CN102117331B - Video search method and system - Google Patents
Video search method and system Download PDFInfo
- Publication number
- CN102117331B CN102117331B CN201110053876.9A CN201110053876A CN102117331B CN 102117331 B CN102117331 B CN 102117331B CN 201110053876 A CN201110053876 A CN 201110053876A CN 102117331 B CN102117331 B CN 102117331B
- Authority
- CN
- China
- Prior art keywords
- querying command
- video
- type information
- object video
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention provides a video search method, comprising the following steps of: receiving an inquiry command; analyzing the inquiry command, and outputting an inquiry command corresponding to the inquiry command and type information of at least one video object; analyzing the important video object type information in the type information of at least one video object; searching the inquiry command and the important video object information in an inverted indexing to obtain a search result; and outputting the search result. The invention has the beneficial effects that: corresponding video object type information is marked on each video object excavated on a network by a machine, the efficiency is higher, the cost is lower, and the error rate is lower; besides, the intention of a search term input by a user also can be analyzed, network video search result which can be paid close attention by the user is returned at first time, thus the search result is easier to meet the user requirement, but also the search efficiency of the user can be improved, the network flow can be reduced, and the network resource can be saved.
Description
Technical field
The present invention relates to search engine technique, relate in particular to a kind of video searching method and system.
Background technology
Growth at full speed along with internet information, has been full of increasing redundant information on network, and for the Internet user of the own required information of search on network, in the face of these information that extend endlessly are undoubtedly as looking for a needle in a haystack.The appearance of search engine is undoubtedly to a certain extent for user's search need has brought convenience.Search engine is a kind of software systems of applying on network, and it collects and discovery information with certain strategy on network, and after information being processed and organized, for user provides the information search service on internet.Conventionally, this software systems provide a web interface, allow user by browser software, submit search word in client to, then return to very soon the relevant information list of search content that may input with user.This list can comprise up to ten thousand entries conventionally, and each entry represents one piece of related web page searching.
Since more than ten years in past, correspondingly, arise at the historic moment in numerous internet search engines and corresponding website, the outstanding person in the middle of this comprises Baidu's search (www.baidu.com) of company of Baidu and Google's search (www.google.cn) of Google.
Along with the development of network technology, user has no longer been satisfied with the just search to text to the requirement of search engine, and a lot of users also wish to Internet video, to search for by search engine.Usually, the Internet video searching out by a search word may comprise a plurality of types, as shown in Figure 1, after searching for by " Liu Dehua " this search word, for the accordingly result that facilitates user to find it to need, in results page, can Search Results be divided into a plurality of tab1 according to the type of Internet video, as: music, film, TV play, comprehensive, information, concert, TV programme, imitation etc., when user clicks certain tab, can jump to the Search Results of such network video type.Yet, in order to reach object like this, when setting up index, need to be first each object video (OBJ) characteristic information excavating from network, and mark its corresponding object video type information (TYPE), so that when retrieval, exportable results page as shown in Figure 1.Yet, in the prior art, the corresponding object video type information of described mark by people for carrying out, not only waste time and energy, cost is higher, and by people for classifying, error rate is also relatively high, affects the Search Results that user obtains in search, waste Internet resources, in addition, in the prior art, when user's input is not searched for the search word of object video type information, still as shown in Figure 1, after searching for by " Liu Dehua " this search word, the results page that it returns, what give tacit consent to is the accordingly result in output " all " this tab, need to treat that user clicks other tab and is, with search word+tab word of user, in index, search for just now, and export Search Results, such way of search, because user's search word not being analyzed in advance, fail to recognize user's search intention, can not directly show accordingly result in the tab that user needs the very first time, cause user again to click, just can obtain it and need result, search efficiency is lower, cause network traffics larger.For example: the film < < Tangshan Earthquake > > of hot showing recently, before, it is a TV play, yet, if user searches for " Tangshan Earthquake " in the near future, the intention of the < < Tangshan Earthquake > > of its search filmization should be higher than the intention of the < < Tangshan Earthquake > > of the acute version of searching television, therefore it is the corresponding Search Results of " film " that result of page searching should directly be shown tab.
Summary of the invention
The object of the present invention is to provide a kind of improved video searching method, it not only can mark corresponding object video type information to each object video excavating on network by machine, also can analysis user the search word intention of input, the very first time is returned to the Internet video Search Results that user's most probable is paid close attention to.
