CN104951468A - Data searching and processing method and system - Google Patents

Abstract

The present application provides a data search processing method and system. Wherein, the method includes: obtaining the first ranking score of each search object in the search results based on the first ranking model, and dividing the first ranking score into a plurality of intervals; classifying each search object into a range according to the first ranking score In the search object set corresponding to each interval; determine the search object with the preset mark in the search object set corresponding to each interval; obtain the second ranking score of the search object with the preset mark based on the second ranking model; use the The second sorting function adjusts the ranking of the search object with the preset mark in the search object set corresponding to the section. As a result, on the premise of ensuring the relevance of search results, the consistency and continuity of the returned search results display are improved, providing users with a good consistent experience, and simplifying the algorithm to reduce data processing complexity, improve efficiency and system processing performance.

Description

Data search processing method and system

technical field

The present application relates to the field of data search, in particular to a data search processing method and system.

Background technique

With the development of Internet technology, more and more users conduct data search through network access and obtain feedback search results. The server-side data search processing technology that executes the search according to the search request and provides the results plays an important role in realizing the user's search purpose, for example, how to process the search results to obtain the results that best meet the user's needs, and how to process the search results to improve the server's performance. Processing performance, optimizing data management efficiency, etc. The existing search processing technology, according to the user's search request, is found by the search engine and the association engine according to the query words (such as keywords), that is, the search engine finds the data object, and the extension engine finds the extended information based on the data object, and then, Process and adjust the data object and the extended information based on the data object and return the output together. For example, embed the found extended information based on the data object into the result of the found data object, and display it to the user who entered the query word.

A common application is to embed paid advertisements into search results in commodity search engines, specifically, as shown in FIG. 1A . (1) The user accesses the product search website through a browser, enters product query words, and presses the search button to request a search. (2) The browser accesses the application server of the website. (3) The application server requests the advertisement engine for the advertisement result of this search (commodity-based advertisement creative result), and at the same time requests the search engine for the commodity search result for this search; (i) Among them, the advertisement engine Logically returns the advertising results, for example: match the keywords purchased by the advertiser according to the query words, get the qualified advertising products, and then decide the ranking according to the maximum expected profit of the advertisement (similarly considering the advertising bid, matching degree, creative quality, etc.), take The advertising ideas of the top m (top m) advertising products are returned as results; (ii) among them, the search engine returns the search results according to certain logic, for example: match the text description of the product according to the query word to get the qualified product, and then According to the degree of matching between the products calculated according to the dimensions of relevance and product quality and the needs of the user who sent the search request, the order of the output products is determined, and the top n (top n) products are returned as the results. (4) The application server obtains the advertisement results and the searched product results, and performs calculations, such as filtering out the corresponding products (advertised products) that already exist in the advertising results from the searched product results; merges the calculated results, and adjusts the ranking; Render the page and return the results to the browser for display to the user who made the search request.

From the process in Figure 1A, the search result returns the output. Taking "commodity trading platform search" as an example, the paid advertisement is displayed next to the searched product, such as the head, tail, right column, etc., as part of the search result, as shown in the figure 2 shown in the right column. Here, the advertisement part is displayed independently, and the browser can directly access the advertisement result obtained from the advertisement engine, and directly display it at the corresponding advertisement position, which can shorten the page processing time. In addition, the output method can also be returned from the search results shown in Figure 1B. As shown in Figure 3, the "bidding ranking" display, the paid advertisement is embedded in the search result, and when the search result is output to the web page, the paid advertisement is also surrounded by a box . Here, the advertisement results and the search results are mixed together, and after the obtained advertisement results and the obtained search results are mixed (for example, using a mixed sorting server), the application server transmits the mixed results to the browser.

The two display output methods after search processing are to display the results of the search engine and the results of the advertising engine on one page. However, both methods have certain defects.

First, because the final displayed results are generated by combining the results of the two engines, and the product sets corresponding to the two engines are different, and the sorting algorithms are different, the final returned results show discontinuous and irrelevant adverse effects, resulting in poor user experience. Inconsistency, especially when ad results and search results are mixed. Therefore, due to the inconsistency of the sorting logic adopted by the two engines, the final returned output results are poor in effect, lack of continuity and relevance, and thus lead to inconsistent user experience. defect.

