Summary of the invention
In view of this, the invention provides information search method and device in a kind of Web Community, thereby reduce user's searching cost.
Technical scheme of the present invention specifically is achieved in that
Information search method in a kind of Web Community, the method comprises:
Receive query word, according in Web Community's internal data, the feature of every kind of data type, determine the correlativity between query word and every kind of data type, according to described correlativity, determines the data type that the information search scope contains; Wherein, described data type, be in advance according to the relation between the hunting zone of Web Community's internal data and user view search, and Web Community's internal data is divided and obtained;
According to described query word, search for community's internal data corresponding to every kind of data type in described information search scope, obtain Search Results corresponding to every kind of data type in described information search scope, the output Search Results.
Information retrieval device in a kind of Web Community, this device comprises hunting zone determination module, search module and output module;
Described hunting zone determination module, for receiving query word, according in Web Community's internal data, the feature of every kind of data type, determine the correlativity between query word and every kind of data type, according to described correlativity, determine the data type that the information search scope contains; Wherein, described data type, be in advance according to Web Community's internal data and user view search hunting zone between relation, Web Community's internal data is divided and is obtained;
Described search module, for according to described query word, search for community's internal data corresponding to every kind of data type in described information search scope, obtains Search Results corresponding to every kind of data type in described information search scope;
Described output module, for exporting Search Results.
As seen from the above technical solution, the present invention is after receiving query word, first according in Web Community's internal data, the feature of every kind of data type, determine the correlativity of the inside, community that query word is corresponding with every kind of data type, data type due to Web Community inside, in advance according to the search intention feature of Web Community's internal data, it is the relation between the hunting zone of Web Community's internal data and user view search, the Web Community internal data is divided and obtained, therefore, by described correlativity, search intention that can the anticipation user, it is the information search scope of user view search, therefore determine the information search scope according to described correlativity, can determine all information search scopes that meet user search intent, it is as follows that this makes the present invention have advantages of:
On the one hand, if user view is searched for community's internal data of multiple community element or numerous types of data, adopt the present invention to be searched for community's internal data of this multiple community element or numerous types of data once, and need not repeatedly select community's element and restart search procedure, on the other hand, because the present invention not all includes all communities internal data in the information search scope simply, thereby, the quantity of irrelevant Search Results can be reduced to a certain extent.Visible, the present invention can reduce user's searching cost.
Embodiment
Fig. 1 is the information search method process flow diagram in Web Community provided by the invention.
As shown in Figure 1, the method comprises:
Step 101, receive query word, and according in Web Community's internal data, the feature of every kind of data type, determine the correlativity between the query word that receives and every kind of data type.
Wherein, described data type, be in advance according to the search intention feature of Web Community's internal data, be the relation between the hunting zone of Web Community's internal data and user view search, Web Community's internal data is divided and is obtained.
For example, the user may be intended in the various hunting zones of search, and a typical hunting zone is the stronger data of real-time such as news, and the data that all real-times in Web Community's internal data are stronger can be divided into a data type.
Step 102, according to described correlativity, determine the data type that the information search scope contains.
By step 101 ~ step 102, can realize anticipation user's search intention, thereby selection meets community's internal data corresponding to the data type of user search intent as the information search scope.
Step 103, according to described query word, search for community's internal data corresponding to every kind of data type in described information search scope, obtains Search Results corresponding to every kind of data type in described information search scope.
Step 104, the output Search Results.
Visible, by method shown in Fig. 1, can realize that all communities internal data corresponding to more than one data type that will meet once user search intent all include the information search scope in, and need not start respectively for different community's elements search routine, and, with all include all communities internal data in the information search scope simply and compare, also can exclude the community's internal data that does not meet user search intent, thereby dwindle the information search scope, reduce the quantity of irrelevant Search Results, therefore, can reduce the cost that the user selects Search Results.
Wherein, in advance according to the search intention feature of Web Community's internal data, be the relation between the hunting zone of Web Community's internal data and user view search, Web Community's internal data is divided to the data type obtained, typically, can include but not limited to: resource sharing data, group data, user data, game application data and/or real time data, and long-tail text data.
Every kind of data type, can corresponding more than one community element.For example, the resource sharing data, bias toward sharing of resource information, and corresponding community's element can comprise novel read module, music playing module etc.
Wherein, according to the feature of data type, the correlativity between the query word determine received and every kind of data type, can adopt accomplished in many ways, and the present invention is not restricted this.For example, the feature of every kind of data type can mean that by proper vector the feature of query word also can mean by proper vector, angle between proper vector by calculating query word and the proper vector of every kind of data type, determine the query word of reception and the correlativity between every kind of data type; Again for example, also can be by the mode of cluster, determine the distance between query word and every kind of data type, according to described distance, determine the correlativity between the query word that receives and every kind of data type.
Wherein, the covering scope of long-tail text data is very extensive, for example user journal is exactly a kind of typical long-tail text data, for any query word, usually corresponding Search Results can be searched in the long-tail text data, therefore, usually, the long-tail text data, as the master data source, for any query word, all should be included the long-tail text data in the information search scope.In other words, step 102 determines that the step of information search scope specifically can comprise: meet resource sharing data, user data, game application data and/or the real time data of predetermined condition by the long-tail text data and with the correlativity of query word, as the information search scope.
Select the cost of Search Results in order further to reduce the user, the present invention also proposes, can be when the output Search Results, corresponding data type information and/or the community's element information of sign Search Results, make the user can understand which data type Search Results belongs to or derive from which community's element, selects Search Results thereby be convenient to the user.