The present invention also aims to provide a kind of video searching system of realizing above-mentioned video searching method.
One of for achieving the above object, a kind of video searching method of the present invention, comprises the following steps:
S1, reception querying command;
S2, analyze described querying command, and output and the corresponding querying command of described querying command and at least one object video type information;
Important object video type information in S3, described at least one the object video type information of analysis;
S4, described querying command and described important object video information are searched in inverted index, obtained Search Results;
S5, export the tab key of described Search Results and described querying command and the combination of object video type information.
As a further improvement on the present invention, described inverted index builds by subordinate's step:
Capture id information and the characteristic information of object video;
By disaggregated model, described characteristic information is carried out to classified calculating, obtain at least one corresponding object video type information;
Described object video type information and described object video id information are integrated to generation inverted index.
As a further improvement on the present invention, described characteristic information comprises: website, duration, text, video rules, URL, issuing time.
As a further improvement on the present invention, described classified calculating is incremental computations.
As a further improvement on the present invention, described disaggregated model is support vector machine.
As a further improvement on the present invention, described S2 step comprises following sub-step:
By knowledge data base, analyze each element in described querying command;
Whether described each element of judgement is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, and obtain described object video type information.
As a further improvement on the present invention, described algorithm is sorting algorithm.
As a further improvement on the present invention, described sorting algorithm is decision tree.
As a further improvement on the present invention, described S2 step comprises following sub-step:
By knowledge data base, analyze each element of the historical query order in described querying command and querying command database;
Whether described each element of judgement is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, and obtain described object video type information.
As a further improvement on the present invention, described algorithm is sorting algorithm.
As a further improvement on the present invention, described sorting algorithm is decision tree.
As a further improvement on the present invention, described S2 step comprises following sub-step:
By natural language processing, extract the various elements in described querying command;
Judge whether described various element is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, and obtain described object video type information.
As a further improvement on the present invention, described algorithm is sorting algorithm.
As a further improvement on the present invention, described sorting algorithm is decision tree.
As a further improvement on the present invention, described S2 step comprises following sub-step:
By natural language processing, extract the various elements of the historical query order in described querying command and querying command database;
Judge whether described various element is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, and obtain described object video type information.
As a further improvement on the present invention, described algorithm is sorting algorithm.
As a further improvement on the present invention, described sorting algorithm is decision tree.
As a further improvement on the present invention, in described step S3, be to filter out described important object video type by user's historical behavior.
Correspondingly, as realizing above-mentioned another object, a kind of video searching system of the present invention comprises:
UI module, for receiving querying command, and is sent to described querying command analysis module by this querying command; And after being assemblied into results page, exports on the tab key of Search Results and described querying command and the combination of object video type information;
Querying command analysis module, for analyzing described querying command, and output and the corresponding querying command of described querying command and at least one object video type information; And for analyzing the important object video type information of described at least one object video type information;
Search module, for described querying command and described important object video information are searched at inverted index, obtains Search Results.
As a further improvement on the present invention, described video searching system also comprises:
Web services module, for receive the querying command transmitting from client by procotol, and forwards this querying command to UI module.
As a further improvement on the present invention, the described results page that described web services module is also returned for receiving described UI module, and described results page is back to client.
As a further improvement on the present invention, described video searching system also comprises:
Disaggregated model module, for receiving id information and the characteristic information of the object video grabbing, and carries out classified calculating by disaggregated model to described characteristic information, obtains at least one corresponding object video type information;
Index module, for integrating generation inverted index by described object video type information and described object video id information.
As a further improvement on the present invention, described characteristic information comprises: website, duration, text, video rules, URL, issuing time.
As a further improvement on the present invention, it is characterized in that, described classified calculating is incremental computations.
As a further improvement on the present invention, described disaggregated model is support vector machine.
As a further improvement on the present invention, described querying command analysis module comprises:
Querying command database, for storing user's historical behavior;
Priori database, for storing known knowledge;
Analytic unit, in conjunction with described querying command database and described priori database, and calculates the query statement receiving by algorithm, obtain at least one object video type information.
As a further improvement on the present invention, described algorithm is sorting algorithm.
As a further improvement on the present invention, described sorting algorithm is decision tree.
As a further improvement on the present invention, then described analytic unit also with by user's historical behavior, described object video type information being filtered out to described important object video type information.