For example: the total collection of products is A, B, C, D, E, F, and the collection of products participating in advertisements is C, D, E, then the collection of products in the search engine is the complete collection of products A ~ F, and the collection of products in the advertising engine Advertisement products C to D. The possibility of the user-initiated search is as follows: the search engine returns results A, C, and F, and the advertising engine returns results C, E, which are mixed and displayed to the user ACEF. Since ACF is displayed according to the search engine sorting rules, E inserted into it will confuse users to judge. From the perspective of advertisement ranking, even if E’s text description is not closely related to the user’s query, if E’s bid is high, it will still be returned to the user. At this time, the overall result for the user’s experience is poor relevance, discontinuous, and inconsistent.

Second, in the existing technology, the application server needs to request two engines. The goals of the two engines are inconsistent, and the sorting conditions considered by each are inconsistent. Returning the final output results requires operations such as merging and deduplication of the target results of the two engines. , resulting in inconsistencies in the final sorting of the same object, thus resulting in increased cumbersome operations such as shuffling and deduplication, increasing the complexity of the computer system, and causing low processing efficiency of the computer system.

Therefore, it is necessary to improve the above-mentioned data search processing solution in the prior art to improve efficiency and provide users with a consistent and good user experience.

Contents of the invention

The main purpose of this application is to provide a data search processing method and system to solve technical problems such as improving the consistency and continuity of the returned search results on the premise of ensuring the relevance of the search results, so as to provide users with a good consistent Sexual experience; further, complex algorithms such as shuffling and deduplication are reduced to solve technical problems such as reducing the complexity of data processing, improving data processing efficiency, and improving the performance of data search and processing systems. in:

One aspect of the present application provides a data search processing method, including: obtaining the first ranking score of each search object in the search result based on the first ranking model; dividing the first ranking score into multiple intervals, according to the first ranking score A sorting point classifies each search object into the search object set corresponding to each interval; determines the search object with the preset mark in the search object set corresponding to each interval; obtains the search object with the preset mark based on the second sorting model The second ranking score of the search object; using the second ranking score to adjust the ranking of the search object with the preset mark in the search object set corresponding to the interval.

Wherein, obtaining the first ranking score of each search object in the search result based on the first ranking model includes: obtaining the search result according to the keyword input by the user, and calculating the score of each search object in the search result based on the first ranking model. For the correlation with keywords, the obtained correlation value is used as the first ranking score.

Wherein, the first sorting score is divided into multiple intervals, and each search object is classified into the search object set corresponding to each interval according to the first ranking score, including: setting one or more correlation thresholds, and the second A ranking score corresponds to the correlation threshold and is divided into multiple intervals; each search object is classified into the search object set corresponding to the interval according to the interval to which the first ranking score belongs.

Wherein, the search object with the preset mark includes: based on the extended information of the search object and records related to the extended information, the preset mark is used to identify that the search object contains the extended information; using the second The sorting function adjusts the sorting of the search object with the preset mark in the search object set of the corresponding interval to which it belongs, including: returning the search result with the sorting adjustment to the user, and at the same time, returning the search object with the preset mark The extended information of the search object is returned to the user.

Wherein, obtaining the second ranking score of the search object with the preset mark based on the second ranking model includes: the second ranking model uses the record to calculate the second ranking for the search object with the preset mark score; use the second sorting score to adjust the ordering of the search object with the preset mark in the search object set of the corresponding interval to which it belongs, including: in the search object set corresponding to each interval, there is a preset mark The search object uses its second ranking score to determine its new ranking position, so as to adjust the ranking positions of all search objects in the search object set.