Wherein, corresponding data type information and/or the community's element information of sign Search Results includes but not limited to following methods:
Method one, show different data types or Search Results corresponding to community's element in different zones.For example, can be according to data type or community's element, the sub-module display of search results, according to each data type or community's internal data corresponding to community's element and the descending order of the correlativity between query word, determine the position relationship between the viewing area, Search Results place that different types of data or community's element are corresponding, show same data type or Search Results corresponding to community's element in same viewing area, show different types of data or Search Results corresponding to community's element in different viewing areas.
Method two, corresponding every Search Results, identify data type and/or community's element information that this Search Results is corresponding.Between each Search Results, can be arranged according to the correlativity with query word.
According to said method provided by the invention, the present invention also provides corresponding device, specifically refers to Fig. 2.
Fig. 2 is the information retrieval device structural drawing in Web Community provided by the invention.
As shown in Figure 2, this device comprises hunting zone determination module 201, search module 202 and output module 203.
Hunting zone determination module 201, for receiving query word, according in Web Community's internal data, the feature of every kind of data type, correlativity between the query word determine received and every kind of data type, according to described correlativity, determine the data type that the information search scope contains; Wherein, described data type, be in advance according to the relation between the hunting zone of Web Community's internal data and user view search, and Web Community's internal data is divided and obtained.
Search module 202, for according to described query word, search for community's internal data corresponding to every kind of data type in described information search scope, obtains Search Results corresponding to every kind of data type in described information search scope.
Output module 203, for exporting Search Results.
Wherein, the data type of Web Community's internal data can comprise resource sharing data, group data, user data, game application data and/or real time data, and the long-tail text data.
Hunting zone determination module 201, can be for by the long-tail text data and with the correlativity of query word, meeting resource sharing data, group data, user data, game application data and/or the real time data of predetermined condition, as the information search scope.
Output module 203 wherein, can be for identifying data type information and/or the community's element information that Search Results is corresponding; Wherein, the every kind of corresponding more than one community of data type element.
Particularly, output module 203, can show for the zone different different data types or Search Results corresponding to community's element; Perhaps, corresponding every Search Results, identify data type and/or community's element information that this Search Results is corresponding.
Output module 203, can be according to data type or community's element, the subregion display of search results, for according to each data type or community's internal data corresponding to community's element and the descending order of the correlativity between query word, determine the position relationship between the viewing area, Search Results place that different types of data or community's element are corresponding, show same data type or Search Results corresponding to community's element in same viewing area, show different types of data or Search Results corresponding to community's element in different viewing areas.
Technical scheme provided by the invention, by analyzing the characteristics of the inner various data of whole Web Community, Web Community's internal data is divided into to numerous types of data, data type can reflect user's search intention, correlativity between every kind of data type of the query word of inputting by analysis user and Web Community's internal data, search intention that can the anticipation user, thereby match the data type that meets user search intent according to query word, and then using Web Community's internal data corresponding to the data type that matches as the information search scope, realization is in Search Results, integrate polytype data, be convenient to the user and select Search Results, particularly when the data type information corresponding to Search Results or community's element information are identified, due to clear and definite icon or official documents and correspondence sign being arranged between different types of data or Search Results corresponding to community's element, can allow the user distinguish the difference between different data types clearly, thereby can further make the user select at a glance the Search Results be satisfied with most, further reduced the cost that the user selects Search Results.
Particularly in the situation that mobile phone screen is less, due to the present invention do not need repeatedly to switch search institute based on community's element and startup search procedure, also do not need all to include all communities internal data in the information search scope, therefore can greatly reduce the cellphone subscriber selects the cost of Search Results on mobile phone screen.Visible, the present invention has met the information search demand of user in Web Community well.
Below, for several concrete examples, technical scheme of the present invention is carried out to exemplary illustration, given example is not used in restriction the present invention:
For example, suppose the title that the query word of user's input is certain novel, can determine that the user is the strongest to the search intention of novel reading, secondly be the group of this novel of discuss and exchange, again secondly for seeing the user of this novel, it is finally the long-tail text data, the data type information that should include hunting zone in is: the resource sharing data, group data, user data and long-tail text data, the position that the Search Results subregion of output is showed is followed successively by from top to bottom or by left-to-right: the Search Results in the resource sharing data, Search Results in group data, Search Results in Search Results in user data and long-tail text data.
Again for example, suppose the title that the query word of user's input is certain money game, can determine that the user is the strongest to the search intention of community's game, secondly be the group of this game of discuss and exchange, again secondly for playing the user of this game, it is finally the long-tail text data, the data type information that should include hunting zone in is: the game application data, group data, user data and long-tail text data, the position that the Search Results subregion of output is showed is followed successively by from top to bottom or by left-to-right: the Search Results in the game application data, Search Results in group data, Search Results in Search Results in user data and long-tail text data.
Again for example, the query word of supposing user's input is the hot word of certain news, can determine that the user is the strongest to the search intention of ageing content, secondly be the group of the hot word of this news of interchange open to discussion, it is finally the long-tail text data, the data type that should include hunting zone in is: real time data, group data and long-tail text data, the position that the Search Results subregion of output is showed is followed successively by from top to bottom or by left-to-right: the Search Results in the Search Results in real time data, the Search Results in group data and long-tail text data.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of making, be equal to replacement, improvement etc., within all should being included in the scope of protection of the invention.