used in formal documents), such as "buzz" (ばずり), and to allow for the dynamic switching of string conversion processing by a second inference unit even while conversion is in progress.

In Figure 4, 301 is the CPU, which starts the BIOS stored in ROM 302, controls the I/O of input and output devices, and executes various data processing operations by running various applications loaded into RAM 303.

304 is the communications unit, which communicates with the server device 1 via the network 21 using a predetermined protocol and performs various data processing in the network environment.

311 is a display. The kana character information entered by the input unit 310 is passed to the server device 1. The server device 1's second inference units 13-1 to 13-N continuously receive conversion string candidates, and the character information is displayed so that it is arranged at the position specified by the cursor.

Furthermore, while the user operates the input unit 310, they continue to receive conversion candidate strings from the server device 1. If the user performs an action to correct the displayed conversion string, the server device 1 selects one of the knowledge servers 13-1-1 to 13-1-N to handle the most optimal second conversion process.

[Data processing on the server side]
Figure 5 is a flowchart showing an example of a character conversion support method for the character conversion support system in this embodiment. Note that (1) to (9) are data processing on the server device 1 side shown in Figure 1, which is realized by the CPU 32 loading the control program stored in the external memory 33 onto the RAM 36 and processing it.

First, the CPU 32 acquires user information (including domain information) from the FEP 102 of one of the client devices CL1 to CLN shown in Figure 1, and kana character information input from the input unit 310 (1).

The CPU 32 passes the kana character information and user information obtained from the client devices CL1 to CLN shown in Figure 1 to the DKS 12 (2). The DKS 12 then selects one of the knowledge servers 13-1-1 to 13-1-N based on the acquired domain information (3).

Next, the CPU 32 passes the string candidate converted by the selected knowledge server via the network 21 to one of the FEP 102s of client devices CL1 to CLN (4).

Next, the CPU 32 determines whether it has received a correction or re-conversion instruction for the conversion string candidate passed from the FEP 102 of any of the client devices CL1 to CLN (5). If the CPU 32 determines that it has received a correction or re-conversion instruction, it then determines whether it should continue to select one of the knowledge servers selected in step (3) (6). If the CPU 32 determines that it will not select the same knowledge server, it then controls the DKS 12 to select one of the knowledge servers 13-1-1 to 13-1-N (7). The CPU 32 then returns the process to step (4).

On the other hand, if in step (6) the CPU 32 determines to continue selecting one of the same knowledge servers 13-1-1 to 13-1-N selected in step (3), it returns to step (4).

On the other hand, in step (5), if a certain amount of time has elapsed since the correction or re-conversion instruction for the conversion string candidate passed from the FEP 102 of any of the client devices CL1 to CLN (8), the CPU 32 determines whether it has continued not to acquire user information (including domain information) from any of the FEP 102 of the client devices CL1 to CLN shown in Figure 1, and whether it has not received text information from the input unit 310, that is, whether it has received kana character information from any of the FEP 102 of the client devices CL1 to CLN (9).

At this point, if the CPU 32 determines that it has received kana character information from one of the FEP 102s of client devices CL1 to CLN, it returns to step (2) and repeats the same process. On the other hand, if in step (9) the CPU 32 determines that it has not received kana character information from one of the FEP 102s of client devices CL1 to CLN, it terminates this process.

[Data processing on client devices CL1 to CLN]
Figure 6 is a flowchart showing an example of a character conversion support method for the character conversion support system illustrating this embodiment. Note that (11) to (19) are data processing on the client device CL1 to CLN side shown in Figure 1, which is realized by the CPU 301 loading a control program stored in an external memory (not shown) onto the RAM 303 and processing it.

First, when each user operating client devices CL1 to CLN instructs access to any of the domains, the CPU 301 launches a browser from external memory (not shown) (11). At that time, each user operates the keyboard on the input unit 310 to input kana characters, for example, using a Romanization method (12), and the FEP 102 outputs the sequentially input kana characters to the server device 1 (13).

At this point, the FEP 102 detects user information (document properties used by the user, services used by the user, applications used by the user) held in the holding unit 102-2 (14), and transfers the detected user information to the server device 1 (15).

Next, the CPU 301 obtains a candidate for the converted string from the knowledge server selected by the server device 1 (one of the second inference units 13-1 to 13-N) via the network 21 (16), and then displays the converted string on the string input line (indicated by the cursor) specified by the browser on the display 311 (17).