Another aspect of the present application provides a data search processing system, including: a first ranking module, which obtains the first ranking score of each search object in the search results based on the first ranking model; Divide into multiple intervals, classify each search object into the search object set corresponding to each interval according to the first sorting point; determine the module, determine the search object with a preset mark in the search object set corresponding to each interval; the second The second ranking score module obtains the second ranking score of the search object with the preset tag based on the second ranking model; the ranking adjustment module uses the second ranking score to adjust the search object with the preset tag in its belonging , The sorting in the search object collection corresponding to the interval.

Wherein, the first ranking sub-module includes: obtaining the search results according to the keywords input by the user, calculating the correlation between each search object and the keywords in the search results based on the first ranking model, and obtaining the correlation value as the first sorting score.

Wherein, the classification module includes: setting one or more correlation thresholds, dividing the first sorting points into multiple intervals corresponding to the correlation thresholds; Classify it into the search object set corresponding to the interval to which it belongs.

Wherein, the search object with a preset mark includes: based on the extended information of the search object and records related to the extended information, the preset mark is used to identify that the search object contains the extended information; the ranking adjustment module, It includes: returning the search results with sorting adjustment to the user, and at the same time, returning the extended information of the search object with the preset mark to the user.

Wherein, the second ranking module includes: the second ranking model uses the records to calculate the second ranking score for the search object with the preset mark; the ranking adjustment module includes: the search object corresponding to each interval In the object set, the search object with the preset mark uses its second ranking score to determine its new ranking position, so as to adjust the ranking positions of all the search objects in the search object set.

Compared with the prior art, according to the technical solution of this application, the unified search engine executes the search and returns the extended information based on the data object directly, avoiding the inconsistency of the user experience, and can ensure that both the relevance of the results and the Priority display for data objects with extended information. However, using the method of dividing intervals to adjust the sorting in a small range does not require complex algorithms and is simple to implement. Furthermore, the results of data objects and extended information based on data objects can be obtained directly through search engines, and consistent sorting rules are adopted, which not only achieves the optimization of data processing and the optimization of the corresponding data processing system, but also achieves a unified and effective user experience. The purpose of improving user experience.

Description of drawings

The drawings described here are used to provide a further understanding of the application and constitute a part of the application. The schematic embodiments and descriptions of the application are used to explain the application and do not constitute an improper limitation to the application. In the attached picture:

1A-1B are application schematic diagrams of existing data search processing technologies;

Figure 2 and Figure 3 are schematic diagrams showing the display effect of the output search results returned by the existing data search processing technology;

Fig. 4 is the flowchart of an embodiment of the data search processing method of the present application;

Fig. 5 is a schematic diagram of an embodiment of the application of the data search processing method of the present application;

6 is a flowchart of an embodiment of sorting search objects in an embodiment of the data search processing method of the present application;

Fig. 7 is a schematic diagram of the application of the preset mark in an embodiment of the data search processing method of the present application

Fig. 8 is a structural block diagram of an embodiment of the data search processing system of the present application.

detailed description

The main idea of this application is to match data objects and user query words during the search process, and obtain ranking points according to the correlation between data objects and user query words; set several correlation thresholds to divide the ranking points into several intervals, and the searched The data object (search object) is classified into the search object set of the corresponding interval; then, various records related to the extended information based on the data object are introduced as factors that affect the ranking of search objects, such as bidding factors. By using the influencing factors, you can use them in In the search object set corresponding to the correlation threshold interval to which each search object belongs, the ranking is adjusted. The two sorts are unified in one search process, which not only ensures the relevance of the search results, but also improves the consistency and continuity of the returned search results, and reduces complex algorithms such as shuffling and deduplication to solve the problem of reducing data processing Complexity, improving data processing efficiency, improving data search processing system performance, and simplifying the processing process can effectively improve system processing efficiency and computing performance, and also provide users with a good consistent experience. The application of this application, for example, in product search, can put the advertiser’s product-based advertising ideas into the product’s search engine, and only need to use the search engine in the search environment to return the search results and advertisement results uniformly, in the search results Direct commercialization can also improve the position of advertised products in search results on the premise of ensuring the relevance of search results, thereby effectively commercializing search engines.

In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and corresponding drawings. Apparently, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

According to an embodiment of the present application, a data search processing method is provided.

