KIM, Yonghwa (104-901, Hak-Maeul Seo-hae Apt. 432-5,Byungbang-dong, Kyeyang-ku, Incheon-shi 407-320, KR)
Claims
[1] A system, which generates the indexing information of multimedia data files using vocal data, the system includes: a vocal data extraction unit which determines whether the multimedia data file, which is registered via web by a multimedia data file provider, includes a vocal data, and if the vocal data is included in the multimedia data file, extracts the vocal data; a data conversion unit which converts the extracted vocal data into a text data; an initial data generation unit which generates an initial data by combining the text data and a time-information data of sentence structure corresponding to the text data; an identification code grant unit which selects a set of data from the initial data, determines whether selection of the set of data is complete, and if selection of the set of data is complete, grants identification code to the combination of time- information data of the selected set of data and the text data of vocal data; a record unit which records the combination of time-information data of selected set of data and the text data of vocal data and the identification code on a metadata of multimedia data file; and a storage unit which stores the multimedia data file in a database.
[2] The system according to claim 1, wherein the initial data includes start time, end time and vocal text data during corresponding time.
[3] The system according to claim 1, wherein the metadata includes a time- information of an arbitrary part of the corresponding file, a vocal text information of the part, a language information of corresponding text and an identification code of the combination of thereof.
[4] A method of generating indexing information of multimedia data file using a vocal data, the method includes: allowing a vocal data extraction unit to determine whether the vocal data is included in the multimedia data file, which was registered by a multimedia data file user via web, and if vocal data is included in the multimedia data file, to extract the vocal data; allowing a data conversion unit to convert the extracted vocal data into a text data; allowing an initial data generation unit to generate initial data by combining text data and time-information data of sentence structure of corresponding text data; allowing an identification code grant unit to select a set of data from the initial data, to determine whether selection of the set of data is completed, and if selection of the set of data is completed, to grant an identification code to the time-information data of selected set of data and text data in vocal form; allowing a record unit to record the combination of time-information data of a set of data and text data of vocal form and identification code in the metadata of the multimedia data file; and allowing a storage unit to store the multimedia data file into a database.
[5] A system, which searches indexing information of multimedia data files, the system includes: an indexing information list reception unit which receives indexing information list of a set of data of corresponding multimedia data file from the multimedia data file; a search and indexing sentence structure selection unit which determines whether a specific part of a video was searched through the indexing information list of the multimedia data file, and selects the indexing sentence structure; an identification code transmission unit which transmits the identification code of the indexing sentence structure to a web server; a time-information transmission unit which transmits time-information data of the multimedia data file, which is indicated by the corresponding identification code of a user's terminal using the multimedia data file; and a video play unit which receives time-information data of the user's terminal using the multimedia data file, and plays the corresponding part of the video.
[6] The system according to claim 5, wherein the search and indexing selection unit includes: a search recognition unit which recognize whether a specific part of the video was searched through the indexing information list of the multimedia data file, and if a specific part of the video was not searched through the indexing information list from above, determines whether a specific part of the video was searched through sentence structures; an indexing sentence structure selection unit which selects the indexing sentence structure to be moved from the indexing information list; a movement unit which moves the indexing information to a sentence structure search interface unit if a specific part of the video was searched through the sentence structure; and a search unit which executes search through the sentence structure search interface unit, and outputs the search results in indexing information list form.
[7] A method of searching indexing information of multimedia data files, the method includes: first step of allowing an indexing information list reception unit to receive an indexing information list of a set of data of the corresponding multimedia data file from the multimedia data file; second step of allowing a search and indexing sentence structure selection unit to determine whether a specific part of the video was searched through the indexing information list of multimedia data filee, and to select the indexing sentence structure; third step of allowing an identification code transmission unit to transmit the identification code of the indexing sentence structure to a web server; forth step of allowing a time-information transmission unit to transmit a time- information data of multimedia data file indicated by the corresponding identification code to a user's terminal using the multimedia data file and a system server; and fifth step of allowing a video play unit to receive the time-information data from the user's terminal using the multimedia data file, and to play the corresponding part of the video.
[8] The method according to claim 7, the second step includes: allowing the search and indexing sentence structure selection unit to select the indexing sentence structure to be moved, from the indexing information list, if a specific part of the video was searched through the indexing information list of multimedia data file.