Next, the CPU 301 determines whether each user has given an editing instruction, such as correcting, modifying, or deleting the converted string displayed on the browser of the display 311 (18). If it determines that an editing instruction has been given, it returns to step (13).

On the other hand, if the CPU 301 determines in step (18) that no editing instruction has been given, it further determines whether kana characters have been entered from the input unit 310 (19). If the CPU 301 determines that no character code has been entered, it terminates the process of acquiring string candidates by the server device 1. If it determines that a character code has been entered, it returns to step (13) and repeats the same process.

[Effects of the First Embodiment]
According to this embodiment, the functional processing of the front-end processor 102 of client devices CL1 to CLN is performed using conversion candidate strings obtained from multiple inference engines (inference engines utilizing LLM) that infer conversion strings provided by the server device 1 connected to the network 21. This allows the same user operating client devices CL1 to CLN to obtain and output highly accurate conversion string candidates for the input character information, even when the attributes of the input device, the attributes of the application being launched, the attributes of the service being launched, and the input method and style of the character information change in various character processing environments.

Furthermore, the system operator may freely choose whether to offer the free service of providing string conversion candidates provided by the server device 1 to general users, or to offer it as a paid service with customized functions for specific users.

Furthermore, while the system described above uses Japanese as an example of a language model, it is also possible to target major foreign language models for the language model.
Furthermore, in environments where users of different languages exchange text information, such as through chat, it is possible to freely construct an environment where similar string conversion processing can be performed by using a system that supports translation models.

[Second Embodiment]
In the above embodiment, an example was described assuming that short sentences are input from client devices CL1 to CLN, but of course, the present invention can also be applied to longer contexts.

In the system described above, it is anticipated that when server device 1 interprets a long context, the response speed until the conversion is visible will be delayed, at the expense of conversion accuracy.

Therefore, each time a user operating client devices CL1 to CLN reads a single character (one token) input from the input unit 310, the FEP 102 immediately sends the input character code and string conversion request to the second inference units 13-1 to 13-N of the server device 1 via the network 21.

In this way, when server device 1 receives a character code and string conversion request, the second inference units 13-1 to 13-N process the conversion to a string and respond to client devices CL1 to CLN with a character code indicating the generated string.

In this process, server device 1 controls the string conversion process to rewind to the beginning of each paragraph (translation unit) of the document before starting.

Therefore, the beginning of each paragraph will undergo repeated string conversion (rewinding), but since the input characters become longer with each conversion, improved conversion accuracy is expected.

As a result, client devices CL1 through CLN will display converted strings that automatically correct any initially mistranslated words as input progresses, ensuring proper correction.

With this type of conversion process, if the character input processing on client devices CL1 to CLN is fast, the burden on the string inference and conversion processing on server device 1 increases. Therefore, server device 1 controls the system to temporarily hold requests from each user in storage until the ongoing string conversion process is completed.

Then, when starting the next process, instead of executing the pending requests sequentially, the system executes only the last request and deletes the previously pending requests.

[Effects of the second embodiment]
As a result, although it appears to client devices CL1 to CLN that server device 1 is converting one character at a time in parallel, server device 1 only needs one thread per user.

Furthermore, since client devices CL1 to CLN intermittently retrieve the results of the string conversion process for long texts from server device 1, the display of client devices CL1 to CLN is controlled so that newly added hiragana strings from the last string sent to server device 1 are concatenated to the end of the last converted text.

By doing this, even if there is conversion latency in the server device 1 during the long chunk inference process, the CPU 301 controls the display on the client device CL1-CLN's display 311 so that the hiragana input is automatically converted to the appropriate text without delay.

Note that while rewind processing is generally performed on a paragraph basis, it can be adjusted from a minimum of one sentence to a maximum of the entire document being processed, based on the additional status on the server device 1.

[Third Embodiment]
In the above embodiment, an example was described assuming that short text is input from client devices CL1 to CLN. However, the display 311 of client devices CL1 to CLN may be configured to control the re-conversion process even if the conversion process has already been completed for the converted string, etc.

As mentioned above, even after performing various string conversion processes, it is still possible that contextual misinterpretations may remain.

In this case, the system may incorporate a configuration that controls the string conversion process so that each user operating client devices CL1 to CLN can utilize the intermediate results of the inference, thereby maximizing the use of the intermediate results and reducing the burden of string input processing on the user.