Referring to FIG. 4 , FIG. 4 is a flowchart 400 of an embodiment of the data search processing method of the present application.

At step S410, search results are obtained according to keywords input by the user.

FIG. 5 is a schematic diagram of an embodiment of the application of the data search processing method of the present application.

A user accesses a search platform, such as a commodity search platform, through a browser. The user can input query terms and send a search request on an open browser, such as inputting a product name and pressing a search button. The browser accesses the application server of the search platform, and the application server receives the search request. The application server requests the search engine to perform a search according to the current search request. The search engine uses the keywords obtained after the preprocessing of the query words to perform a search to obtain search results.

Among them, the search engine uses keywords to match the text descriptions of each data object in its general collection of all data objects. A text description with a high degree of correlation (correlation) of the keyword, so that it is determined that the data object corresponding to the text description is a searched result. These searched data objects are the data objects that match the keywords, and are used as search objects in the search results. The relevance here is search relevance or retrieval relevance.

Among them, some data objects have extended information based on the data object. For such search objects with extended information, they can be marked with preset marks so as to distinguish them from search objects without extended information. Further, the extended information is stored corresponding to the data object having the extended information. Further, various records related to data objects with extended information may also be stored.

Taking product search as an example, the user enters a keyword such as a product name through a browser, and presses the search button, the browser accesses the application server of the product search platform, and the product name is passed to the application server. Then the application server requests the search engine to execute this product search, such as through the similarity of the retrieval model (IR model based on algebra theory, IR model based on probability statistics, IR model based on set theory, machine learning model based on statistics, etc.) Calculate the degree, find all the product descriptions that match the keyword of this search, that is, the product name, and obtain the products (data objects) corresponding to these product descriptions. Among them, some products belong to advertising products, that is, the products also correspond to advertising ideas (extended information based on data objects). Such advertising products can be pre-marked to distinguish them from products without advertising creativity.

For example, FIG. 7 shows a schematic diagram of the application of preset marks in the embodiment of the data search processing method of the present application. Advertisers can manage their advertising products through the advertising system. Advertisers edit advertising ideas for the products they want to advertise (ie, advertising products) and place bids. Advertising products and corresponding bids will enter the search offline processing system in real time and merge with the original offline processing data. For example, mark the advertising products that are currently in normal delivery status in the search product collection, and record the corresponding advertising ideas and bids at the same time. bid. The data objects merged by the search offline processing system can be entered into a large collection of search engines and provided to the search engines for search services. These advertised items have preset flags if they are found by the search engine.

At step S420, the search objects in the obtained search results are sorted.

Specifically, the search engine may first sort the search results according to the search ranking logic, and then select the search objects in the search results that meet the conditions corresponding to the threshold according to the preset relevance threshold, and, for the search results in the For search objects with preset tags, adjust the sort order according to their records, and return the adjusted search results (for example, to the application server, etc.).

Among them, the selection of the relevance threshold needs to ensure that the search relevance is not affected. For example, the correlation between the keyword and the text description of the data object when the search engine is searched and a series of other factors are linearly combined to determine the numerical value (score) of the sort order of the searched data object (search object), that is, the ranking (for example, 1-100 points, 100 points are ranked first), and the correlation threshold is selected based on the scores of the linear combination. In this way, search objects whose scores are within a certain threshold can be reordered according to various situations such as those with preset tags, so as to avoid affecting the original search experience and take into account the actual needs of search objects with preset tags .

For example: when a product search engine sorts the results, it can determine the sorting order by linearly combining the keywords and text correlations calculated during the search and a series of commercial values (for example: sales in the past 30 days, return rate, etc.) This goes to select the correlation threshold.