[9] The method according to Claim 7, the second step includes: allowing a search recognition unit to determine whether a specific part of the video was searched through sentence structures, if a specific part of the video was not searched through the indexing information list of multimedia data file; allowing a movement unit to move the indexing information to the sentence structure search interface method if a specific part of the video was searched through the sentence structure; allowing a search unit to execute the search through the sentence structure search interface method and to output the search results in indexing information list form; and allowing an indexing sentence structure selection unit to select the indexing sentence structure to be moved, from the outputted indexing information list. |
Description
THE SYSTEM AND METHOD FOR GENERATING INDEXING
INFORMATION OF MULTIMEDIA DATA FILE USING VOCAL
DATA AND RETRIEVING INDEXING INFORMATION OF
MULTIMEDIA DATA FILE Technical Field
[1] The present invention is related to the system and method of generating indexing information of multimedia data file using vocal data and the system and method of searching indexing information of multimedia data file. Especially, the present invention is related to generating indexing information of multimedia data file using vocal data, and by using the generated indexing information, allowing the multimedia data file users to search wanted contents. Background Art
[2] Until now, along with the increase in demand for multimedia data, a variety of media players have appeared and streaming service has quickly expanded. In addition to real-time live videos, VOD(Video on Demand) users have also increased. Since most VOD and multimedia data files are mostly constituted of video and audio, users had to rely on multimedia data file s title or limited information provided by the webpage to know the information of multimedia data file. Even if the limited information is not provided, the user had to watch the multimedia data file.
[3] Additionally, to know the overall content of the multimedia data file, the user has to watch the entire corresponding multimedia data file. It is difficult under current circumstance to watch only the important parts. If the multimedia data file user selects random parts to watch, it is difficult to exactly understand the content of multimedia data file. Disclosure of Invention
Technical Problem
[4] To solve the problem above, the present invention aims to provide a system and method which allows users to quickly search only necessary information in the multimedia data file by generating indexing information, with vocal data as its basis, on important parts of the multimedia data file. Technical Solution
[5] The present invention presents a system, which generates the indexing information of multimedia data files using vocal data, the system includes: a vocal data extraction unit which determines whether the multimedia data file, which is registered via web by
a multimedia data file provider, includes a vocal data, and if the vocal data is included in the multimedia data file, extracts the vocal data; a data conversion unit which converts the extracted vocal data into a text data; an initial data generation unit which generates an initial data by combining the text data and a time-information data of sentence structure corresponding to the text data; an identification code grant unit which selects a set of data from the initial data, determines whether selection of the set of data is complete, and if selection of the set of data is complete, grants identification code to the combination of time-information data of the selected set of data and the text data of vocal data; a record unit which records the combination of time-information data of selected set of data and the text data of vocal data and the identification code on a metadata of multimedia data file; and a storage unit which stores the multimedia data file in a database.
[6] Additionally, the present invention presents A method of generating indexing information of multimedia data file using a vocal data, the method includes: allowing a vocal data extraction unit to determine whether the vocal data is included in the multimedia data file, which was registered by a multimedia data file user via web, and if vocal data is included in the multimedia data file, to extract the vocal data; allowing a data conversion unit to convert the extracted vocal data into a text data; allowing an initial data generation unit to generate initial data by combining text data and time- information data of sentence structure of corresponding text data; allowing an identification code grant unit to select a set of data from the initial data, to determine whether selection of the set of data is completed, and if selection of the set of data is completed, to grant an identification code to the time-information data of selected set of data and text data in vocal form; allowing a record unit to record the combination of time-information data of a set of data and text data of vocal form and identification code in the metadata of the multimedia data file; and allowing a storage unit to store the multimedia data file into a database.
[7] Additionally, the present invention presents A system, which searches indexing information of multimedia data files, the system includes: an indexing information list reception unit which receives indexing information list of a set of data of corresponding multimedia data file from the multimedia data file; a search and indexing sentence structure selection unit which determines whether a specific part of a video was searched through the indexing information list of the multimedia data file, and selects the indexing sentence structure; an identification code transmission unit which transmits the identification code of the indexing sentence structure to a web server; a time-information transmission unit which transmits time-information data of the multimedia data file, which is indicated by the corresponding identification code of a user's terminal using the multimedia data file; and a video play unit which receives
time-information data of the user's terminal using the multimedia data file, and plays the corresponding part of the video.