Specifically, in server device 1, the second inference units 13-1 to 13-N secure a conversion buffer on memory resources (not shown), and buffer all the plaintext (raw hiragana characters entered by each user and the trajectory of flick gestures) before the current paragraph conversion on this buffer. Therefore, it is configured to allow undoing (rewinding) even after the conversion has been confirmed.

Therefore, in server device 1, the second inference units 13-1 to 13-N maintain a confidence level (0.0 to 1.0 (probability notation)) for each phrase of the inference result, and instruct the FEP 102 of client devices CL1 to CLN to highlight a portion of the generated document according to this confidence level.

[Effects of the third embodiment]
This allows each user of client devices CL1 to CLN to quickly identify the areas to be corrected on the screen of display 311, return to that point, selectively revert to the pre-conversion state on a phrase-by-phrase basis, and then proceed to the appropriate correction work.

Therefore, unlike simple kana-kanji conversion, this embodiment allows users to rewind and perform editing even after the character conversion process has been finalized, thus reducing the psychological burden on each user when rewinding.

Furthermore, the highlighting function based on the confidence level mentioned above is not mandatory. Alternatively, when a user manipulates the cursor on the screen, the system may implement a control that rewinds only the relevant phrase and selectively displays only the original, unconverted string.

Furthermore, the string conversion inference process performed by the second inference units 13-1 to 13-N in server device 1 may involve simple input errors or insufficient expression in the original text or conversation.

Therefore, even if only a single incorrect particle is corrected, controlling the system to re-evaluate the entire paragraph allows the improved accuracy of the correction to ripple through to other parts, automatically proofreading the surrounding text.

[Fourth Embodiment]
In the above embodiment, an example was described assuming that short sentences are input from client devices CL1 to CLN. However, when each user inputs text from the input unit 310, a lightweight first inference unit 102-1 may be provided on the client devices CL1 to CLN to flexibly handle the situation, such as removing typos and noise before the user input.

In this way, instead of relying solely on the string conversion inference processing by the server device 1, the first inference unit 102-1 on the client devices CL1 to CLN cooperates with the inference processing on the server device 1. This eliminates user-specific key input errors, restricts the transfer of user-specific strings that interfere with inference to the second inference units 13-1 to 13-N to the server device 1, and prevents contamination of the inference processing in the second inference units 13-1 to 13-N, thereby enabling the acquisition of highly accurate string conversion processing results from the server device 1.

[Effects of the fourth embodiment]
In this case, the local first inference unit 102-1 on the client device CL1 to CLN side, assuming the input unit 310 is a keyboard, does not target the Romanized string but directly acquires information about the keystroke itself and applies a predetermined inference process specific to each user, thereby enabling it to extract a more accurate string and reflect it in the inference process on the server device 1 side.

Thus, when the text input mode can be selectively switched between direct keyboard input and flick input, by selecting the second inference units 13-1 to 13-N according to the respective input mode, an improvement in string conversion accuracy can be expected for each input mode.

[Effects of the fourth embodiment]
In the above embodiment, an example was described assuming that short sentences are input from client devices CL1 to CLN. However, when each user inputs text from the input unit 310, a lightweight first inference unit 102-1 may be provided on the client devices CL1 to CLN to flexibly handle the situation, such as removing typos and noise before the user input.

Furthermore, the first inference unit 102-1 is reserved for each user on the client device CL1 to CLN side and is equipped with a dedicated personal dictionary memory.

[Fifth Embodiment]
In the above embodiment, an example was described assuming that short sentences are entered from the keyboard of the input unit 310 from client devices CL1 to CLN. In particular, there are individual differences in the movement of fingertips and the speed at which fingertips strike keys. For example, by measuring and learning the frequency of input fluctuations and typos specific to Roman alphabet input and implementing control that reflects this in the string conversion candidates, it is possible to improve the conversion accuracy.

In particular, since the characteristics of the key input type differ between users who input "i" using "xi" and those who input "li," the system is controlled so that both "seaman" and "seaman n" output "seaman" as the conversion string candidate.

[Effects of the Fifth Embodiment]
According to this embodiment, in the second inference units 13-1 to 13-N that utilize LLM on the server device 1 side, input strings that could cause significant errors in generation can be corrected in advance.

Furthermore, in client devices CL1 to CLN that employ an input unit 310 for flick input of text strings, it becomes possible to correct errors specific to touch mistakes in the placement of kana characters.

