KR102229130B1 - Apparatus for providing of digital broadcasting using real time translation - Google Patents

Abstract

The present invention relates to an apparatus for providing captions for digital broadcasting, and more particularly, to an apparatus for providing captions for digital broadcasting, capable of providing captions for digital broadcasting through real-time translation.
In addition, according to the present invention, there is provided a caption extraction unit for extracting and decoding captions from digital broadcasting data streamed at the request of a user terminal; A caption analysis unit for analyzing caption data extracted by the caption extracting unit in units of morphemes and reconstructing caption data including direct reading symbols; A translation request unit for requesting a translation of the caption data reconstructed by the caption analysis unit to a predetermined translation service device; A translation result aggregator for receiving and collecting the translated subtitle data translated by the translation request; And a translation caption transmission unit that receives the input of the language to be translated from the user terminal and transmits the translated direct reading direct translation caption data collected through the translation result collecting unit in synchronization with the digital broadcasting data. A caption providing service device is provided.

Description

Service device for providing closed captions for digital broadcasting through real-time translation {APPARATUS FOR PROVIDING OF DIGITAL BROADCASTING USING REAL TIME TRANSLATION}

The present invention relates to an apparatus for providing captions for digital broadcasting, and more particularly, to an apparatus for providing captions for digital broadcasting, capable of providing captions for digital broadcasting through real-time translation.

As a method of transmitting a broadcast signal has recently been developed, a digital TV to which a digital transmission method is applied is spreading.

In particular, high-definition and high-quality digital broadcasting broadcasted through satellite, terrestrial, and cable have several advantages, and can provide high-definition, high-quality sound, and various additional information compared to conventional analog broadcasting.

As one of various additional services, digital broadcasting has a demand for a broadcast subtitle machine translation system that extracts the subtitle signal from the broadcast signal and automatically translates it into the desired language so that the viewer can watch the broadcast in a language that can be understood. The situation is increasing.

Here, the broadcast caption refers to a caption signal inserted into an image signal transmitted by a broadcaster, and recently, many broadcasters have implemented such caption broadcasting for the hearing impaired. In the United States, closed captioning has been mandatory for 13-inch or larger TVs since 1990, and in Korea, TV broadcasters and CATV are expanding closed-captioned broadcasting programs, and the continuous expansion of foreign-language closed-caption broadcasting such as CNN, NHK, and AFKN is expected. It is expected.

However, in TV broadcast subtitles, colloquial and written language are mixed and used not only in news, but also in various broadcasting genres such as drama, liberal arts current affairs, and entertainment programs. It is technically difficult to develop a machine translation system for broadcast subtitles with translation capabilities.

In order to solve such technical difficulties, Korean Patent Publication No. 1997-56985 (published date: July 31, 1997) has a separate translation unit for Korean and foreign languages in the TV, so that broadcast subtitle data is provided in the native language or foreign language according to the user's request. A'TV with a broadcast subtitle translation function' that can be displayed to provide convenient viewing to viewers is disclosed.

However, since the TV having the broadcast subtitle translation function extracts broadcast subtitle data from the received broadcast signal, translates the subtitle data according to the user's request, and displays it on the TV screen, it is possible to separately support broadcast subtitles. There is a downside to buying a TV. In addition, when a broadcast signal is input through a media device other than a TV such as a satellite set-top box, a video player, and a DMB terminal, there is a problem that a caption translation function cannot be provided.

Korean Patent Publication No. 1997-56985 (Publication date: 1997.07.31)

The present invention was invented to solve the above problems, and when a user selects a desired digital broadcasting program and language, the selected digital broadcasting program is streamed while analyzing subtitles and translating the analyzed subtitles into a desired language to provide a digital broadcasting program and a digital broadcasting program. An object of the present invention is to provide a service device for providing closed captions for digital broadcasting that is synchronized and transmitted together.

The present invention includes a caption extraction unit for extracting and decoding captions from digital broadcasting data streamed at the request of a user terminal; A caption analysis unit for analyzing caption data extracted by the caption extracting unit in units of morphemes and reconstructing caption data including direct reading symbols; A translation request unit for requesting a translation of the caption data reconstructed by the caption analysis unit to a predetermined translation service device; A translation result aggregator for receiving and collecting the translated subtitle data translated by the translation request; And a translation caption transmission unit receiving input from the user terminal for a language to be translated, and transmitting the translated direct-read translation caption data collected through the translation result collecting unit in synchronization with the digital broadcasting data.

According to the present invention, when viewing a digital broadcasting program, a user can simultaneously view a subtitle translated into a language desired by the user, thereby maximizing the user's convenience. In particular, there is an effect that digital broadcasting can be viewed immediately without having to wait for a digital broadcasting program to be translated and converted into a database.