[8] Additionally, the present invention presents A method of searching indexing information of multimedia data files, the method includes: first step of allowing an indexing information list reception unit to receive an indexing information list of a set of data of the corresponding multimedia data file from the multimedia data file; second step of allowing a search and indexing sentence structure selection unit to determine whether a specific part of the video was searched through the indexing information list of multimedia data filee, and to select the indexing sentence structure; third step of allowing an identification code transmission unit to transmit the identification code of the indexing sentence structure to a web server; forth step of allowing a time- information transmission unit to transmit a time-information data of multimedia data file indicated by the corresponding identification code to a user's terminal using the multimedia data file and a system server; and fifth step of allowing a video play unit to receive the time-information data from the user's terminal using the multimedia data file, and to play the corresponding part of the video.
[9]
Advantageous Effects
[10] The present invention makes search of existing multimedia data files via words or sentence structures (including keyword search) possible without the processing by the multimedia data file provider, because a separate vocal extraction device extracts vocal text and time-information data, and automatically creates, registers and stores file indexing information. Therefore, it is possible for the user to quickly understand important contents without playing the corresponding multimedia data file during searching information. Moreover, this increases the efficiency of information search because the user does not have to read unneeded information.
[11] Additionally, the user can search a specific word or phrase among the contents of a multimedia data file and quickly move to the corresponding part. Therefore, an efficient and dynamic information search is possible because the user can quickly understand the contents of the multimedia data file without having to watch the whole file.
Brief Description of the Drawings
[12] FIG 1 is a composition diagram of the present invention showing the system, which generates indexing information of multimedia data file using vocal data.
[13] FIG 2 is a drawing of the present invention showing the method of extracting vocal data from multimedia data file and converting it into text data.
[14] FIG 3 is a drawing of the present invention showing the method of extracting initial
data by matching vocal text data extracted form multimedia data file and synchronization of time-information data. [15] FIG 4 to FIG 6 is drawings of the present invention showing the method of granting identification code to a set of data. [16] FIG 7 is a drawing of the present invention showing the process of adding the combination of time-information data and vocal text data, and the identification code granted, to metadata of multimedia data file. [17] FIG 8 is a composition diagram of the present invention showing the system of searching indexing information of multimedia data file using vocal data. [18] FIG 9 is a composition diagram of the search and indexing sentence structure selection unit of FIG 8. [19] FIG 10 is a drawing of the present invention showing the action process among the user s terminal of multimedia data file, web server and system server. [20] FIG 11 is a drawing of the present invention showing the user interface method of search process through indexing information list among multimedia data file search processes. [21] FIG 12 is a drawing of the present invention showing the user interface method of search process through word or phrase search among multimedia data file search processes. [22] FIG 13 is a flow chart showing the method of generating indexing information of multimedia data file. [23] FIG 14 is a flow chart showing the method of searching indexing information of multimedia data file.
Best Mode for Carrying Out the Invention [24] The following is a detailed explanation of present invention's execution example through drawing. [25] FIG 1 is a composition diagram of the present invention showing a system, which generates indexing information of multimedia data file using vocal data. As illustrated in FIG 1, the system(lOO), which generates indexing information of multimedia data file using vocal data includes vocal data extraction unit(l 10), data conversion unit(120), initial data generation unit(130), identification code grant unit(140), record unit(150) and storage unit(160). [26] The vocal data extraction unit(l 10) concludes whether vocal data is included in multimedia data file registered on web by the multimedia data file provider, and if vocal data is included in multimedia data file, extracts vocal data. [27] The data conversion unit(120) converts extracted vocal data into text data.
[28] The initial data generation unit(130) generates initial data by combining text data
and time-information data of corresponding text data sentence structure.
[29] The identification code grant unit(140) selects a set of data from initial data, concludes whether selection of a set of data is completed, and if selection of a set of data is completed, grants identification code to the combination of time-information data of selected set of data and text data of vocal data. A set of data is data the user selected on user's terminal.
[30] The record unit(150) records combination of time-information data of selected set of data and text data of vocal data, and identification code on metadata of multimedia data file.
[31] The storage unit(160) stores multimedia data file in database.