[Sixth Embodiment]
In this embodiment, when a keyboard is used in client devices CL1 to CLN, the correlation between the time interval of keystrokes and the delimiter of phrases (tokens) may be used to control the system so that when the length of the keystroke interval exceeds a certain threshold, the following processing is executed.

Specifically, in server device 1, when each user of client devices CL1 to CLN uses the keyboard, if the length of the keystroke interval exceeds a certain threshold, a control mechanism is executed to insert a hidden delimiter.

[Effects of the sixth embodiment]
According to this embodiment, although user input is not echoed, by inserting the above-mentioned delimiter when the string is internally broken down into phrases, it is possible to suppress a chain reaction of misinterpretations that arises from errors in phrase division that span conjunctions such as "wa" and "e," which are likely to occur when converting long sentences all at once.
Furthermore, the same issue is expected to occur with voice input as well, so the same countermeasures can be expected to have the same effect.

[Seventh Embodiment]
In this embodiment, if there are multiple application services used by the same user in client devices CL1 to CLN, it is expected that the input tendencies will change even for the same user. Therefore, the server device 1 may be configured to perform control to switch the selection order among the second inference units 13-1 to 13-N that utilize LLM, so as to use the higher-ranking second inference unit.

Specifically, for example, when the same user performs text input for business purposes, the system will control the selection order so that it uses the highest-ranking second inference unit among the second inference units 13-1 to 13-N to select the target document name and the strings obtained during the process of converting from the beginning to the paragraphs.

[Effects of the 7th Embodiment]
According to this embodiment, from the perspective of each user operating client devices CL1 to CLN, it is expected that natural-sounding text will be generated according to the type of application being run, even though they are creating documents using the same interface from the same document creation software or spreadsheet software. In addition, when RAG, which combines external information retrieval, is adopted, it becomes unnecessary to go through AI chat during the document creation process, improving convenience.
Here, RAG is an abbreviation for Retrievable-Augmented Generation, which means search augmented generation.

On the other hand, for conversation and social media use, it becomes possible to perform long string inference processing across conversations, and simultaneously improve the accuracy of proper noun conversion by recording life logs.

Furthermore, by having the client application 100 provide the conversation partner's text, it becomes possible to include the previous conversation text from the previous partner or group in the string conversion inference. This is expected to effectively suppress misinterpretations and accidental messages, which frequently occur in conversation and social networking applications due to the short length of individual messages.

Furthermore, in the character conversion support system, the functions of the second inference units 13-1 to 13-N in server device 1 and the first inference unit 102-1 in client devices CL1 to CLN are clearly separated. This configuration eliminates the need for server device 1 to acquire user-specific user information (information derived from typing habits) and manage its history, thus preventing information linked to individual attributes (individual profiles) from leaking from server device 1.

Furthermore, if it is necessary to refer to the activity history of each user of client devices CL1 to CLN for business purposes, it is also possible to incorporate a strict account management function into the system independently of this.
Furthermore, information linked to individual attributes is limited to low-level statistical information such as typing habits, and this information is stored only on the individual's client devices CL1 to CLN, and can be controlled so that server device 1 does not need to know about it.

Furthermore, in this embodiment, the output unit that outputs the string candidates converted by the server device 1 is not limited to a display unit, but may also be a device that outputs the converted string candidates as audio information in each language (speaker, earphone with microphone, goggle device, etc.).

The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program (a program conforming to the flowcharts shown in Figures 5 and 6 above). Furthermore, it can also be realized by a circuit (e.g., an ASIC) that implements one or more functions.

The disclosures relating to the present invention described above can be summarized to at least the following:

(1) A character conversion support system in which a client device having a front-end processor and an input unit for string processing and a server device that acquires strings from the front-end processor communicate via a predetermined communication medium to coordinate string conversion processing, wherein the client device has a configuration in which the front-end processor acquires strings and user information input by the input unit, and sequentially transfers these to the server device each time characters are input from the input unit, and each time the transfer occurs, the server device identifies the context based on the input history including the user information and the strings sequentially input from the input unit, re-evaluates the context by rewinding to the beginning of previously input sentences, performs conversion processing in real time using one of a plurality of inference engines using a large-scale language model based on the re-evaluation result, and returns a candidate converted string that reflects the identified context and the strings sequentially input from the input unit as a response to the client device, and the client device performs processing to display the returned candidate converted string .

( 2 ) The string entered by the input unit is characterized by being entered by keyboard input or flick input.

( 3 ) The server device is characterized by having a configuration that selects one of a plurality of inference engines based on the character processing pattern included in the user information.

