Embodiment
In order to make the application's technical matters to be solved, technical scheme and beneficial effect is clearer, understand, below in conjunction with drawings and Examples, the application is further elaborated.Should be appreciated that specific embodiment described herein only in order to explaining the application, and be not used in restriction the application.
Can not filter out the commodity of high-quality in order to solve the e-business network site server that exists in the prior art, provide suitable merchandise news for the buyer as the side of browsing accurately, the path that makes the side of browsing browse merchandise news increases, increase the problem that the operation complexity of suitable commercial product is searched by the side of browsing, a kind of information providing method that the embodiment of the present application provides, by system's realization as shown in Figure 1, this system comprises information-providing server 1 and client 2.Wherein:
Client 2 be used for to obtain the search word of user's input, and described search word is offered information-providing server 1, and the information display that information-providing server 1 is provided is to the user.
Information-providing server 1 is according to the search word of user's input, in conjunction with search engine to the Query Result of the information of the publisher that stores in information publication database issue and the navigation algorithm correlativity calculation result to the information of the side of the browsing click of storing in the information acquisition database, in conjunction with default priority regulation rule, the priority that provides of each information category is provided, and is provided the information of mating with described search word to the user according to the described priority that provides.
When information-providing server 1 was used for e-commerce website, the information of publisher's issue of its storage was the merchandise news of seller's issue, and the information that the side of browsing of storage clicks also refers to the merchandise news that the buyer clicks.
The flow process of above-mentioned information providing method as shown in Figure 2, its principle comprises the steps: as shown in Figure 3
Step S11: the search word that obtains user's input.
The user can use client inputted search word, general user as the side of browsing can search own interested information by the mode of inputted search word, server side is searched the information that the publisher shows according to the search word of the side of browsing user input, for the side of browsing user provides.
Step S12: according to the search word that obtains, the information of the publisher who stores in Query Information publication database issue, the information content of determining to comprise in each information category the search word of acquisition accounts for the quantitative proportion of the information total quantity of the search word that comprises acquisition.
The process of quantification ratio specifically comprises: comprise the total quantity of information of the search word of acquisition in the information of publisher's issue of determining to store in the information publication database, and the first quantity of information that comprises the search word of acquisition in each information category; Respectively each first quantity of going out of calculative determination and the ratio of total quantity, the information content that obtains comprising in each information category search word accounts for the ratio of the information total quantity of the search word that comprises acquisition.
Wherein, store the information of all publisher's issues in the information publication database, as shown in Figure 3, search for these information by search engine inquiry, this is the true statistics to the information of publisher's issue, can obtain accordingly embodying the relation of industry distribution data of incidence relation of the information of the side's of browsing search word and publisher issue, wherein the industry distribution data can be understood as the distribution situation of certain product in each information category.
Step S13: according to the search word that obtains, the information that the side of browsing that stores in the Query Information acquisition database clicks, according to the clicked number of times of information of the search word that comprises acquisition in each information category, determine the search word of acquisition and the relevance weight of each information category.
The process of determining relevance weight specifically comprises: comprise the clicked total degree of information of the search word of acquisition in the information of determining the side of browsing that stores in the information acquisition database to click, and the first clicked number of clicks of information that comprises the search word of acquisition in each information category; Respectively each first number of clicks of going out of calculative determination and the ratio of total degree, the clicked number of times of information that obtains comprising in each information category the search word of acquisition accounts for the relevance weight of the clicked total degree of information of the search word that comprises acquisition.
Wherein, store the image data of the information that the side of browsing is clicked in the information acquisition database, record the various information that each side of browsing once clicked.As shown in Figure 3, pass through navigation algorithm, according to image data, the search word of the calculating side of browsing input and the correlativity of each information category, this is the statistics to the side's of browsing operation behavior data, can obtain accordingly embodying the industry distribution statistics of the side's of browsing search word and the incidence relation of the side's of browsing operation behavior.
Step S14: according to above-mentioned definite quantitative proportion and relevance weight, determine to comprise the prediction priority of each information category of the search word of acquisition.Wherein, prediction priority is that a priority level of each information category of tentatively determining according to quantitative proportion and relevance weight puts in order, with each information category by the prediction priority arrangement after, follow-up again according to will be wherein designated the information category of priority the priority that provides of each information category is provided after adjusting.
The process of determining prediction priority specifically comprises: obtain default first weighting factor corresponding with quantitative proportion and second weighting factor corresponding with relevance weight; The product of the quantitative proportion that calculates respectively and the first weighting factor, and the product of the relevance weight that obtains and the second weighting factor, and to the summation of the product that obtains obtain comprising the predicted value of each information category of the search word of acquisition; Determine prediction priority according to the size of the predicted value that obtains.
