Method and system of searching and collating video files, establishing semantic group, and program storage medium therefor

1. A method of searching and collating video files, comprising:

retrieving a semantic keyword according to video information, wherein the semantic keyword comprises title information;

searching according to the semantic keyword in a video platform and obtaining a plurality of candidate video files;

parsing each of the candidate video files by using a video semantic searching and collating process, selecting a plurality of matched video files and grouping the matched video files; and

gathering the grouped matched video files in at least one video channel set,

wherein the step of retrieving the semantic keyword according to the video information comprises:

retrieving the semantic keyword from metadata of the video information or from source information related to the video information,

if the metadata of the video information comprises a program title and a category, directly parsing the program title and the category in the metadata to serve as the semantic keyword, and

if the metadata of the video information does not comprise the program title or the category, retrieving the program title and the category corresponding to an electronic program guide (EPG) of the video information through a network according a currently viewed channel number and a current time.

2. The method according to claim 1 , wherein the EPG complies with service information (SI) in moving picture experts group-2 (MPEG-2) transport stream.

3. The method according to claim 1 , wherein the EPG of the video information is metadata describing program information which is retrieved by using an EPG parser.

4. The method according to claim 1 , the video semantic searching and collating process step comprising

filtering the candidate video files, grouping and collating all of or part of the filtered candidate video files in a set according to the semantic keyword.

5. The method according to claim 4 , wherein the video semantic searching and collating process further comprising a program semantic parsing method, the program semantic parsing method filtering the candidate video files by using a filter style so as to filter out the candidate video files matching the filter style; and

parsing the candidate video files by using a semantic group so as to gather semantics of the candidate video files.

6. The method according to claim 5 , wherein the semantic group comprises an episode semantic group, a date semantic group, a clip number semantic group, a definition semantic group or any combination thereof.

7. The method according to claim 4 , further comprising a step of performing program semantic search algorithm to combine program title and semantic information as at least one set of searching conditions to search in the candidate video files and obtain the matched video files.

8. The method according to claim 1 , wherein the step of gathering the grouped matched video files in the at least one video channel set comprises

based on program semantic search algorithm, changing semantics to search the matched video files having consecutive episodes to establish the at least one video channel set.

9. The method according to claim 1 , further comprising:

parsing validity for the matched video files of the at least one video channel set after a predetermined time period to divide the matched video files in the at least one video channel set into a valid program sub set and an invalid program sub set.

10. The method according to claim 1 , further comprising:

periodically changing the semantic keyword to search a plurality of updated candidate video files; and

parsing the updated candidate video files by the video semantic searching and collating process and selecting a plurality of updated matched video files to update the at least one video channel set.

11. The method according to claim 1 , further comprising:

providing the matched video files in the at least one video channel set with a recommended list to select the at least one video channel set.

12. A system of searching and collating video files, comprising:

a retrieving unit, retrieving a semantic keyword according to video information, wherein the semantic keyword comprises title information;

a searching unit, searching according to the semantic keyword in a video platform to obtain a plurality of candidate video files; and

a processing unit, parsing each of the candidate video files by using a video semantic searching and collating process, selecting a plurality of matched video files and grouping the matched video files so as to gather the grouped matched video files in at least one video channel set,

wherein the retrieving unit retrieves the semantic keyword from a part of metadata of the video information or from source information related to the video information,

wherein if the metadata of the video information comprises a program title and a category, the program title and the category in the metadata is directly parsed to serve as the semantic keyword, and

if the metadata of the video information does not comprise the program title and the category, the program title and the category corresponding to the EPG is retrieved through a network according a currently viewed channel number and a current time.

13. The system according to claim 12 , wherein the EPG is service information (SI) in transport stream complying with moving picture experts group-2 (MPEG-2) standard.

14. The system according to claim 12 , wherein the EPG of the video information is metadata describing program information which is retrieved by using an EPG parser.

15. The system according to claim 12 , wherein the processing unit filters the candidate video files by using the video semantic searching and collating process, groups and collates all of or part of the filtered candidate video files in a set according to the semantic keyword.