Compared with prior art, the invention has the beneficial effects as follows: the present invention can mark corresponding object video type information to each object video excavating on network by machine, and efficiency is higher, cost is lower, and error rate is lower; In addition the present invention also can analysis user the search word intention of input, the very first time is returned to the Internet video Search Results that user's most probable is paid close attention to, not only Search Results is more easily met consumers' demand, simultaneously, also can improve user search efficiency, reduce network traffics, save Internet resources.
Accompanying drawing explanation
Fig. 1 is the video search result page in prior art;
Fig. 2 is that video searching system of the present invention and client realize interactive fundamental diagram;
Fig. 3 is the module map of video searching system one embodiment of the present invention;
Fig. 4 is the submodule figure of querying command analysis module one embodiment of the present invention;
Fig. 5 is the submodule figure of another embodiment of querying command analysis module of the present invention;
Fig. 6 is the process flow diagram of video searching method one embodiment of the present invention;
Fig. 7 is the process flow diagram of index building method of the present invention;
Fig. 8 is the process flow diagram of S2 step 1 embodiment of the present invention;
Fig. 9 is the process flow diagram of another embodiment of S2 step of the present invention.
Embodiment
Below with reference to each embodiment shown in the drawings, describe the present invention.But these embodiments do not limit the present invention, the conversion in the structure that those of ordinary skill in the art makes easily according to these embodiments, method or function is all included in protection scope of the present invention.
Video searching system of the present invention 10 shown in Fig. 2 is realized interactive fundamental diagram with client 20.In present embodiment, this client 20 comprises a browser, and client can open search engine by this browser, and input inquiry order in search engine, general, the querying command of this input is text message, certainly, this querying command can also be pictorial information, video information etc.Described video searching system 10 receives client by network and inputs to the querying command in described browser, and after this querying command is searched for, Search Results is back to this browser.Wherein, this video searching system 10 can comprise one or more server, this client 20 can comprise one or more subscriber terminal equipments, as personal computer, notebook computer, wireless telephone, personal digital assistant (PDA) or other department of computer science communication system of unifying.
These servers and terminal device all comprise some basic modules on framework, as bus, disposal system, storage system, one or more input/output and communication interface etc.Bus can comprise one or more wires, is used for realizing the communication between server or each assembly of terminal device.Disposal system comprises that all types of being used for carry out processor or the microprocessor of instruction, treatment progress or thread.Storage system can comprise the dynamic storagies such as random access storage device (RAM) of storing multidate information, with the static memories such as ROM (read-only memory) (ROM) of storage static information, and the mass storage that comprises magnetic or optical record medium and respective drive.Input system arrives server or terminal device for user's input information, as keyboard, mouse, writing pencil, sound recognition system or bioassay system etc.Output system comprises display for output information, printer, loudspeaker etc.Communication interface is used for making server or terminal device and other system or system to communicate.Between communication interface, can be connected in network by wired connection, wireless connections or light, make video searching system 10,20 of clients realize mutual communication by network.Network can comprise that Local Area Network, wide area network (WAN), telephone network are as combination of internet, the Internet or above-mentioned these networks of public switch telephone network (PSTN), enterprises etc.
On server and terminal device, all include for management of system resource, control the operating system software that other program is moved, and the application software that is used for realizing certain functional modules.As shown in Figure 3, described video searching system 10 can be divided into two parts, and part set up in video search part and index.Wherein, described video search partly comprised web services module 101, with the UI module 102 of web services module 101 interactive communications, with querying command analysis module 103 and the search module 104 of described UI module 102 interactive communications; Described index is set up part and is comprised disaggregated model module 105 and the index module 106 of communicating by letter with described disaggregated model module 105.It is worth mentioning that, these modules can store and run in same server, also can store and operate in multiple servers.
Described web services module 101 is for receiving the querying command transmitting from client 20 by procotol, and forward this querying command to UI module 102, in addition, the results page that this web services module 101 is also returned for receiving described UI module 102, and described results page is back to client 20.
The querying command that described UI module 102 transmits for receiving described web services module 101, and this querying command is sent to described querying command analysis module 103, simultaneously, receive querying command+object video type information that described querying command analysis module 103 returns, again described querying command+object video type information is sent in described search module 104 and is searched for, in addition, the Search Results that described UI module 102 is also returned for receiving described search module 104, and described Search Results is assemblied into after results page, return to described results page to described web services module 101.