[32] FIG 2 is a drawing of the present invention showing the method of extracting vocal data from multimedia data file, and converting it into text data. As illustrated in FIG 2, the vocal data extraction unit(l 10) concludes whether vocal data(104) is included in multimedia data file(102), and if vocal data(104) is included in multimedia data file(102), extracts vocal data(104) after separating it with time-information data(105). Also, the data conversion unit(120) converts extracted vocal data(104) into vocal text data(106)
[33] FIG 3 is a drawing of the present invention showing the method extracting initial data by matching vocal text data extracted from multimedia data file and synchronization of time-information data. As illustrated in FIG 3, in case of the present invention, initial data is generated by combining vocal text data extracted and converted from multimedia data file registered by multimedia data file provider, and time-information data of corresponding vocal text data. The corresponding process is consisted of single sentence unit(201), The expression form of time-information data includes start time(202), end time(204), and vocal text data(206) during corresponding time. Time-information data(202, 203, 204, 205) includes time data(202, 204) expressed in hour, minute and second, and SCR(System Clock Reference) information or OCR(Optical Character Reader) information (203, 205). For example, in case of the corresponding sentence(206) "DDDD OOO DDDDDD.", the time-information data is from start time Start = 00:02:50,00009001(202,203)toendtimeEnd=00:02:55,00013500(204 ,205).
[34] FIG 4 to FIG 6 is a drawing of the present invention showing the method of granting identification code to a set of data. In each language, smallest units such as basic phoneme, syllable, alphabet and radicals(basic unit of Chinese characters) exist, which are called basic units. In Korean phonemes such as "D D D D ... D D D", in English alphabets such as "a b c d e ... x y z" and in Chinese radicals such as " — ' D ... D D" are considered as basic units. Likewise, other languages consider its smallest unit as basic unit using the same method as above.
[35] Additionally, arrange basic units of each language in order, and assign numbers in
order. The assigned numbers should decimal numbers between 1-255, with a size of 8 bits. For example, as in FIG 4, the 24 basic units (phoneme) of Korean, "D D D D ... D D D" are granted decimal numbers 1-24(301), the 26 basic units(alphabet) of English, "a b c d e ... x y z" are granted decimal numbers 1-26(302), and the 214 basic units (radicals) of Chinese, " — ' D ... D D" are granted decimal numbers 1-214 in order(303).
[36] Additionally, break units(spacing between words or syllables) exist in a selected set of data sentence. Among the units(spacing between words or syllables) which consist a sentence, for the first 12, extract the first basic unit(phoneme, syllable, alphabet, or radicals etc.) of each letter of each unit. For example, as in FIG 5, from Korean sentence "DDD DDDD DDD DDDD DD DD DD DD DDD DDDD DDDDD.", extract the first basic unit(phoneme), "D", "D "D", "D", "D", "D", "D", "D", "D", "D", of the first 12 words(311).
[37] Create a 96 bit(12 byte) partial identification code by combining number line corresponding to the first basic units of the first 12 words. For example, as in FIG 5, for Korean partial identification code, combine number "3" corresponding to "D", number "9" corresponding to "D", number "1" corresponding to "D", number "7" corresponding to "D", number "8" corresponding to "D", number "8" corresponding to "D", number "8" corresponding to "D", number "6" corresponding "D", and number "10" corresponding to "D"(rest omitted)(312). If the unit(word or syllable) of a sentence is less than 12, fill in number "0" for the empty slots. Thus, by combining the first 12 phoneme, 96 bit partial identification code, "3.9.1.7.8.8.8.6.10.7.8.0" can be made(313).
[38] Likewise, partial identification code can be granted and searched for English sentences using the same method of extracting basic units(alphabet) of the first letters of the first 12 breaking units(words). For example, in case of the sentence "A network of people or institutions is a large number of them that have a connection with each other and work together as a system.", extract first syllables "a", "n", "o", "p", "o", "i", "i", "a", "1", "n", "o" and "t". A 96 bit partial identification code, " 1.14.15.16.15.9.9.1.12.124.15.20" can be made by combining corresponding numbers of the 12 alphabets.
[39] Partial identification code of a Chinese sentence is generated using the same method. For example, in the case of sentence, "^fRD W it J# i^ffMH^Hffi", the radicals of first syllable^), son "ja"(^F) is the 39 radicals, thus number "39" is granted. Likewise, the radicals of "^R" is mouth "gu"( P ), which is the 30 radicals, thus number"30" is granted(restomitted). Combining all these numbers will generate a 96bit partial identification code,"39.30 0.0."
[40] To prevent overlap with pre-existing identification codes, a 32 bit(4 byte) index is attached to the 96 bit partial identification code generated using the same method as above. The number representing the index is a 8 bit decimal number, assigned between 0-255, and automatically increases by 1 when the next index is generated. For
example, as in FIG 6, if 26 partial identifications code already exists when partial identification code "3.9.1.7.8.8.8.6.10.7.8.0" is generated(321), index "0.0.0.26" is attached(322), and index "0.0.0.27" is attached to the 28 identification code due to the automatic increase of 1.