16. The system according to claim 15 , wherein the processing unit performs a program semantic parsing method adopting a filter style to filter the candidate video files so as to filter out the candidate video files matching the filter style and adopting a semantic group parsing semantics of the candidate video files to collect the semantics of the candidate video files.

17. The system according to claim 16 , wherein the filter style comprises a universal filter semantic group and a user-defined filter semantic group, wherein

the universal filter semantic group serves to filter part of the candidate video files which have titles comprising program type restriction semantics, and the user-defined filter semantic group serves to filter part of the candidate video files which have titles comprising undesired semantics.

18. The system according to claim 16 , wherein the semantic group comprises an episode semantic group, a date semantic group, a clip number semantic group, a definition semantic group or any combination thereof.

19. The system according to claim 15 , wherein the processing unit performs program semantic search algorithm to combine program title and semantic information as at least one set of searching conditions to search in the candidate video files and obtain the matched video files.

20. The system according to claim 12 , wherein the processing unit gathers the grouped matched video files in the at least one video channel set to be based on program semantic search algorithm, changes semantics to search the matched video files having consecutive episodes to establish the at least one video channel set.

21. The method according to claim 6 , further comprising:

compiling at least one first-language episode term, at least one first-language date term, at least one first-language part number term or at least one first-language clip term in a first language respectively into the episode semantic group, the date semantic group, a part number semantic group or the clip number semantic group;

compiling at least one second-language episode term, at least one second-language date term, at least one second-language part number term or at least one second-language clip term in a second language respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group;

compiling synonyms of the at least one second-language episode term, synonyms of the at least one second-language date term, synonyms of the at least one second-language part number term or synonyms of the at least one second-language clip term respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group; and

compiling synonym abbreviations of the at least one second-language episode term, synonym abbreviations of the at least one second-language date term, synonym abbreviations of the at least one second language part number term or synonym abbreviations of the at least one second-language clip term respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group.

22. The method according to claim 21 , wherein a format of the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group is a regular expression.

23. The method according to claim 21 , wherein after step of compiling the synonym abbreviations of the at least one second-language episode term, the synonym abbreviations of the at least one second-language date term, the synonym abbreviations of the at least one second language part number term or the synonym abbreviations of the at least one second-language clip term respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group, the method further comprises:

compiling a filter semantic group and a definition semantic group.

24. The method according to claim 21 , before the step of compiling the synonyms of the at least one second-language episode term, the synonyms of the at least one second-language date term, the synonyms of the at least one second-language part number term or the synonyms of the at least one second-language clip term respectively into the episode semantic group, the method further comprises:

identifying the synonyms of the at least one second-language episode term, the at least one synonyms of the second-language date term, the synonyms of the at least one second-language part number term or the synonyms of the at least one second-language clip term by using a synonym kit.

25. A non-volatile and non-transitory computer storage medium for storing a computer program, which configured for an electronic apparatus to perform a method of searching and collating video files comprising steps of:

retrieving a semantic keyword according to video information, wherein the semantic keyword comprises title information;

searching according to the semantic keyword in a video platform and obtaining a plurality of candidate video files;

parsing each of the candidate video files by using a video semantic searching and collating process, selecting a plurality of matched video files and grouping the matched video files;

gathering the grouped matched video files in at least one video channel set,

compiling at least one first-language episode term, at least one first-language date term, at least one first-language part number term or at least one first-language clip term in a first language respectively into the episode semantic group, the date semantic group, a part number semantic group or the clip number semantic group;

compiling at least one second-language episode term, at least one second-language date term, at least one second-language part number term or at least one second-language clip term in a second language respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group;

compiling synonyms of the at least one second-language episode term, synonyms of the at least one second-language date term, synonyms of the at least one second-language part number term or synonyms of the at least one second-language clip term respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group; and

compiling synonym abbreviations of the at least one second-language episode term, synonym abbreviations of the at least one second-language date term, synonym abbreviations of the at least one second language part number term or synonym abbreviations of the at least one second-language clip term respectively into the episode semantic group, the date semantic group, the part number semantic group or the clip number semantic group.