The querying command that described querying command analysis module 103 transmits for receiving described UI module 102, and by analysis after described querying command, return to querying command+at least one object video type information to described UI module 102, preferably, if after querying command analysis module 103 is analyzed, what return is a querying command+mono-object video type information, a described object video type information is important object video type information, and 102 of described UI modules are sent to search module 104 by described querying command+important object video type information, if after querying command analysis module 103 is analyzed, what return is querying command+a plurality of object video type informations, for example: querying command+the first object video type information, the second object video type information ... etc., now, the querying command that UI module 102 analyzes described querying command analysis module 103+important object video type information exports search module 104 to, search for, and other object video type information is only assemblied into a tab key by UI module and shows in the results page of output, described querying command analysis module 103 also screens for the importance degree of the object video type information to described, filter out important object video type.How to obtain this important object video type information, will in following Fig. 4, Fig. 5, elaborate.
Described search module 104 is for receiving the querying command+important object video type information of described UI module 102 inputs, and by searching in the inverted index of described querying command+important object video type information in described index module 106, and return to Search Results to described UI module 102, preferably, this Search Results is the object video set corresponding to the querying command+important object video type information of UI module 102 inputs.
Described disaggregated model module 105 is for receiving object video (OBJ) characteristic information grabbing from network, usually, this characteristic information can comprise: website (the object video type that different website may provide is different), duration, text, video rules (Pattern, the description rule that refers to querying command, the input habit that can represent user, for example: user is when searching television is acute, often can input " TV play name+numeral " this special querying command, during search film, often can input " movie name+film " this special querying command, in general, described video rules is what manually sum up), URL, issuing time etc., and calculate at least one corresponding object video type information (TYPE) by disaggregated model, preferably, described disaggregated model module is SVM(support vector machine), the calculating adopting is a kind of incremental computations, be to only have when occurring changing just to calculate, wherein, this disaggregated model obtains by machine learning mode, be first by manually selecting a collection of sample, be by manually a collection of characteristic information being classified, obtain corresponding object video type information, secondly by the mode of machine learning, obtain this disaggregated model, the mode classification of this machine learning, those of ordinary skills can skillfully grasp by prior art, do not repeat them here.Analytic unit also screens for the importance degree of the object video type information to described, filters out important object video type.
Described index module 106 is for receiving the object video type information of the object video id information that grabs from network and 105 outputs of described disaggregated model module, and described object video id information and described object video type information are generated as to inverted index, for described search module 104, search for.
Wherein, as shown in Figure 4, in first embodiment of the invention, described querying command analysis module 103 comprise analytic unit 1031, with the querying command database 1032 of described analytic unit interactive communication, and with the priori database 1033 of described analytic unit 1031 interactive communications.
Described analytic unit 1031 is for receiving described querying command, and call described querying command database 1032 and priori database 1033, by extended mode in conjunction with described querying command database 1032 and priori database 1033 after, output querying command+at least one object video type information.Concrete steps are: first receive described querying command, secondly each element of analyzing in described querying command by knowledge data base 1033, as make the name of an article, collection of drama number etc., preferably, (adopting the historical query order in querying command database 1032 is mainly for cumulative data amount by knowledge data base 1033, to analyze each element of the historical query order in described querying commands and querying command database 1032, to obtain enough elements), and judge by machine whether described each element is consistent with the video rules of summing up in advance, if meet, obtain the video rules corresponding with described querying command, finally, by described video rules, obtain described object video type information again, it is worth mentioning that: usually, described video rules and described object video type information are relations one to one, be by sorting algorithm, described video rules to be calculated, can obtain described object video type information, preferably, in best mode for carrying out the invention, described video rules is to calculate described object video type information by decision tree, the application of described decision tree, those of ordinary skills can skillfully grasp by prior art, do not repeat them here, certainly, in other embodiments of the present invention, also can pass through C4.5, or SVM, or the sorting algorithm such as IF/ELSE calculates described object video type information by described video rules.
Described analytic unit 1031 is also in conjunction with described querying command database 1032, importance degree to described object video type information screens, judge the most important object video type information of possibility concerning user, such as the user's clicking rate by record in querying command database 1032, querying command history etc. screened.For example: the querying command of user's input is " Tangshan Earthquake ", the historical behavior of searching for as inquiring user by described querying command database 1032(, or user clicks the historical behavior of Search Results etc.) can learn, " Tangshan Earthquake " that inquiry object video type information is film is recently more, can judge for " Tangshan Earthquake " of this search, may most important object video type information should be film.