Wherein, the first weighting factor is the weighting coefficient of quantitative proportion when calculating predicted value of determining, the second weighting factor is the weighting coefficient of relevance weight when calculating predicted value of determining, the value of the first weighting factor and the second weighting factor can arrange as required, preferably, the span of the first weighting factor and the second weighting factor is 0~1.
Wherein, when determining prediction priority according to the size of the predicted value that obtains, can be according to the predicted value size order of each information category, determine the prediction priority of each information category, for example: the prediction priority of the information category of predicted value maximum is the highest, and the prediction priority of the information category of predicted value minimum is minimum; Vice versa.
For example as shown in Figure 3, calculate by the side's of browsing industry intention calculate platform: the size of the predicted value of each information category of the search word that comprises acquisition that obtains is expressed as f (x), then:
Wherein, C represents the search word that obtains and the relevance weight of each information category, is to call the user search word navigation algorithm anticipatory data that navigation algorithm obtains to distribute; X represents the weighting factor corresponding with relevance weight, is a threshold values of determining according to website platform True Data distribution situation; The information content that B/A represents to comprise in each information category the search word of acquisition accounts for the quantitative proportion of the information total quantity of the search word that comprises acquisition, B is that the information that comprises search word that obtains of calling search engine is at each information category, be the product data distribution situation of every profession and trade, A is that the overall product data of the information that comprises search word in all row that calling search engine obtains distribute; (1-x) the expression weighting factor corresponding with quantitative proportion.
Certainly, this just calculates a kind of optional form of predicted value, can also calculate by some modes of texturing, and for example: two weighting factor sums are not equal to 1, etc.
Step S15: according to the prediction priority of determining and default priority regulation rule, determine the priority that provides of each information category.Wherein, the rule that above-mentioned definite prediction priority is adjusted that the priority regulation rule refers to dispose, wherein can specify certain/priority of a little information categories.
Determining provides the process of priority specifically to comprise: inquire about default priority regulation rule, obtain the assigned priority of selected information classification in the information category that comprises search word; According to assigned priority prediction priority is adjusted, obtain each information category corresponding priority is provided.
Assigned priority refer to certain/a little information categories can dispose its priority, the priority of for example specifying certain information category is 1, the priority of another information category is 5 etc.Refer on the basis of prediction priority arrangement order, the information category of wherein having specified priority be adjusted to place, corresponding priority position to predicting that priority is adjusted according to assigned priority.
Configuration intervention rule as shown in Figure 3 arranges and can realize arranging the priority regulation rule, and the intervention rule that comprises search word shown in Figure 3 realizes obtaining the assigned priority of selected information classification in the information category that comprises search word.
Step S16: provide the information of mating with search word to the user according to the priority that provides of determining.
As shown in Figure 3, information-providing server provides the information of mating with search word according to the client that provides priority to use to the user of determining, and presents to the user by client.When providing priority to provide information to the user, can be according to providing priority order from high to low all to show the user with the information of search word coupling.Also can specify the categorical measure of the information that offers the user, when having specified categorical measure, according to priority from high to low order determines to provide to the user information of which information category.For example only provide the prediction priority bit in the information of the information category of front three, the first five or first eight bits to the user, specifically former positions are determined according to the categorical measures of appointment.
As shown in Figure 4, take user's inputted search word mp3 as example, describe the specific implementation process of above-mentioned information providing method in detail.
After user's inputted search word mp3, server gets access to this search word, calls the product information total amount that the search engine that the mass data inquiry service is provided obtains the side's of browsing search word product information industry data distribution and single industry; Namely obtain the information content that comprises search word mp3 in the total quantity of information of the publisher issue that comprises search word mp3 and each information category.
For example shown in Fig. 4, information category or say the product information industry comprise mobile phones, car video, car dvd player ... Deng.Such as (the Mp3 among the figure, car dvd player, information content) expression (side's of browsing search word, the product information industry, the information content that the side's of browsing search word inquires in this product information industry), the data set that search engine returns sequentially can be arranged from high to low according to information content.
After user's inputted search word mp3, server gets access to this search word, call the navigation algorithm part, the navigation algorithm part can provide the strong industry of relevance according to the user search word and browse catalogue, obtains the side's of browsing search word expection industry data and distributes and corresponding weight.Namely obtain the information content that comprises search word mp3 in each information category of the total quantity of the information that comprises search word mp3 of clicking the side of browsing and the side of browsing click.
Again for example shown in Fig. 4, navigation algorithm partly obtains (mp3, mobile phones, weight) expression (side's of browsing search word, the product information industry, the relevance weight of the side's of browsing search word and this product information industry), the data set order partly returned of navigation algorithm also can be arranged from high to low by weighted value.This part is actually and obtains the side's of browsing industry intention data distribution.