Continuing from the aforementioned product search example, the search engine first calculates the similarity between the text description of the product and the user’s query word according to, for example, Boolean model, vector space model, probability model, language model or machine learning ranking model, that is, determines the correlation through similarity calculation. Assuming that the products A~I can be matched, and the products A~I are sorted, the obtained sequence is ABCDEFGHI, and each product has a sorting score. Then pre-set the correlation threshold as "20 points", and take the first 6 products A~F that satisfy this condition with a ranking score greater than or equal to 20 points and place them in the high-correlation score range, that is, the searches that meet the correlation threshold The result is commodity ABCDEF, and commodities G, H, and I that are less than "20 points" are classified into the low-correlation score range. Furthermore, in the interval greater than or equal to 20 points, for the advertised products C, E, and F with preset marks in the searched products A to F that are currently in normal delivery status, adjust C, E, and F according to their bid records, etc. The sorting order of is E, C, F, and then return the adjusted products in order and output them as EABCDF; similarly, in the interval of less than 20 points, for the searched products G, H, I with preset marks currently Advertisement products H and I in normal delivery status are adjusted to IGH according to their bid records and so on.

In one embodiment, referring to the flowchart of an embodiment of sorting search objects in an embodiment of the data search processing method of the present application shown in FIG. 6 (step S420), it is introduced to sort the search objects in the obtained search results Sorting and how sort adjustments are handled.

At step S610, the first ranking score of each search object in the search results is obtained based on the first ranking model.

Among them, the first sorting model is to find the data object (that is, the search object) through the similarity calculation of the retrieval model in the process of matching documents according to the keywords divided by the user query. The calculation of the similarity is to find out the correlation and the degree of correlation between the keyword and the data object, that is, to calculate the correlation. In one embodiment, the numerical value/score of the relevance of each search object can be obtained as a sorting score according to the similarity calculation; A linear combination of relevance and a range of other factors to determine the ranking score for each search object. Furthermore, a mathematical model for determining the ranking output of each search object according to the ranking score, or in other words, a search ranking logic for searching data objects and sorting search results (all searched data objects).

Wherein, the ranking score of each search object in the search results is obtained through the sorting operation of the first ranking model, which is called the first ranking score. The first ranking model can use language model, probability model, Boolean model, machine training model, etc. to calculate the ranking score of each search object.

Based on the above example of product search, for the sake of simplicity and clarity, only the numerical value/score of correlation obtained by similarity calculation is used for illustration. The commodity search engine will use the retrieval model to search and match according to the user's query (which can be divided into several keywords), such as using the Boolean model, vector space model, etc. to calculate the similarity between the query and the text description of each commodity The ranking scores of the searched commodities A˜I are the first ranking scores of each commodity in the search results obtained based on the first ranking model.

At step S620, the first ranking score is divided into a plurality of intervals, and each search object is classified into a search object set corresponding to each interval according to the first ranking score.

Dividing the first sorting score into a plurality of intervals may be performed by setting several correlation thresholds such as "20 points" and "10 points" in advance. For example, the first sorting score is divided into three intervals: interval 1 "greater than or equal to 20 points", interval 2 "greater than or equal to 10 points and less than 20 points", and interval 3 "less than 10 points". In this way, each interval is preset with a threshold, and the first ranking score of each search object can be compared with the threshold to determine whether the first ranking score falls into the interval. Once it falls, the first ranking score corresponds to the threshold. The search object can be included in the search object set corresponding to the interval.

Following the above product search example, among products A to I, assuming a preset threshold value of "20 points", the first sorting score is divided into two intervals, the first interval of greater than or equal to 20 points and the second interval of less than 20 points. The first ranking scores of commodities G, H, and I are 19, 18, and 17 in turn, and they will be classified into the corresponding commodity set II={G, H, I} in the second interval, where H and I are currently in normal delivery status Advertisement products; and the first ranking score of the product ABCDEF is from large to small and both are greater than 20, then it will be classified into the product set corresponding to the first interval I={A, B, C, D, E, F}, where C , E, and F are advertising products that are currently in normal delivery status.

At step S630, the search objects with preset marks in the search object set corresponding to each interval are determined.

In the set of search objects corresponding to each interval, there are one or more search objects, some of which also have preset tags, which can identify that the search object (that is, the data object) has data based on the extended information, and these extended information is normal and correct.