1 is a network connection configuration diagram of a caption providing apparatus according to an embodiment of the present invention.
2 is a detailed configuration diagram of a caption providing apparatus according to an embodiment of the present invention.
3 is an exemplary diagram of caption data including direct reading symbols according to the present invention.
4 is an exemplary diagram in which direct reading direct translation caption data is provided.
5 is an exemplary diagram in which Korean word order translated caption data is provided.
Fig. 6 is a block diagram of the translation service device of Fig. 1;

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. The configuration of the present invention and its effect will be clearly understood through the detailed description below. Note that prior to the detailed description of the present invention, the same components are denoted by the same reference numerals as possible even if they are displayed on different drawings, and a detailed description will be omitted when it is determined that the gist of the present invention may be obscure for known configurations. do.

1 is a network connection configuration diagram of a caption providing apparatus according to an embodiment of the present invention, and FIG. 2 is a detailed configuration diagram of a caption providing apparatus according to an embodiment of the present invention.

First, referring to FIG. 1, the subtitle providing service apparatus 100 according to an exemplary embodiment of the present invention basically performs a function of translating and providing English subtitles of a video into Korean in real time.

In particular, the caption providing service apparatus 100 according to an embodiment of the present invention interlocks between the broadcast operator and the user terminal 200 to stream digital broadcast data provided by the broadcast operator, while simultaneously displaying English subtitles of digital broadcast data in Korean. It is translated into and provided in real time. As such, the caption providing service apparatus 100 according to an embodiment of the present invention is characterized by providing a real-time translation technology.

The broadcaster plays a role of producing, editing, and changing broadcast content and providing it to the broadcast receiving unit 1. The broadcaster may include a program provider (PP), a terrestrial or content provider (CP), and the like.

Digital broadcast data provided by a broadcaster may include both existing broadcast data and various video data on the Internet. For example, it includes real-time broadcasting data, video on demand (VOD) data, cable broadcasting data, and the like.

In general, digital broadcast data is received through the broadcast receiving unit 1 and converted into a format suitable for the environment of the user terminal 200 through the streaming processing unit 2.

Then, it is transmitted to the user terminal 200 through the broadcast transmission unit 3 in a converted form.

In this case, the transmission method to the user terminal 200 may use one of broadcasting or unicasting, or a combination thereof.

Alternatively, the digital broadcast data may be pre-built in the video on demand (VOD) server 4 and then extracted and provided according to the request of the user terminal 200.

The streaming processing unit 2 will be described in detail in FIG. 2 below.

Based on this configuration, the caption providing service apparatus 100 according to an embodiment of the present invention includes a caption extraction unit 110, a caption analysis unit 120, a translation request unit 130, a translation result collection unit 140, and a translation caption. It can be implemented by including the management unit 150 and the translation subtitle transmission unit 160.

The caption extracting unit 110 extracts and decodes captions from digital broadcasting data streamed through the broadcast receiving unit 1. When the digital broadcasting data includes caption data, extraction thereof can be easily performed. That is, the digital broadcasting data received through the broadcast receiving unit 1 usually includes video data and audio data, and if there is textual caption data that is distinguished from these, it is collected and provided as a single file. Therefore, the extraction of subtitle data can be easily performed by a person skilled in the art.

As shown in FIG. 1, the caption extracting unit 110 may directly extract from digital broadcast data streamed through the broadcast receiving unit 1 at the rear end of the broadcast receiving unit 1, but through the broadcast receiving unit 1 The digital broadcasting data to be streamed may be extracted in connection with the moving picture separation unit 21 of the streaming processing unit 2 that separates the streaming digital broadcasting data into video, audio, and subtitles.

When extracting, the caption start time for synchronizing with the image data is also extracted from the caption data.

At this time, the caption start time is normally matched with the playing time of the video data when it is received together with the video data through the broadcast receiving unit 1, but a closed caption system is used to secure viewing rights for the hearing impaired. Accordingly, when the subtitle data is additionally included, a delay of several seconds may occur compared to the playback time of the image data.

The caption analysis unit 120 analyzes the caption data extracted by the caption extraction unit 110 into morpheme units such as words, characters, and symbols, and reconstructs the form of a sentence to request translation including a direct reading symbol.

Looking at this in more detail, the caption analysis unit 120 analyzes the morpheme based on the input caption data.

The caption analysis unit 120 analyzes a sentence component consisting of a combination of two or more of a subject, a predicate, an object, a bore, and other formula words to determine any one of the first to fifth formats, and the sentence component and Generate first result data in sentence format.