Described querying command database 1032, for storing user's historical behavior, comprises user's historical query order.
Described priori database 1033 is for storing known knowledge.For example: " Liu Dehua " is a people's name; " Tangshan Earthquake " is movie name and TV play name etc.
Wherein, as shown in Figure 5, in second embodiment of the invention, described querying command analysis module 103 comprises analytic unit 1031, querying command database 1032, priori database 1033, and natural language processing module 1034.
Described analytic unit 1031 is for receiving from the various elements of described natural language processing module 1034 outputs, and call described querying command database 1032 and priori database 1033, by extended mode in conjunction with described querying command database 1032 and priori database 1033 after, output querying command+at least one object video type information.Concrete steps are: first receive from the various elements of described natural language processing module 1034 outputs, secondly by machine, judge that whether described each element is consistent with the video rules of summing up in advance, if meet, obtains the video rules corresponding with described querying command, finally, by described video rules, obtain described object video type information again, it is worth mentioning that: usually, described video rules and described object video type information are relations one to one, be by sorting algorithm, described video rules to be calculated, can obtain described object video type information, preferably, in best mode for carrying out the invention, described video rules is to calculate described object video type information by decision tree, the application of described decision tree, those of ordinary skills can skillfully grasp by prior art, do not repeat them here, certainly, in other embodiments of the present invention, also can pass through C4.5, or SVM, or the sorting algorithm such as IF/ELSE calculates described object video type information by described video rules.
Described analytic unit 1031 is also in conjunction with described querying command database 1032, importance degree to described object video type information screens, judge important object video type information concerning user, such as the user's clicking rate by record in querying command database 1032, querying command history etc. screened.For example: the querying command of user's input is " Tangshan Earthquake ", the historical behavior of searching for as inquiring user by described querying command database 1032(, or user clicks the historical behavior of Search Results etc.) can learn, " Tangshan Earthquake " that inquiry object video type information is film is recently more, can judge for " Tangshan Earthquake " of this search, may most important object video type information should be film.
Described querying command database 1032, for storing user's historical behavior, comprises user's historical query order.
Described priori database 1033 is for storing known knowledge.For example: " Liu Dehua " is a people's name; " Tangshan Earthquake " is movie name and TV play name etc.
Described natural language processing module 1034 is for receiving the querying command of described UI unit 102 inputs, and described querying command is carried out exporting described analytic unit 1031 to after natural language processing, this natural language processing is for extracting the various elements in described querying command, as make the name of an article, collection of drama number etc., understand user's search intention.
As shown in Figure 6, the video searching method of an embodiment of the present invention, comprises the following steps:
S1, reception querying command; Preferably, this querying command be user by the browser input in client to web services module 101, this web services module 101, after obtaining described querying command, can forward this querying command to UI module 102;
S2, analyze described querying command, and export querying command+at least one object video type information; Wherein, this analysis completes by querying command analysis module 103;
Important object video type information in S3, described at least one the object video type information of analysis, preferably, if after querying command analysis module 103 is analyzed, what return is a querying command+mono-object video type information, a described object video type information is important object video type information, and 102 of described UI modules are sent to search module 104 by described querying command+important object video type information, if after querying command analysis module 103 is analyzed, what return is querying command+a plurality of object video type informations, for example: querying command+the first object video type information, the second object video type information ... etc., now, by querying command analysis module 103, in described a plurality of object video type informations, analyze important object video type information, UI module 102 exports querying command+important object video type information to search module 104, search for, and other object video type information is only assemblied into a tab key by UI module and shows in the results page of output,
S4, described querying command+important object video information is searched in inverted index; Wherein, described search completes by search module 104, querying command+important object video type information that described search module 104 is inputted for receiving described UI module 102, and by searching in the inverted index of described querying command+important object video type information in described index module 106;
S5, export described Search Results.Wherein, this Search Results is back to described UI module 102 from described search module 104, and by described UI module, described Search Results is assemblied into after results page, return to described results page to described web services module 101, thereby be back to client browser by described web services module 101.