Following the product search example above, in the product set I={A, B, C, D, E, F}, C, E, and F are advertising products that are currently in normal delivery status, and the product set II={G, H, I} Among them, H and I are advertising products that are currently in normal delivery status. The default flag of an advertised product, which indicates that the product has a product-based ad creative and is in normal delivery status. For a clear and brief description, the product set I will be taken as an example below, and the advertised products C, E, and F can be found in the product set I through preset tags.

At step S640, a second ranking score of the search objects with preset marks is obtained based on a second ranking model.

In one embodiment, for search objects with preset tags in a section, the ranking score calculation is performed based on the second ranking model. Wherein, the second sorting model may also be a kind of sorting logic, and the second sorting model may be adjusted and designed according to actual needs, this is only an example, and the present application should not be understood as being limited thereto.

For example, each search object with preset tags includes extended information (extended information based on the data object), corresponding various records and other feature information, and various records and extended information can be used to design sorting rules or logic. The second sorting model, and get the second sorting score according to such rules or logic, for example: determine which characteristic information values, or which characteristic information are calculated to represent the sorting order of the search object, that is, the value As the second ranking score, the method of determining the value or calculating the value is the second ranking model.

For each interval, the second ranking score of the search object with the preset mark can be obtained based on the second ranking model, and the second ranking score is the adjusted ranking score of the search object.

Inheriting the above example of product search, take the advertised products C, E, and F in the product set I corresponding to the first interval as an example to place advertising ideas and pay for them using the pay-per-click (CPC) advertising model to illustrate the second ranking model. Acquisition of ranking points (including bids from advertisers, calculation of ranking points, advertising deduction, etc.). The adjustments for Commodity Set II of the second interval are similar. Here, for clear and concise description, only the commodity set I and the ranking of the adjusted first place are taken as an example.

Advertiser, that is, the owner who provides advertisement products C, E, F. Advertiser's bid, that is, the advertiser can bid for the advertised product to be displayed under a certain query word/keyword. The bid is recorded, that is, mark the advertised product in the searched product set that is currently in a normal delivery state as described in step S410, and record the corresponding advertising creative, the bid displayed under the keyword, and the advertising quality of the collected advertised product score etc. Table 1 shows advertisers' bids for advertised commodities, and Table 2 shows the advertisement quality scores of advertised commodities.

Table 1:

commodity Sort Adjusted Bids C 1 Yuan E. 1.5 yuan f 0.8 yuan

Table 2:

commodity quality score C 60 E. 50 f 30

Among them, the sorting logic of the design is the sorting formula: expected return (second ranking score) = bid * quality score. In addition, the deduction logic can also be designed, that is, the deduction formula is: actual deduction = next bid * next quality score / quality score + 0.01. Then the second ranking score/expected income of advertising products C, E, and F are obtained in order: 1*60=60, 1.5*50=75, 0.8*30=24, and the order of the three is 2, 1, 3. The actual deduction of advertising products C, E, and F is as follows: 24/60+0.01=0.41, 60/50+0.01=1.21, 0.8 (the last digit will be deducted according to its bid price). The second ranking score and the actual deduction are calculated as shown in Table 3.

table 3:

commodity second ranking score to sort Actual deduction C 60 2 0.41

E. 75 1 1.21 f twenty four 3 0.8

In addition, other methods can be used for advertiser bidding, second ranking score calculation (for subsequent search engine ranking adjustment), advertising fee deduction, etc., without affecting the core of the application. For example: use the cost per thousand (CPM) advertising model to place and pay for advertising ideas, and can also directly use the advertiser's bid for thousand impressions to calculate the second ranking score, and so on.

At step S650, the second ranking score is used to adjust the ranking of the search object with the preset mark in the search object set of the corresponding interval to which it belongs.

In the set of search objects corresponding to each interval, according to the needs and according to certain rules, the second sorting score obtained by the second row model can be used for the search objects with preset marks to adjust the corresponding intervals of the respective intervals. The sort order of search objects with preset tags in the search object collection.