Then partly carry out conformity calculation by the side's of browsing industry intention calculate platform, carry out conformity calculation according to the user search word navigation algorithm anticipatory data that product industry True Data distributes and navigation algorithm partly the provides distribution that search engine provides, for example can carry out conformity calculation according to the computing formula of above-mentioned f (x).
For example: the data that search engine provides comprise { (mp3, mobile phones, product quantity A), (mp3, car video, product quantity B), (mp3, car dvd player, product quantity C), (mp3, navigation﹠amp; GPS, product quantity D), (mp3, mp3 players, product quantity E), (mp3, speaker, product quantity F), (mp3, digital photo frame, product quantity G), (mp3, car mp3 player, product quantity H)---}.Be that search engine provides the quantity that comprises the information of search word mp3 in the above-mentioned product industry of enumerating (such as mobile phones etc.).
The data that navigation algorithm partly provides comprise: (mp3, mobile phones, weight a), (mp3, mp3players, weight b), (mp3, navigation﹠amp; GPS, weight c), (mp3, car mp3 player, weight d), (mp3, mp4 players, weight e), (mp3, car video, weight f), (mp3, cardvd player, weight g), (mp3, speaker, weight h)---}.Be that navigation algorithm provides in the above-mentioned product industry of enumerating (such as mobile phones etc.) and comprises the information of search word mp3 and the relevance weight of search word mp3.
Configuration Interference service platform part, intervening the navigation algorithm anticipatory data distributes, increase the more reasonable and correctness distribution of data, dispose and intervene the high priority data consideration, f (x) data are leaned on rear arrangement, and go coincidence and reach data to rearrange, for example: the regulation rule that the configuration intervention is set is data acquisition { (mp3, mp3 players, 2), (mp3, mp4 players, 6), (mp3, digital photo frame, 0)---}, wherein, 2,6 ordinal positions of indicating to arrange, the related data of 0 this data industry of expression will be deleted, it can be a small amount of that data are intervened in configuration, can be not other with magnitude with f (x), i.e. and the assigned priority of a predetermined fraction information category.
The f (x) that is intended to obtain after calculate platform calculates by the side's of browsing industry is { (mp3, mobile phones), (mp3, car video), (mp3, mp3 players), (mp3, car dvd player), (mp3, mp4 players), (mp3, navigation﹠amp; GPS), (mp3, speaker), (mp3, car mp3 player), (mp3, digital photo frame)----after, in conjunction with the result that configuration is intervened, the data acquisition after the merging that obtains comprises { (mp3, mp3 players, 2), (mp3, mp4 players, 6), (mp3, digital photo frame, 0), (mp3, mobile phones), (mp3, car video), (mp3, mp3 players), (mp3, car dvd player), (mp3, mp4 players), (mp3, navigation﹠amp; GPS), (mp3, speaker), (mp3, car mp3 player), (mp3, digital photo frame)----,
After the first step merges, the priority arrangement that provides of each information category of going heavily to process with the positional alignment in the 3rd step of carrying out again second step sequentially is: { (mp3, mobile phones), (mp3, mp3 players), (mp3, car video), (mp3, car dvd player), (mp3, navigation﹠amp; GPS), (mp3, mp4 players), (mp3, speaker), (mp3, car mp3 player) }.
Based on above-mentioned information providing method, a kind of information provider unit that the embodiment of the present application also provides, this device can be arranged in the server in as shown in Figure 5 the information providing system, the structure of this device comprises as shown in the figure: ratio-dependent module 10, weight determination module 20, prediction priority determination module 30 and priority determination module 40 is provided.
Ratio-dependent module 10, be used for the search word according to the user's input that obtains, the information of the publisher that stores in Query Information publication database issue, the information content of determining to comprise in each information category the search word of acquisition accounts for the quantitative proportion of the information total quantity of the search word that comprises acquisition.
Weight determination module 20, be used for the search word according to the user's input that obtains, the information that the side of browsing that stores in the Query Information acquisition database clicks, according to the clicked number of times of information of the search word that comprises acquisition in each information category, determine the search word of acquisition and the relevance weight of each information category.
Prediction priority determination module 30, be used for the quantitative proportion of each information category of determining according to ratio-dependent module 10 and the relevance weight of each information category that weight determination module 20 is determined, determine to comprise the prediction priority of each information category of the search word of acquisition.
Priority determination module 40 is provided, be used for the prediction priority of determining according to prediction priority determination module 30 and the priority regulation rule of presetting, the priority that provides of each information category is provided, and is provided the information of mating with the search word that obtains to the user according to the priority that provides that prediction priority determination module 30 is determined.