( 4 ) The character processing methods included in the user information are characterized by including the properties of the document used by the user, the attributes of the service used by the user, and the attributes of the application used by the user.

( 5 ) A character conversion support method in which a client device having a front-end processor and an input unit for string processing and a server device that acquires a string from the front-end processor communicate via a predetermined communication medium to coordinate string conversion processing, characterized in that, in the client device, the front - end processor acquires a string and user information input by the input unit, and sequentially transfers these to the server device each time a character is input from the input unit, and in the server device, each time a transfer occurs, identifies a context based on the input history including the user information and the string sequentially input from the input unit, rewinds to the beginning of a previously input sentence and re-evaluates the context, executes a conversion process in real time using one of a plurality of inference engines using a large-scale language model based on the re-evaluation result, returns a candidate converted string generated reflecting the identified context and the string sequentially input from the input unit as a response to the client device, and performs a process in the client device to display the returned candidate converted string .

The above system example describes a method where language conversion processing is initiated by passing a key code to FEP102 installed on client devices CL1 to CLN. However, this is a technique primarily used as preprocessing in the field of natural language processing. It is also possible to construct a system incorporating an analysis tool, such as mecab, that decomposes text into morphemes (the smallest units in which a word has meaning) based on the grammar and part-of-speech information of the target language.

Furthermore, the multiple second inference units 13-1 to 13-N on the server device 1 learn the converted strings and their degree of modification, either independently or in coordination with each other. This allows them to convert strings to specific strings tailored to each user's requirements, enabling them to evolve to present a timely array of conversion string candidates adapted to user-specific expressions and a wide range of versatile conversion candidates.

Furthermore, since the language types valid in client devices CL1 to CLN are already set up, it is possible to configure a system with worldwide translation processing capabilities even when different languages are used between client devices by arranging the multiple second inference units 13-1 to 13-N on the server device 1 side according to the language used.

Furthermore, by appropriately expanding features such as automatic typographical error correction, OS-specific language processing, automatic conversion triggers, customization options for individual user environments, context recognition, modeless functionality, and the ability to independently evolve the DNN, user convenience can be further improved.

1. Server devices CL1-CLN, Client device 21, Network 50, DNS

Claims (5)

A character conversion support system in which a client device having a front-end processor and input unit for character processing and a server device that acquires character strings from the front-end processor communicate via a predetermined communication medium to coordinate character conversion processing,
The aforementioned client device
The front-end processor acquires the string and user information input by the input unit.
Each time a character is input from the input unit, these characters are sequentially transferred to the server device.
The server device is
Each time the aforementioned transfer is performed, the context is identified based on the user information and the input history, which includes strings sequentially entered from the input unit.
Rewind to the beginning of the previously entered text and re-evaluate the context.
Based on the re-evaluation results, the conversion process is performed in real time using one of several inference engines that employ large-scale language models.
The system returns a candidate for a converted string, generated in accordance with the identified context and the strings sequentially input from the input unit, to the client device as a response.
The character conversion support system is characterized by having a client device that performs processing to display the returned conversion string candidates . The character conversion support system according to claim 1 , characterized in that the character string entered by the input unit is entered by keyboard input or flick input. The server device is
The character conversion support system according to claim 1 , characterized in that it includes a configuration to select one inference engine from among a plurality of inference engines based on the character processing method included in the user information. The character processing method included in the user information is characterized in that it includes the properties of the document used by the user, the attributes of the service used by the user, and the attributes of the application used by the user, as described in claim 1 . A character conversion support method in which a client device having a front-end processor and input unit for character processing and a server device that acquires a character from the front-end processor communicate via a predetermined communication medium to coordinate character conversion processing,
In the aforementioned client device,
The front-end processor acquires the string and user information input by the input unit.
Each time a character is input from the input unit, these characters are sequentially transferred to the server device.
In the aforementioned server device,
Each time the aforementioned transfer is performed, the context is identified based on the user information and the input history, which includes strings sequentially entered from the input unit.
Rewind to the beginning of the previously entered text and re-evaluate the context.
Based on the re-evaluation results, the conversion process is performed in real time using one of several inference engines that employ large-scale language models.
The system returns a candidate for a converted string, generated in accordance with the identified context and the strings sequentially input from the input unit, to the client device as a response.
In the aforementioned client device,
A character conversion support method characterized by performing a process to display the returned conversion string candidates .