As shown in Figure 7, the inverted index in the video searching method of an embodiment of the present invention, builds by following steps:
Id information and the characteristic information of S100, crawl object video (OBJ), preferably, after grabbing described characteristic information, by described characteristic information input disaggregated model module 105, usually, this characteristic information can comprise: website (the object video type that different website may provide is different), duration, text, video rules (Pattern, the description rule that refers to querying command, the input habit that can represent user, for example: user is when searching television is acute, often can input " TV play name+numeral " this special querying command, during search film, often can input " movie name+film " this special querying command, in general, described video rules is what manually sum up), URL, issuing time etc.,
S101, by disaggregated model, described characteristic information is carried out to classified calculating, obtain at least one corresponding object video type information (TYPE), preferably, described disaggregated model is SVM(support vector machine), the calculating adopting is a kind of incremental computations, be to only have when occurring changing just to calculate, wherein, this disaggregated model obtains by machine learning mode, be first by manually selecting a collection of sample, be by manually a collection of characteristic information being classified, obtain corresponding object video type information, secondly by the mode of machine learning, obtain this disaggregated model,
S102, described object video type information and described object video id information are integrated and generated inverted index.Wherein, generating inverted index completes by index module 106.
As shown in Figure 8, in one embodiment of the present invention, above-mentioned S2 step completes by following sub-step:
The querying command that S21, reception are exported by UI module;
S22, each element of analyzing in described querying command by knowledge data base 1033, preferably, by knowledge data base 1033, analyze each element (adopting the historical query order in querying command database 1032 is mainly for cumulative data amount, to obtain enough elements) of the historical query order in described querying command and querying command database 1032;
Whether S23, described each element of judgement are consistent with the video rules of summing up in advance;
S24, if meet, by algorithm, calculate described video rules, and the object video type information described in obtaining, it is worth mentioning that: usually, described video rules and described object video type information are relations one to one, be by sorting algorithm, described video rules to be calculated, can obtain described object video type information, preferably, in best mode for carrying out the invention, described video rules is to calculate described object video type information by decision tree, the application of described decision tree, those of ordinary skills can skillfully grasp by prior art, do not repeat them here, certainly, in other embodiments of the present invention, also can pass through C4.5, or SVM, or the sorting algorithm such as IF/ELSE calculates described object video type information by described video rules,
If S25 does not meet, do not process.
As shown in Figure 9, in another embodiment of the present invention, above-mentioned S2 step completes by following sub-step:
The querying command that S21 ', reception are exported by UI module;
S22 ', by natural language processing, extract the various elements in described querying command, preferably, by natural language processing, extract each element (adopting the historical query order in querying command database 1032 is mainly for cumulative data amount, to obtain enough elements) of the historical query order in described querying command and querying command database 1032;
S23 ', judge that whether described various element is consistent with the video rules of summing up in advance;
S24 ', if meet, by algorithm, calculate described video rules, and the object video type information described in obtaining, it is worth mentioning that: usually, described video rules and described object video type information are relations one to one, be by sorting algorithm, described video rules to be calculated, can obtain described object video type information, preferably, in best mode for carrying out the invention, described video rules is to calculate described object video type information by decision tree, the application of described decision tree, those of ordinary skills can skillfully grasp by prior art, do not repeat them here, certainly, in other embodiments of the present invention, also can pass through C4.5, or SVM, or the sorting algorithm such as IF/ELSE calculates described object video type information by described video rules.
S25 ', if do not meet, does not process.
Known by foregoing description, the present invention can mark corresponding object video type information to each object video excavating on network by machine, and efficiency is higher, cost is lower, and error rate is lower; In addition the present invention also can analysis user the search word intention of input, the very first time is returned to the Internet video Search Results that user's most probable is paid close attention to, not only Search Results is more easily met consumers' demand, simultaneously, also can improve user search efficiency, reduce network traffics, save Internet resources.
Be to be understood that, although this instructions is described according to embodiment, but not each embodiment only comprises an independently technical scheme, this narrating mode of instructions is only for clarity sake, those skilled in the art should make instructions as a whole, technical scheme in each embodiment also can, through appropriately combined, form other embodiments that it will be appreciated by those skilled in the art that.
Listed a series of detailed description is above only illustrating for feasibility embodiment of the present invention; they are not in order to limit the scope of the invention, all disengaging within equivalent embodiment that skill spirit of the present invention does or change all should be included in protection scope of the present invention.