For example, the second ranking score of a certain search object is its new ranking score, which is compared with the ranking scores of other search objects in the search object set of the corresponding interval to which it belongs. It is adjusted to the frontmost position (first position) of the set; for example, the second ranking score of a certain search object is compared with the second ranking score of other search objects with preset marks in the search object set of the corresponding interval to which it belongs Largest, the search object with the preset mark is prioritized and only the first position is adjusted, and it can be adjusted to the front position of the collection, and the other search objects are sorted from large to small; and so on.

Wherein, the adjustment of the foremost position (the first position) is performed in the above-mentioned way, and the adjustment of the second, third and fourth positions can be done by analogy according to the above example.

Following the example of product search above, in the product set I corresponding to the first section, the products A to F are sorted in the ABCDEF manner. From the second sorting scores calculated above, it can be known that among the advertised products C, E, and F, the product E ranks second. The ranking score is 75, the highest, product C is 60 points, and product F is the lowest, 24 points.

In one case, according to the rule that the advertised product is prioritized and only the first position is adjusted, the product E can be placed at the first position of the product set I. Then the adjusted sorting of each commodity in the commodity set I corresponding to the first interval is EABCDF.

In another situation, if there is a click on the advertisement and you want to charge all the advertisement fees, you can use the second sorting method for all the advertised commodities C, E, and F in the commodity collection of each interval, such as the commodity collection I in the first interval, according to the second ranking. Adjust them all. Example 1: If the advertised product has absolute priority, the final ranking can be adjusted to ECFABD according to the aforementioned second ranking points. Example 2: If it is stipulated that each advertised product maintains the original order of the first ranking score, that is, the order of CEF, then combined with the adjustment of the second ranking score, for example, it is stipulated to move forward by at most one-tenth and take the entire position, and the advertised product E is the most Can be adjusted forward by 7 digits (75/10 and rounded), C can be adjusted forward by up to 6 digits (60/10 and rounded), F can be adjusted forward by up to 2 digits (24/10 and rounded), according to the original ABCDEF The sorting is adjusted to CEABFD according to the first 6 digits of C, the first 7 digits of E, and the first 2 digits of F. Example 3: If it is agreed that the first ranking score and the second ranking score are superimposed to adjust the final ranking score, assuming that the first ranking score of the product ABCDEF is: 120, 100, 50, 40, 30, 10, according to the adjustment of the second ranking score (addition), the final sorting points are: 120, 100, 110, 40, 105, 34, and the sorting adjustment is: ACEBDF.

At step S430, the sorted search results are returned to the user.

Specifically, the application server obtains the sorted search results from the search engine, that is, the adjusted sorted search results, renders the browser page, and returns the search results to the browser. Each search object in the search results will be displayed in the order specified by the sorting on the browser. Moreover, for a search object with a preset tag, it will also return extended information based on the data object along with the search object (data object).

Following the above product search example, in the product set I corresponding to the first section, the application server obtains the sorted product results from the product search engine, renders the browser page, and returns the product results to the browser. Commodities A to F will be displayed to the user in the sort order of EABCDF on the browser. At the same time, the advertising creatives of the advertising products E, C, and F are returned to the user along with the advertising products E, C, and F. Further, if the user is interested in the advertisement idea of product E, and clicks on product E, the advertiser will deduct 1.21 yuan as shown in Table 3.

Fig. 8 schematically shows a structural block diagram of an embodiment of a data search processing system according to the present application.

According to an embodiment of the present application, the system 800 may include: a search module 810, which obtains search results according to the keywords input by the user, and the specific functions may refer to the processing described in step S410; a sorting module 820, which sorts the obtained search results Each search object is sorted, and the specific realized functions can refer to the processing described in step S420; the output module 830 returns the sorted search results to the user, and the specific realized functions can be referred to the processing described in step S430.

Wherein, the ranking module 820 also includes: a first ranking score module (not shown), which obtains the first ranking score of each search object in the search results based on the first ranking model. For the specific functions, please refer to the processing described in step S610; The classification module (not shown) divides the first sorting score into multiple intervals, and classifies each search object into the search object set corresponding to each interval according to the first sorting score. For specific functions, please refer to the step The processing described in S620; the determination module (not shown), which determines the search object with a preset mark in the search object set corresponding to each interval, the specific functions can refer to the processing described in step S630; the second sorting sub-module (not shown shown), based on the second ranking model to obtain the second ranking score of the search object with the preset mark, the specific implementation function can refer to the processing described in step S640; the ranking adjustment module (not shown), using the second The sorting component adjusts the sorting of the search object with the preset flag in the search object set of the corresponding section to which it belongs, and the specific function may refer to the processing described in step S650.