In addition, the subtitle analysis unit 120 receives the first result data from which the sentence components and sentence format of the sentence data are analyzed, and receives nouns, verbs, modifiers, infinitives, gerunds, participles, comparative grades, equal conjunctions, conjunctions, relative pronouns, and relations. Second result data for classifying sentence data for each part-of-speech is generated by performing sentence component analysis with any one of a plurality of parts-of-speech groups including adverbs or a combination thereof.

In addition, the caption analysis unit 120 receives the first result data and the second result data and displays a specific symbol for each part of speech in sentence data to generate caption data including the direct reading symbol. Phrases are underlined or displayed with original characters, triangles, arrows, etc., and symbols according to an embodiment of the present invention are merely an example, and such symbol indications may be variously adopted and displayed.

3 is an exemplary diagram of caption data including direct reading symbols according to the present invention.

Referring to the drawings, according to an embodiment of the present invention, the caption analysis unit 120 classifies each sentence component or part-of-speech and performs specific symbol processing. If this is described in detail, first, when the classified sentence component or part-of-speech is a verb Mark the word or phrase by underlining it.

In addition, when the modifier (preposition + noun) displays the "/" symbol on the word or the front side of the passage, in the case of the supine 'to' part, and the source to accommodate this "to" later, the underlined lower side Mark it.

' 기호를 표시한다.At this time, in the case of the injection phrase (V·ing noun (S), pp noun) , as described above, '

'Sign.

In addition, if the classified sentence component or part of speech is part of a part (V· ing noun, pp) , an arrow on the upper side in the noun direction of the front or rear side of the word or phrase modified by the corresponding word or phrase in the corresponding word V·ing, pp) or phrase. To mark.

In addition, if the classified sentence component or part of speech is comparative, the'△' sign that accepts it is indicated in the'as' or'than' part of the word or phrase, and if the classified sentence element or part of speech is a conjunction, the word or phrase A'△' or'○' sign that accepts this is indicated. If the conjunction is a co-ordinate conjunction, a'△' symbol is displayed, and if the conjunction is a subordinate conjunction, a'○' symbol is displayed.

'기호를 함께 표시한다.In addition, if the classified sentence component or part of speech is a relative pronoun or a relative adverb, mark the word or phrase with a'○' sign to accommodate it, and move forward with the relative pronoun or relative adverb.

'Mark together.

'기호를 표시한다.In addition, if it is determined that a conjunction, a relative pronoun, or a relative adverb is omitted through the first result data and the second result data, a'∨' symbol is displayed in the case of a conjunction in the omitted position of the sentence data, and a relative pronoun or a relative adverb If '

'Display the symbol.

Meanwhile, the translation request unit 130 pre-determines a device to be translated (hereinafter, a translation service device), and transmits the subtitle data including the direct reading symbol analyzed by the subtitle analysis unit 120 to the translation service device 300. And request a translation.

The translation service device 300 automatically translates the caption data including the direct reading direct mark requested through the translation request unit 130 into Korean. The translation service device 300 may be classified as a separate and independent device as a component linked to an external translation company, and may be included in the caption providing service device 100 of the present invention.

Meanwhile, the translation service apparatus 300 divides the sentence by extracting all possible split point candidates while reading the input sentence from start to finish. Here, the split point candidates used are punctuation marks, modifier phrases, and modifier clauses.

Here, the punctuation marks are period (.), question mark (?), exclamation mark (!), comma (,), middle dot (·), colon (: ), colon (;), hatched (/), double quotation mark ( “”), single quotation marks (''), parentheses (( )), curly braces, ({ }), brackets ([ ]), braces (-), dashes (--), tildes (~), ellipsis (.. ....) and so on.

At this time, a modifier or a modifier clause is marked with a'/' sign, a'△' sign, or a'○' sign.

However, the translation service device 300 excludes a noun or noun clause from the split point candidate if a noun or noun clause is followed in the case of a comma. In this case, the translation service apparatus 300 does not exclude from the dividing point candidate when the word following the comma is including serving as a preposition.

In addition, in the case of a modifier phrase, the translation service apparatus 300 excludes the word from the split point candidate when the preceding word is an injection phrase.

' 기호가 표시되어 있는 경우에 분할점 후보에서 제외한다. 다만 번역부 (3-1)는 '/' 기호의 뒤에 전치사를 앞에 둔 동명사가 위치하는 경우(/전치사+동명사)에는 분할점 후보에서 제외하지 않는다. That is, in the case of the'/' symbol, the translation service device 300 has a word in front of the'/'

If the 'symbol is marked, it is excluded from the candidate for splitting points. However, the translation department (3-1) does not exclude it from candidates for split points in cases where a gerund with a preposition in front of the'/' sign is located (/preposition + synonym).