Claims (19)
1. a video searching method, is characterized in that, described video searching method comprises the following steps:
S1, reception querying command, described querying command is text message;
S2, analyze described querying command, and output and the corresponding querying command of described querying command and at least one object video type information;
Important object video type information in S3, described at least one the object video type information of analysis;
S4, described querying command and described important object video information are searched in inverted index, obtained Search Results;
S5, export described Search Results;
In described step S2, analyzing described querying command specifically comprises:
By priori database or natural language processing, extract each element in described querying command or extract described querying command and querying command database in each element of historical query order;
Whether described each element of judgement is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, obtain described object video type information.
2. video searching method according to claim 1, is characterized in that, described inverted index builds by subordinate's step:
Capture id information and the characteristic information of object video;
By disaggregated model, described characteristic information is carried out to classified calculating, obtain at least one corresponding object video type information;
Described object video type information and described object video id information are integrated to generation inverted index.
3. video searching method according to claim 2, is characterized in that, described characteristic information comprises: website, duration, text, video rules, URL, issuing time.
4. video searching method according to claim 2, is characterized in that, described classified calculating is incremental computations.
5. video searching method according to claim 2, is characterized in that, described disaggregated model is support vector machine.
6. video searching method according to claim 1, is characterized in that, described algorithm is sorting algorithm.
7. video searching method according to claim 6, is characterized in that, described sorting algorithm is decision tree.
8. video searching method according to claim 1, is characterized in that, in described step S3, is to filter out described important object video type by user's historical behavior.
9. a video searching system, is characterized in that, described video searching system comprises:
UI module, for receiving querying command, described querying command is text message, and this querying command is sent to described querying command analysis module; And after being assemblied into results page, exports Search Results;
Querying command analysis module, for analyzing described querying command, and output and the corresponding querying command of described querying command and at least one object video type information; And for analyzing the important object video type information of described at least one object video type information;
Search module, for described querying command and described important object video information are searched at inverted index, obtains Search Results;
Wherein said querying command analysis module, when analyzing described querying command, is specifically carried out:
By priori database or natural language processing, extract each element in described querying command or extract described querying command and querying command database in each element of historical query order;
Whether described each element of judgement is consistent with the video rules of summing up in advance;
If meet, by algorithm, calculate described video rules, obtain described object video type information.
10. video searching system according to claim 9, is characterized in that, described video searching system also comprises:
Web services module, for receive the querying command transmitting from client by procotol, and forwards this querying command to UI module.
11. video searching systems according to claim 10, is characterized in that, the described results page that described web services module is also returned for receiving described UI module, and described results page is back to client.
12. video searching systems according to claim 9, is characterized in that, described video searching system also comprises:
Disaggregated model module, for receiving id information and the characteristic information of the object video grabbing, and carries out classified calculating by disaggregated model to described characteristic information, obtains at least one corresponding object video type information;
Index module, for integrating generation inverted index by described object video type information and described object video id information.
13. video searching systems according to claim 12, is characterized in that, described characteristic information comprises: website, duration, text, video rules, URL, issuing time.
14. video searching systems according to claim 12, is characterized in that, described classified calculating is incremental computations.
15. video searching systems according to claim 12, is characterized in that, described disaggregated model is support vector machine.
16. video searching systems according to claim 9, is characterized in that, described querying command analysis module comprises:
Querying command database, for storing user's historical behavior;
Priori database, for storing known knowledge;
Analytic unit, for by priori database or natural language processing, extract each element in described querying command or extract described querying command and querying command database in each element of historical query order; Whether described each element of judgement is consistent with the video rules of summing up in advance; If meet, by algorithm, calculate described video rules, obtain described object video type information.
17. video searching systems according to claim 16, is characterized in that, described algorithm is sorting algorithm.
18. video searching systems according to claim 17, is characterized in that, described sorting algorithm is decision tree.