[41] The 128 bit( 16 byte) overall identification code is generated by combining the 96 bit partial identification code and 32 bit index generated using the method above. For example, as in (c) of FIG 5, from "DDD DDDD DDD DDDD DD DD DD DD DDD DDDD DDDDD", 12 basic units can bt extracted to generate partial identification code "3.9.1.7.8.8.8.6.10.7.8.0", and if 26 of the corresponding identification code is generated, index representing 27 ,"0.0.0.26", is attached to create identification code,"3.9.1.7.8.8.8.6.10.7.8.0.0.0.0.26", of the corresponding sentence.
[42] FIG 7 is a drawing of the present invention showing the process of adding the combination of time-information data, vocal data and identification code granted, to metadata of multimedia data file. As illustrated in FIG 7, multimedia data file(401) is consisted of time-information of corresponding file's set unit(406), vocal text information of that unit(407), language information(405) of corresponding text, header unit(402) including metadata(404, 405, 406, 407), where identification code(404) of language information is recorded and data unit(403), where actual data is stored.
[43] When the multimedia data file provider selects one or more set parts of a multimedia data file, the writing tool of identification code generation unit executes the process of granting identification code to the combination of time-information data(406), vocal text information(407) and language information of corresponding text(405), of the corresponding part. If the multimedia data file provider selects many set parts, a large number of data combinations exist. At this point, the writing tool stores the combination of one or more identification codes and data in the header unit(402)'s metadata(404, 405, 406, 407) of multimedia data file(401).
[44] FIG 8 is a composition diagram of the present invention showing the system which searches indexing information of multimedia data file using vocal data, and FIG 9 is a composition diagram of search and indexing sentence selection unit of FIG 8. As illustrated in FIG 8 and FIG 9, the system(500) which searches indexing information of multimedia data file using vocal data includes indexing information list reception unit(510), search and indexing sentence structure selection unit(520), identification code transmission unit(530), time-information transmission unit(540) and video play unit(550). Also, the search and indexing sentence structure selection unit(520) includes search recognition unit(521), movement unit(522), search unit(523) and indexing sentence structure selection unit(524).
[45] The indexing information list reception unit(510) receives indexing information list of corresponding multimedia data file's set part from the multimedia data file.
[46] The search and indexing sentence structure selection unit(520) concludes whether video at a certain part was searched, through indexing information list of multimedia data file, and selects indexing sentence structure. Search and indexing sentence structure selection unit(520) includes search recognition unit(521), movement unit(522), search unit(523), search result output unit(524) and indexing sentence structure selection unit(525). The search recognition unit(521) recognizes whether a certain part of a video was searched through indexing information list of multimedia data file, and if a certain part of a video was not searched through indexing information list, it concludes whether a certain part of a video was searched through sentence structures. Then, the movement unit(522) moves it to sentence structure search interface method of indexing information, if a certain part of a video was searched through sentence structures. Also, the search unit(523) executes the search through sentence structure search interface method, and outputs search results in an indexing information list form. Moreover, the indexing sentence structure selection unit(524) selects indexing sentence structures, which it is attempting to move to, from the indexing information list.
[47] The identification code transmission unit(530) transmits identification code of indexing sentence structures to the web server.
[48] The time-information transmission unit(540) transmits time-information data of multimedia data file, which the corresponding identification code is indicating, to the streaming server and user's terminal.
[49] The video play unit(550) receives time-information data from user's terminal of multimedia data file, and plays the video of the corresponding part.
[50] FIG 10 is a drawing of the present invention showing movement process between the user's terminal of multimedia data file and the system server. As illustrated in FIG 10, the multimedia data file user uses the user's terminal(603, 604) to search and play wanted multimedia data file, and is able to move to search the wanted of a multimedia data file. The user's terminal of multimedia data file(604, 604) connects to the system web server(601), searches wanted contents via the web page, and requests wanted multimedia data file to watch. Then, the streaming server(602) transmits corresponding multimedia data file to the user interface method for play. The user interface method executes search through indexing information list or word and phrase during play of multimedia data file, and if a specific location of multimedia data file is requested, the multimedia data file of the corresponding location is transmitted.