In addition, the translation service device 300 adds virtual division points to the front and back of the verb to be divided.

That is, in the case of the verb, the translation service apparatus 300 adds the virtual splitting points before and after the underline is located in the case of the verb.

In addition, the translation service apparatus 300 excludes from the dividing point candidates in the case of a co-ordinate conjunction, and in the case of a non-verb, that is, a noun or the like.

However, the translation service device 300 adds virtual splitting points before and after the verb in the co-ordinate conjunctions when but is followed by a verb.

That is, the translation service device 300 is marked with a'△' sign in the case of a co-ordinate conjunction, and is excluded from the split point candidate when there is no underscore sign indicating a verb after the'△' sign, and when the co-ordinate conjunction is but If there is an underscore symbol, add virtual splitting points before and after.

When the modifier clause is a relational pronoun or a relational adverb, the translation service device 300 translates a limited portion defined by the relational pronoun or relational adverb by replacing the verb in the modifier clause.

'기호를 함께 표시되어 있는바, '○'기호에 '

'기호가 있는 경우에 한정 부분으로 수식어절의 동사를 대체하여 번역한다. 도 4와 이와 같이 직독직해 번역 자막 데이터가 제공되는 예시도이다.That is, when the modifier clause is a related pronoun or a related adverb, the translation service device 300 displays a'○' sign and moves forward with a related pronoun or a related adverb.

'The symbol is marked together, and the'○' symbol is marked with '

'If there is a sign, it is translated by replacing the verb in the modifier clause with a limited part. As shown in FIG. 4, it is an exemplary diagram in which direct-read translation caption data is provided.

On the other hand, the translation service device 300 not only provides directly read translated caption data, but also provides translated caption data suitable for Korean word order.

In this case, the translation service apparatus 300 may reverse the order of modifiers related to the subject.

For example, if the English sentence is "On a clear day, a crewmember on a merchant ship sailing across the caribbean Sea peers out at the horizon through his telescope" It is often translated as "to stare at the horizon through".

In order to prevent such erroneous translation, the translation service device 300 extracts nouns in front of the verb.

At this time, the translation service device 300 extracts nouns having a strong semantic element, and the translation service device 300 selects a flight attendant or a merchant ship, for example. As such, the translation service apparatus 300 performs the following operation when there are at least two nouns in front of the verb.

In addition, the translation service device 300 extracts sentences including the corresponding word from big data by inputting the crew member and the merchant ship from big data, and calculates the sentence ratio in which the merchant ship is expressed following the crew member from the extracted sentences, and vice versa. Following the merchant ship, the crew calculates the expressed sentence ratio, creates the Korean word order translation subtitle data arranged in the order of the largest sentence ratio, and provides it to the user through the translation result collection unit 140 as shown in FIG. do.

As an example, the translation service device 300 has a ratio of sentences in which a merchant ship is expressed following a crew member in the extracted sentences is 30%, and on the contrary, a sentence ratio calculated by calculating the sentence ratio expressed by a crew member following a merchant ship is 70%. The backside translation service device 300 generates Korean word order translation caption data translated as "on a clear day, a crew member gazes at the horizon through his telescope on a merchant ship crossing the Caribbean Sea."

That is, the translation service device 300 reads the semantic unit of the modifier after reading the semantic unit of the subject through this process, and the semantic unit of the verb is To be read So that it becomes a natural sentence.

Meanwhile, the translation service device 300 performs a verb post-processing process for correcting a translation error in relation to a verb.

To this end, the translation service device 300 extracts search nouns before and after the verb, and searches and extracts sentences including related search nouns from big data.

In addition, the translation service device 300 extracts a verb expression having a meaning similar to the meaning of the corresponding verb as a verb replacement candidate, and after calculating the sentence ratio for the extracted verb replacement candidates, the largest verb replacement with the largest sentence ratio Substitute the verb expression as a candidate.

As an example, translation services device (300) "There is a black flag up high in his mast." About "A black flag is flying high on its mast." Translation 1 car, and the "flag", "Mast" After extracting a plurality of representative nouns of the first translation, they are searched in big data and sentences including the plurality of representative nouns are collected.

At this time, the sentences that can be collected by the translation service device 300 are, for example, "The flags that have been torn by the wind and the torn flags dance sadly on the mast", "The flags are fluttering through the high-rise masts", "Broken At the end of the mast, there was a terrible torn flag," and so on, and the translation service device 300 extracts a verb expression having a meaning similar to that of the corresponding verb as a verb replacement candidate and calculates a sentence ratio.