19. video searching systems according to claim 16, is characterized in that, so described analytic unit is also with by user's historical behavior, described object video type information being filtered out to described important object video type information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110053876.9A CN102117331B (en) | 2011-03-07 | 2011-03-07 | Video search method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110053876.9A CN102117331B (en) | 2011-03-07 | 2011-03-07 | Video search method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102117331A CN102117331A (en) | 2011-07-06 |
| CN102117331B true CN102117331B (en) | 2014-09-17 |
Family
ID=44216101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110053876.9A Active CN102117331B (en) | 2011-03-07 | 2011-03-07 | Video search method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102117331B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140379731A1 (en) * | 2012-02-27 | 2014-12-25 | Amol Sunil Dixit | Video search |
| CN102955858B (en) * | 2012-11-09 | 2016-07-27 | 北京百度网讯科技有限公司 | The search ordering method of video file, system and server |
| CN103324682B (en) * | 2013-05-31 | 2017-06-06 | 百度在线网络技术（北京）有限公司 | It is a kind of that the method and device of video is searched for based on picture |
| TWI575391B (en) * | 2013-06-18 | 2017-03-21 | 財團法人資訊工業策進會 | Social data filtering system, method and non-transitory computer readable storage medium of the same |
| CN104182440A (en) * | 2014-02-26 | 2014-12-03 | 无锡天脉聚源传媒科技有限公司 | Video searching method and system |
| CN106777287B (en) * | 2016-12-29 | 2020-07-31 | 上海聚力传媒技术有限公司 | Method and equipment for providing video search result |
| CN109597929A (en) * | 2018-09-21 | 2019-04-09 | 北京字节跳动网络技术有限公司 | Methods of exhibiting, device, terminal and the readable medium of search result |
| CN114511022B (en) * | 2022-01-24 | 2022-12-27 | 百度在线网络技术（北京）有限公司 | Feature screening, behavior recognition model training and abnormal behavior recognition method and device |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060020624A (en) * | 2003-05-12 | 2006-03-06 | 코닌클리케 필립스 일렉트로닉스 엔.브이. | Method of searching for media objects |
| CN101369287A (en) * | 2008-09-25 | 2009-02-18 | 四川长虹电器股份有限公司 | Method for implementing audio/video data excavation |
| CN101419614A (en) * | 2008-12-03 | 2009-04-29 | 深圳市迅雷网络技术有限公司 | Video resource clustering method and device |
| CN101916265B (en) * | 2010-07-30 | 2012-07-25 | 百度在线网络技术（北京）有限公司 | Inquiry based method and device for showing local resource of user equipment |
-
2011
- 2011-03-07 CN CN201110053876.9A patent/CN102117331B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN102117331A (en) | 2011-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102117331B (en) | Video search method and system | |
| CN102096717B (en) | Search method and search engine | |
| CN101334792B (en) | Personalized service recommendation system and method | |
| CN101408894B (en) | System and method for enhancing search relevancy using semantic keys | |
| CN102073725B (en) | Method for searching structured data and search engine system for implementing same | |
| CN103339623B (en) | It is related to the method and apparatus of Internet search | |
| CN102073726B (en) | Structured data import method and device for search engine system | |
| CN102708174B (en) | Method and device for displaying rich media information in browser | |
| CN104246748B (en) | System and method for determining situation | |
| CN102135985B (en) | Method and system for searching by calling search result of third-party search engine | |
| CN102054003B (en) | Methods and systems for recommending network information and creating network resource index | |
| CN106096056A (en) | A kind of based on distributed public sentiment data real-time collecting method and system | |
| CN111708740A (en) | Mass search query log calculation analysis system based on cloud platform | |
| CN102110170B (en) | System with information distribution and search functions and information distribution method | |
| CN111666490A (en) | Information pushing method, device, equipment and storage medium based on kafka | |
| CN102063476A (en) | Video searching method and system | |
| CN101681372A (en) | Method and system for providing relevant information to a user of a device in a local network | |
| WO2008141295A1 (en) | Keyword generation system and method for online activity | |
| CN103092962B (en) | A kind of method and system issuing internet information | |
| CN102999586A (en) | Method and device for recommending website | |
| CN102521232B (en) | Distributed acquisition and processing system and method of internet metadata | |
| CN102073735A (en) | Searching method and searching system | |
| CN105007314A (en) | Big data processing system oriented to mass reading data of readers | |
| CN110232126A (en) | Hot spot method for digging and server and computer readable storage medium | |
| CN104090923A (en) | Method and device for displaying rich media information in browser |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170110 Address after: 100193 room 303-305, building, East District, No. 10, Wang Dong Road, Beijing, Haidian District Patentee after: Beijing small mutual Entertainment Technology Co., Ltd. Address before: 100085 Beijing, Haidian District, No. ten on the ground floor, No. 10 Baidu building, layer 2 Patentee before: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY Co.,Ltd. |