[51] FIG 11 is a drawing of the present invention showing user interface method of search process through indexing information list among multimedia data file search process. As illustrated in FIG 11, the multimedia data file user plays the multimedia data file on the user interface means(701) provided by the web page and can move to
the corresponding part of the multimedia data file is possible by selecting wanted contents from the indexing information list of multimedia data file listed.
[52] The user interface means(701) is consisted of the play unit(702) and the search unit(703). The media player(704), which plays the multimedia data file, is inserted in the play unit(702). The media player(704) is consisted of the display window(705), which shows the multimedia data file, the frame bar(706), which controls the play- point of multimedia data file, and control button(707), which controls the play of videos. If the indication point of frame bar(706) is moved, the multimedia data file plays the contents corresponding to the indication point. Multimedia data file indexing information window(708) and page movement button(710) is included in the search unit(703).
[53] As in FIG 2 to FIG 7, stored identification code, time-information data and text data of a corresponding part are stored in the metadata of multimedia data file. Indexing information list reception unit of the present invention receives information from the metadata of multimedia data file, and outputs the vocal text data of corresponding multimedia data file in list form on the indexing information window(708) of user interface method. Thus, vocal text data of information at a specific part, selected as a set part of multimedia data file by the multimedia data file provider, is presented with the multimedia data file play unit(702).
[54] The user of multimedia data file can control and play multimedia data file with the play unit(702), refer to the multimedia data file's indexing information list(708) of the search unit(703), and select the indexing sentence structure(709). At this point, the identification code transmission part of the present invention transmits identification code corresponding to the selected indexing sentence structure(709) to the web server, and requests time-information data of corresponding vocal text data. The time- information transmission unit transmits time-information data of corresponding vocal text data through the streaming server, and can play the multimedia data file unit of the corresponding part on the user's terminal. Thus, by quickly moving to the corresponding part of selected information through selection of the indexing sentence structure(709) from the indexing information list(708) of multimedia data file, quickly watching and easily searching a set part is possible. If multiple indexing information are assigned, the contents of rest of the indexing information can be played in order.
[55] If page movement button(710) is selected from the search menu(710), movement to user interface method, which is able to search words or phrases, is possible.
[56] FIG 12 is a drawing of the present invention showing the user interface method of search process of searching via words or phrases, among multimedia data file search processes. As illustrated in FIG 12, the multimedia data file user plays the multimedia data file on the user interface means(711) provided through the web page, and searches
wanted information among multimedia data files registered on the database using indexing information search tool(718), which is provided. Also, movement to the corresponding part of the multimedia data file is possible through selection of information from result list(719).
[57] The user interface means(711) is consisted of the play unit(712) and the search unit(713). The media player(714), which plays the multimedia data file, is inserted in the play unit(712). The media player(714) is consisted of the display window(715), which shows the multimedia data file, the frame bar(716), which controls the play- point of multimedia data file, and the control button(717), which controls the play of multimedia data files. If the indication point of the frame bar(716) is moved, the multimedia data file plays the multimedia data corresponding to the indication point. Indexing information search tool(718), search result output window(719) and page movement button(721) is inserted in the search unit(713).
[58] The multimedia data file user inputs words or phrases of search words in the indexing information search tool(718) of multimedia data file, during executing search. The indexing information list reception unit of the present invention receives information related to the corresponding search from the metadata of registered multimedia data file, and outputs the search result, which is vocal text data, on the search result output window (719) in list form(702).
[59] The multimedia data file user can play and control multimedia data files on the play unit(712), execute word or phrase search using the indexing information tool(718) of multimedia data file of the search unit(713), and select wanted indexing sentence structures(720) from the search results.
[60] At this point, the identification code transmission unit of the present invention transmits the identification code corresponding to the selected indexing sentence structure(720) to the web server, and requests the time-information data of corresponding vocal text data. The time-information transmission unit transmits time- information data of corresponding vocal text data to the streaming web server, and transmits multimedia data file unit of the corresponding part to the user's terminal. Thus, by selecting the indexing sentence structure(720) from the indexing information list(719) of the multimedia data file, and moving to the part corresponding to the selected information(720), quickly watching and easily searching a set part is possible. If multiple indexing information is assigned, the contents of rest of the indexing information can be played in order.
[61] If list menu(721) is selected from page movement buttons(721), movement to user interface method, which is searchable through indexing information list, is possible.