As a result of calculating the sentence ratio, the translation service device 300 includes, for example, "Dance a sad dance", which is an alternative verb candidate extracted from "On the mast, the flags torn by the wind and the torn flags dance sad". The sentence ratio was 2%, and the alternative verb candidate "fluttered" extracted from the sentence "The flag flutters through the high-rise mast". The sentence ratio was 25%, "the end of the mast had a terribly torn flag hanging". If 8% of the expressions include "was stretched" in the expression, then "it depends " with the largest sentence proportion "flutters".

Alternatively, the translation service device (300) is the "A black flag is flying high on its mast."" There are black flag hangs high in their masts," as the primary translation, and "flag", "Mast" for After extracting the first translation into a plurality of search nouns, it is searched in big data, and sentences including the plurality of search nouns are collected, and each sentence is classified into first to n alternative candidate translations.

Thereafter, the translation service apparatus 300 extracts a plurality of reference representative words from the primary translation, but measures the similarity between the plurality of reference representative words and the plurality of first to n alternative representative words, and based on the result. By inferring the degree of similarity between the corresponding verb and the verb substitution candidates of 1 to n, the verb substitution candidate with the greatest similarity is substituted with the corresponding verb.

Measurement of the degree of similarity between the corresponding verb and the plurality of first to n replacement candidates may be performed by examining the presence or absence of a common word between the plurality of reference representative words and the plurality of first to n alternative representative words, It can be calculated according to Equation (1) below.

(Equation 1)

Here, n represents the number of a plurality of replacement representative words extracted from any one replacement candidate translation. In this case, when the i-th word among the replacement representative words is commonly present in the reference representative word and the substitute representative word, the Si value has a value of 1.

When the i-th word among the reference representative words does not exist in the substitute representative word, the Si value has a value of 0.

Meanwhile, the translation result aggregating unit 140 receives and aggregates the translated caption data (hereinafter, translated caption data-including direct translation caption data and Korean word order translated caption data) from the translation service device 300.

In this case, the translation result aggregator 140 may match the reproduction time of the translated caption data based on the audio signal of the digital broadcasting data. That is, the translated caption data is synchronized with the digital broadcasting data by matching and collecting the corresponding caption data in the section in which the audio signal of the digital broadcasting data is generated.

For synchronization, the translation result aggregating unit 140 checks the time for receiving the translation caption data and the time requested by the translation request unit 130 to calculate the time required for the translation process. Accordingly, when the time required for the translation process is delayed after a predetermined time elapses, it may be requested to delay the image data of the digital broadcast data by the elapsed time.

In addition, when the original caption data is delayed by a few seconds compared to the playback time of the video data in the digital broadcasting data (e.g., when the original caption data is closed caption data input in real time by personnel to protect the viewing rights of the hearing impaired), The translation result aggregator 140 may request to delay the reproduction time of the image data based on a reproduction time of the caption data (ie, original caption data) extracted from the digital broadcasting data and a delay time between the reproduction time of the image data. That is, the translation result aggregator 140 may request to delay the reproduction time of the image data by the delay time based on a value obtained by calculating a delay time between the reproduction time of the original caption data and the reproduction time of the image data.

The translation caption management unit 150 manages data such as storing the original caption data for digital broadcasting data and the translated caption data collected through the translation result collection unit 140 in a database and reconfiguring it into a file format suitable for later use. .

The translation caption transmission unit 160 receives input of a language to be translated from the user terminal 200, and transmits the translated caption data collected through the translation result collection unit 140 in synchronization with streaming digital broadcasting data.

In particular, when a delay request for image data of digital broadcasting data is generated by the translation result collecting unit 140, the translated caption transmitting unit 160 delays and transmits the image data of the digital broadcasting data by a delay time.

Next, referring to FIG. 2, the streaming processing unit 2 may include a configuration for separating digital broadcasting data streamed through the broadcast receiving unit (1 of FIG. 1) into video and audio, processing the data, and then synchronizing it.

As a configuration for this, the streaming processing unit 2 includes a moving picture separation unit 21, a voice signal characteristic analysis unit 22, a voice start time recording unit 23, and an integration unit 24.

The moving picture separating unit 21 separates video data and audio data from digital broadcast data streamed through the broadcast receiving unit (1 in FIG. 1). Specifically, the moving picture separating unit 21 may include an image extracting unit 21a for extracting image data from digital broadcast data streamed through a broadcast receiving unit, and an audio extracting unit 21b for extracting audio data.

The voice signal feature analysis unit 22 analyzes the voice data extracted through the voice extracting unit 21b to detect a voice start time in the video. The audio start time is the basis for determining the subtitle start time when synchronizing the subtitle data.

Voice data can be analyzed using frequency, peak, edge detection, and noise analysis of the voice signal.