Preferably, aforementioned proportion determination module 10 comprises the total quantity of information of the search word of acquisition and the first quantity that comprises the information of search word in each information category in the information of the concrete publisher's issue that is used for determining that the information publication database is stored; The first quantity that each information category of going out of calculative determination is corresponding respectively and the ratio of above-mentioned total quantity, the information content that obtains comprising in each information category the search word of acquisition accounts for the ratio of the information total quantity of the search word that comprises acquisition.
Preferably, above-mentioned weight determination module 20, the clicked total degree of information that comprises the search word of acquisition in the information of the concrete side of the browsing click that is used for determining that the information acquisition database is stored, and the first clicked number of clicks that comprises the information of search word in each information category; The first number of clicks that each information category of going out of calculative determination is corresponding respectively and the ratio of total degree, the clicked number of times of information that obtains comprising in each information category the search word of acquisition accounts for the relevance weight of the clicked total degree of information of the search word that comprises acquisition.
Preferably, above-mentioned prediction priority determination module 30, concrete for obtaining the first weighting factor corresponding to default quantitative proportion and the second weighting factor corresponding to relevance weight; Calculate respectively the product of quantitative proportion corresponding to each information category relevance weight corresponding with each information category of sum of products of the first weighting factor and the second weighting factor, and to the summation of the product that obtains, obtain comprising the predicted value of each information category of the search word of acquisition; Determine to comprise the prediction priority of each information category of the search word of acquisition according to the size of the predicted value that obtains.
Preferably, the above-mentioned priority determination module 40 that provides, the assigned priority of concrete information category selected information classification for obtaining the search word that comprises acquisition; According to the assigned priority that obtains prediction priority is adjusted, obtain comprising acquisition search word each information category priority is provided.
Above-mentioned information provider unit also can be regarded the side's of browsing industry intention as and determine platform, can be arranged in the server of network side, also can be set to separately an independently equipment.
As shown in Figure 6, at server side, the side's of browsing industry intention is determined platform and search engine module, the data statistics module of navigation algorithm and the priority regulation rule module setting acting in conjunction that configuration is intervened, aggregation products industry True Data (search engine), navigation algorithm anticipatory data (navigation algorithm), artificial operation data (configuration is intervened) carries out comprehensive computing, realizes determining of priority is provided.Wherein, the priority regulation rule that configuration is intervened arranges to provide interface is set, allow the operation staff to dispose intervention, input the assigned priority of some selected information classification, the inquiry of priority regulation rule, the regulation rule corresponding to search word of inquiring user input is such as assigned priority etc.The side's of browsing industry intention determines that platform provides the output interface to each e-commerce website.
As shown in Figure 6, the side's of browsing industry intention determines that platform provides the serviceization mode, and has open OpenAPI interface mode, so that any application call to be provided, for example, www.aliexpress.com uses, and search.aliexpress.com uses, and www.alibaba.com uses, and call with the OpenAPI interface mode etc. the partner site, for example, be used for the navigation of keyword search results page classification in aliexpress.com, the keyword search results page is pressed sales volume ordering etc.;
So long as need to judge the application of the side's of browsing industry intention, all directly call this platform, do not need each application to control oneself and again develop these codes, resource reuse, maintenance and renewal are convenient, and cost is low.
Said method is according to natural Search Results and the comprehensive operation result of user click data, intervene the result in conjunction with the distribution of actual products information industry data and the distribution of expection industry data with artificial the operation, obtaining rational data distributes, more precisely determine the side's of browsing intention, reduce the side's of browsing browse path, strengthen website viscosity, filter out the more product of high-quality, prevent swindle, provide a unified serviceization platform, with artificial operation data intervention, and come into force in real time, the serviceization of providing mode can be called flexibly, safeguards and expand, and improves resource reuse.
Wherein, search engine data and navigation algorithm data merge, can prevent effectively that the user from clicking maliciously swindle, the interference normal data distributes, and configuration is intervened the data merging and come into force in real time, can avoid in some Sparse situations, the accurate problem of data, unified platform of final formation in the serviceization mode, offers external call flexibly.
Above-mentioned explanation has illustrated and has described the application's preferred embodiment, but as previously mentioned, be to be understood that the application is not limited to the disclosed form of this paper, should not regard the eliminating to other embodiment as, and can be used for various other combinations, modification and environment, and can in invention contemplated scope described herein, change by technology or the knowledge of above-mentioned instruction or association area.And the spirit and scope that the change that those skilled in the art carry out and variation do not break away from the application, then all should be in the protection domain of the application's claims.