[62] FIG 13 is a flow chart of the present invention showing the method of generating indexing information of multimedia data files. When FIG 13 is examined with FIG 1,
according to the present invention, the method of generating indexing information of multimedia data files is the following.
[63] First of all, the multimedia data file provider connects (SlOl) to the web server, and registers(S102) the multimedia data file.
[64] Then, the vocal data extraction unit(l 10) concludes whether vocal data is included in the multimedia data file. If vocal data is included in the multimedia data file, the vocal data extraction unit(l 10) extracts(S104) the vocal data, and the data conversion unit(120) converts(S105) the extracted vocal data into text data. However, if vocal data is not included in the multimedia data file, re-registers(S102) the multimedia data file.
[65] Then, the initial data generation unit(130) combines text data and time-information data of corresponding text data to generate(S106) initial data.
[66] Then, the identification code grant unit(140) selects(S107) a set of data from initial data, and concludes(S108) whether selection of a set of data is completed. If selection of a set of data is completed, the identification code grant unit(140) grants(S109) identification code to the combination of time-information data of a set of data and text data of vocal data. However, if selection of a set of data is not completed, the identification code grant unit(140) concludes whether selection of a set of data is completed until the selection of a set of data is completed.
[67] Then, the record unit(150) records(S 110) the time-information data of a set of data, text data of vocal data, and the identification code combination granted to this on the metadata of the multimedia data file. The storage unit(160) stores(Sl 11) and transmits the multimedia data file in the database.
[68] FIG 14 is a flow chart of the present invention showing the method of searching indexing information of multimedia data files. When FIG 14 is examined with FIG 8 and FIG 9, the method of searching indexing information of multimedia data files is the following.
[69] First of all, the multimedia data file user connects to the web server and searches(S201) multimedia data files on web pages using words or phrases.
[70] Then, select the multimedia data file from the indexing information list and play(S202) it through user interface method.
[71] Then, the indexing information list reception unit(510) receives(S203) indexing information list of a set part of the corresponding multimedia data file from the multimedia data file already played.
[72] Then, the search and indexing sentence structure selection unit(520) concludes(S204) whether the user searched the wanted part of a video through the indexing information list of the multimedia data file. If a specific part of a video is searched through indexing information list of the multimedia data file, the search and indexing sentence structure selection unit(520) selects(S209) the indexing sentence
structure, attempting to move to, from the indexing information list.
[73] Then, the identification code transmission unit(530) transmits the identification code of the selected indexing sentence structure to the web server.
[74] Then, the time-information transmission unit(540) transmits the time-information data of multimedia data file indicated by the corresponding identification code to the user's terminal, which is using the multimedia data file.
[75] Then, the video play unit(550) receives time-information data from the user's terminal, which is using the multimedia data file, and plays(S212) the corresponding part of the video.
[76] However, if a specific part of a video is not searched through the indexing information list of the multimedia data file during the process of S204, the process of S205-S208 is executed instead of the process of S204-209. Thus, the search recognition unit(521) concludes(S205) whether a specific part of the video is searched through sentence structures. Then, if a specific part of the video is searched through sentence structures, the movement unit(522) moves(S206) to the sentence structure interface method of the indexing information. Then, if the search unit(523) executes search using the search tool provided by the sentence structure search interface method, the search results are outputted(S207) in the form of indexing information list. Then, the indexing sentence structure selection unit(524) selects(S208) the indexing sentence structure, attempting to move to, from the search results list.
[77] As seen above, the present invention generates file search index by extracting the vocal data of the multimedia data file, which includes video and vocal. Also, by granting an identification code system to the words and phrases selected by the multimedia data file user, the multimedia data file user can search wanted contents using identification codes, and easily move to a specific part, attempting to search, in a file.
[78] Using the same method as above, the present invention collects massive amounts of multimedia data files, generates file indexing information, and by data-basing the identification code system, allows the user of the web page to search various multimedia data files using words or phrases, and use the related multimedia data information. Thus, the user can select wanted information from the multimedia data file, which represents a set part of the multimedia data file, and while playing the multimedia data file, move to a desired location and continue watching the multimedia data file.
[79] As seen above, although, the technological concept behind method of generating indexing information of multimedia data file using vocal data and searching indexing information of multimedia data file are described with the attached diagrams, this only explains the most exemplary execution example of the present invention, not limit the present invention. Also, to a person who possesses common knowledge of this
technology field, it is clear that various modification and imitation is possible without breaking away from the technological concept of the present invention.