The audio start time recording unit 23 records times in the video at the points where the audio start time is determined based on the information analyzed by the audio signal characteristic analysis unit 22.

The integration unit 24 combines the image data and audio data extracted through the image extraction unit 21a and the audio extraction unit 21b to form one broadcast data. At this time, the voice data is integrated in a form including the voice start time through the voice start time recording unit 23.

In addition, the integrator 24 may receive the translated caption data translated in real time by the caption providing service device (100 of FIG. 1) described above and integrate it together with the image data.

6 is a diagram showing the configuration of a translation service apparatus according to the present invention.

6, the translation service apparatus according to the present invention includes a sentence splitter 10, a direct reading translator 12, a Korean word order translator 14, a word order sorter 16, and a post-processing unit 18. .

First, the sentence divider 10 extracts all possible split point candidates while reading the input sentence from start to finish. Here, the split point candidates used are punctuation marks, modifier phrases, and modifier clauses.

Here, the punctuation marks are period (.), question mark (?), exclamation mark (!), comma (,), middle dot (·), colon (: ), colon (;), hatched (/), double quotation mark ( “”), single quotation marks (''), parentheses (( )), curly braces, ({ }), brackets ([ ]), braces (-), dashes (--), tildes (~), ellipsis (.. ....) and so on.

At this time, a modifier or a modifier clause is marked with a'/' sign, a'△' sign, or a'○' sign.

However, the sentence divider 10 excludes a noun or noun clause from a candidate for a split point if a noun or a noun clause is followed by a comma. In this case, the sentence divider 10 does not exclude from the dividing point candidates when the word following the comma is including serving as a preposition.

In addition, the sentence divider 10 excludes a modifier phrase from the split point candidate when the preceding word is an injection phrase.

' 기호가 표시되어 있는 경우에 분할점 후보에서 제외한다.That is, in the case of the'/' symbol, the sentence divider 10 has a word in front of the'/'

If the 'symbol is marked, it is excluded from the candidate for splitting points.

In addition, the sentence divider 10 adds virtual splitting points to the front and back in the case of verbs to distinguish them.

That is, the sentence divider 10 adds an underline in the case of a verb, and adds virtual splitting points before and after the underline.

In addition, the sentence divider 10 excludes from the dividing point candidates in the case of a co-ordinate conjunction, and in the case of a non-verb, that is, a noun or the like.

However, the sentence divider 10 adds virtual splitting points before and after when the verb is after but in the equal conjunction.

That is, the sentence divider 10 is marked with a'△' sign in the case of a co-ordinate conjunction, and is excluded from the split point candidate when there is no underscore sign indicating a verb after the'△' sign. If there is an underscore symbol, an imaginary split point is added before and after.

Next, the direct reading direct translation translator 12 receives the divided English sentences and translates them according to the sentence division order to generate the translated subtitle data.

From here, Direct reading Translator 12 Direct reading Sign Marked In English sentences Direct reading The meaning unit operation of the symbol is performed, and based on this, the meaning unit Direct reading Carry out.

The Korean word order translator 14 receives the direct read direct translation caption data translated by the direct read direct translation translator 12 and adjusts the order according to the Korean word order to generate the Korean word order translated caption data.

At this time, the direct-reading translator 12 translates the qualifier clause by replacing the limited part defined by the relation pronoun or the relation adverb with the verb of the qualifier clause when the modifier clause is a relation pronoun or a relation adverb.

'기호를 함께 표시되어 있는바, 직독직해 번역기(12)는 '○'기호에 '

'기호가 있는 경우에 한정 부분으로 수식어절의 동사를 대체하여 번역한다. That is, when the modifier clause is a related pronoun or a related adverb, the sentence divider 10 displays a'○' sign and moves forward with a related pronoun or a related adverb.

'As the symbol is marked together, the direct reading translator (12) means'○' in the'○' symbol.

'If there is a sign, it is translated by replacing the verb in the modifier clause with a limited part.

Meanwhile, the Korean word order translator 14 may reverse the order of modifiers related to a subject.

For example, if the English sentence is "On a clear day, a crewmember on a merchant ship sailing across the caribbean Sea peers out at the horizon through his telescope" It is often translated as "to stare at the horizon through".

In order to prevent such erroneous translation, the word order sorter 16 extracts nouns in front of the verb.

At this time, the word order sorter 16 extracts nouns having strong semantic elements, and the word order sorter 16 selects a crew member and a merchant ship as an example. As such, the word order sorter 16 performs the following operation when there are at least two nouns in front of the verb.

In addition, the word order sorter 16 extracts sentences containing the word from big data by inputting the crew and the merchant ship from big data, and calculates the proportion of sentences in which the merchant ship is expressed following the crew from the extracted sentences, and vice versa. Following the merchant ship, the flight attendant calculates the expressed sentence ratio, and creates and provides the Korean word order translated subtitle data arranged in the order of the largest sentence ratio to the user.

In the word order sorter 16, the proportion of sentences in which the crew and the merchant ship are expressed in the extracted sentences as an example is 30%, and on the contrary, the sentence ratio calculated by calculating the proportion of the sentences expressed by the crew after the merchant ship is 70%. The word order sorter 16 on the back generates Korean word order translated subtitle data translated as "On a clear day, on a merchant ship crossing the Caribbean Sea, the crew stares at the horizon through his telescope."

On the other hand, the post-processing unit 18 performs a verb post-processing process to correct a translation error in relation to the verb.

To this end, the post-processing unit 18 extracts search nouns before and after the verb, and searches and extracts sentences including related search nouns from big data.

In addition, the post-processing unit 18 extracts a verb expression with a meaning similar to the meaning of the corresponding verb as a verb substitution candidate, calculates the sentence ratio for the extracted verb substitution candidates, and then substitutes the largest verb with the largest sentence ratio. Replace the verb expression as a candidate

As an example, the post-processing suhaenggi 18 "has a black flag up high in the mast," with respect to "A black flag is flying high on its mast." Translated first in, and "flag", a "mast" After extracting a plurality of representative nouns of the first translation, they are searched in big data and sentences including the plurality of representative nouns are collected.

At this time, the sentences that can be collected by the post-processing implementer (18) are, for example, "The flags that are torn by the wind and the torn flags dance sadly on the mast", "The flags are fluttering through the high-rise masts", "Broken At the end of the mast, there was a terrible torn flag," and so on, and the post-processing unit 18 extracts the verb expression with a meaning similar to that of the corresponding verb as a verb replacement candidate and calculates the sentence ratio.

As a result of calculating the sentence ratio, the post-processing unit 18 includes, for example, "Dance a sad dance", which is an alternative verb candidate extracted from "On the mast, the flags torn by the wind and the torn flags dance sad". The sentence ratio was 2%, and the alternative verb candidate "fluttered" extracted from the sentence "The flag flutters through the high-rise mast". The sentence ratio was 25%, "the end of the mast had a terribly torn flag hanging". If 8% of the expressions include "was stretched" in the expression, then "it depends " with the largest sentence proportion "flutters".

Alternatively, the processing suhaenggi 18 is an "A black flag is flying high on its mast.""There is a black flag up high in the mast," as the primary translation, and the "flag", "poles" for after After extracting the first translation into a plurality of search nouns, it is searched in big data, and sentences including the plurality of search nouns are collected, and each sentence is classified into first to n alternative candidate translations.

Thereafter, the post-processing unit 18 extracts a plurality of reference representative words from the primary translation, but measures the similarity between the plurality of reference representative words and the plurality of first to n alternative representative words, and based on the result. By inferring the degree of similarity between the corresponding verb and the verb substitution candidates of 1 to n, the verb substitution candidate with the greatest similarity is substituted with the corresponding verb.

Measurement of the degree of similarity between the corresponding verb and the plurality of first to n replacement candidates may be performed by examining the presence or absence of a common word between the plurality of reference representative words and the plurality of first to n alternative representative words, It can be calculated according to Equation (1) below.

(Equation 1)

Here, n represents the number of a plurality of replacement representative words extracted from any one replacement candidate translation. In this case, when the i-th word among the replacement representative words is commonly present in the reference representative word and the substitute representative word, the Si value has a value of 1.

When the i-th word among the reference representative words does not exist in the substitute representative word, the Si value has a value of 0.

The above description is merely illustrative of the present invention, and various modifications may be made without departing from the technical spirit of the present invention by those of ordinary skill in the technical field to which the present invention pertains. Therefore, the embodiments disclosed in the specification of the present invention do not limit the present invention. The scope of the present invention should be interpreted by the following claims, and all technologies within the scope equivalent thereto should be interpreted as being included in the scope of the present invention.

1: broadcast receiving unit 2: data processing unit
3: Broadcasting unit 4: VOD server
100: subtitle providing service device
110: subtitle extraction unit 120: subtitle analysis unit
130: translation request unit 140: translation result collection unit
150: translation subtitle management unit 160: translation subtitle transmission unit
200: user terminal 300: translation service device

Claims (15)

A caption extractor for extracting and decoding captions from digital broadcasting data to be streamed at the request of a user terminal;
A caption analysis unit for analyzing caption data extracted by the caption extracting unit in units of morphemes and reconstructing caption data including direct reading symbols;
A translation request unit for requesting a translation of the caption data reconstructed by the caption analysis unit to a predetermined translation service device;
A translation result aggregator for receiving and collecting the translated subtitle data translated by the translation request; And
A translation caption transmission unit receiving input from the user terminal for a language to be translated and transmitting the translated direct reading translation caption data collected through the translation result collecting unit in synchronization with the digital broadcasting data,
The translation service apparatus includes: a sentence divider for dividing a sentence by extracting all possible segmentation point candidates while reading an input sentence from start to finish;
A direct-reading translator for receiving the divided English sentences from the sentence divider and translating them according to the order in which the sentences were divided to generate direct-reading translated caption data;
A Korean word order translator for generating subtitle data translated in Korean word order by receiving the direct reading direct translation subtitle data translated by the direct reading direct reading translator and adjusting the order according to the Korean word order:
A word order sorter that arranges positions of modifiers related to a subject in the Korean word order translated subtitle data,
The split point candidates are punctuation marks, modifiers, and modifier clauses, and the word order sorter extracts nouns in front of the verb, extracts sentences including the corresponding word from big data, and calculates the sentence ratio for each noun order from the extracted sentences. And creating a Korean word order translation arranged according to the calculated sentence ratio. The method of claim 1,
The caption analysis unit analyzes the morpheme based on the input caption data, displays a specific symbol for each part of speech in the caption data, and generates caption data including the symbol by direct reading, a digital broadcasting caption providing service device through real-time translation. The method of claim 2,
The subtitle analysis unit analyzes a sentence component consisting of a combination of two or more of a subject, a predicate, an object, a bore, and other modal words to determine any one of the first to fifth forms, and the sentence component and sentence form accordingly Generates the first result data, and receives the first result data from which the sentence components and sentence format of the subtitle data are analyzed, and receives the nouns, verbs, modifiers, infinitives, gerunds, participles, comparative grades, equal conjunctions, conjunctions, relative pronouns, and relative adverbs Generates second result data for classifying subtitle data by part of speech by performing sentence component analysis with any one or a combination of a plurality of parts of speech groups including, and receiving the first result data and the second result data A service device for providing captions for digital broadcasting through real-time translation that displays specific symbols for each part-of-speech in the caption data and generates caption data including the symbols by direct reading. The method of claim 3,
A service device for providing captions for digital broadcasting through real-time translation in which the specific symbol generated by the caption analysis unit is displayed by underlined in a corresponding word or phrase, or displayed as an original character, triangle, or arrow. The method of claim 1,
The translation service device receives caption data including the direct reading symbol generated by the caption analysis unit, performs direct reading direct translation, generates the translated caption data directly read, and provides it to the translation result aggregator through real-time translation. Subtitle provision service device. The method of claim 5,
The translation service device receives the caption data including the direct reading symbol generated by the caption analysis unit and performs Korean word order translation to generate Korean word order translated caption data. The method of claim 6,
The translation result collection unit receives and aggregates the translated subtitle data in Korean word order translated by the translation request,
The translation caption transmission unit receives the input of the language to be translated from the user terminal and transmits the translated Korean translated caption data collected through the translation result collecting unit in synchronization with the digital broadcasting data. Device. delete delete delete delete The method of claim 7,
A service apparatus for providing subtitles for digital broadcasting through real-time translation, further comprising: a post-processing unit performing a verb post-processing process for correcting a translation error in relation to a verb in the Korean word order translation subtitle data by the Korean word order translator. The method of claim 12,
The post-processor extracts a plurality of search nouns, searches for and extracts sentences including related search nouns from big data, extracts a verb expression with a meaning similar to the meaning of the corresponding verb as a verb substitution candidate, and extracts verb substitution candidates. A service device providing subtitles for digital broadcasting through real-time translation that substitutes the corresponding verb expression as the largest verb substitution candidate with the largest sentence ratio after calculating the sentence ratio for. The method of claim 12,
The post-processing unit extracts a plurality of search nouns of the primary translation, searches it in big data, collects sentences including the plurality of search nouns, and classifies each sentence into first to n alternative candidate translations, Extracting a plurality of reference representative words from the primary translation, measuring the similarity between the plurality of reference representative words and a plurality of first to n alternative representative words, and replacing the corresponding verb with the first to n verbs based on the results A service device providing subtitles for digital broadcasting through real-time translation that infers the degree of similarity between candidates and replaces the candidate with the largest similarity with the corresponding verb. The method of claim 1,
The translation result collection unit
A service apparatus for providing closed captions for digital broadcasting through real-time translation, characterized in that matching the reproduction time of the translated direct reading direct translation caption data based on the audio signal of the digital broadcasting data.

Images