Since the processing and functions realized by the system of this embodiment basically correspond to the method embodiments shown in the above-mentioned Figures 1 to 7, for the details not described in this embodiment, you can refer to the relevant descriptions in the previous embodiments. I won't go into details here.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

Memory may include non-permanent storage in computer-readable media, in the form of random access memory (RAM) and/or nonvolatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer readable media.

Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can be implemented by any method or technology for storage of information. Information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Flash memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, Magnetic tape cartridge, tape magnetic disk storage or other magnetic storage device or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media excludes non-transitory computer-readable media, such as modulated data signals and carrier waves.

It should also be noted that the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus comprising a set of elements includes not only those elements, but also includes Other elements not expressly listed, or elements inherent in the process, method, commodity, or apparatus are also included. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The above description is only an embodiment of the present application, and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of the claims of the present application.

Claims (10)

1. a data search disposal route, is characterized in that, comprising:

The first sequence point of each object search in Search Results is obtained based on the first order models;

This first sequence graduation is divided into multiple interval, each object search is referred in each interval corresponding object search set according to this first sequence point;

Determine, in each interval corresponding object search set, there is the object search presetting mark;

There is described in obtaining based on the second order models the second sequence point of the object search presetting mark;

Utilize the sequence of object search in its corresponding interval object search set having described in this second sequence point adjustment and preset mark.

2. the method for claim 1, is characterized in that, obtains the first sequence point of each object search in Search Results, comprising based on the first order models:

Keyword according to user's input obtains described Search Results, calculates the correlativity of each object search and keyword in Search Results based on described first order models, using the relevance values obtained as the first sequence point.

3. method as claimed in claim 1 or 2, is characterized in that, this first sequence graduation is divided into multiple interval, is referred to by each object search in each interval corresponding object search set, comprises according to this first sequence point:

One or more relevance threshold is set, by this first sequence point corresponding described relevance threshold, is divided into multiple interval;

Each object search, according to its first sequence point affiliated interval, is referred in this affiliated interval corresponding object search set.

4. the method as described in one of claim 1-3, is characterized in that,

Described have the object search presetting mark, and comprising: based on extend information and the record relevant to extend information of this object search, described default mark comprises described extend information for identifying this object search;

Utilize the object search that to have described in this second sequence point adjustment and preset mark belonging to it, sequence in corresponding interval object search set, comprise: the Search Results completing sequence adjustment is returned to user, meanwhile, the described extend information with the object search presetting mark is returned to user.

5. method as claimed in claim 4, is characterized in that,

There is described in obtaining based on the second order models the second sequence point presetting the object search marked, comprising: described second order models utilizes described record, to described, there is object search calculating second sequence point presetting mark;

Utilize the object search that to have described in this second sequence point adjustment and preset mark belonging to it, sequence in corresponding interval object search set, comprising:

In each interval corresponding object search set, there is the object search presetting mark and utilize its second sequence point, determine its new sorting position, to adjust the sorting position of object searches all in this object search set.

6. a data search disposal system, is characterized in that, comprising:

First sequence sub-module, obtains the first sequence point of each object search in Search Results based on the first order models;

Classifying module, is divided into multiple interval by this first sequence graduation, is referred to by each object search in each interval corresponding object search set according to this first sequence point;

Determination module, determines to have the object search presetting mark in each interval corresponding object search set;

Second sequence sub-module, has the second sequence point of the object search presetting mark based on the second order models described in obtaining;

Sequence adjusting module, utilize the object search that to have described in this second sequence point adjustment and preset mark belonging to it, sequence in corresponding interval object search set.

7. system as claimed in claim 6, is characterized in that, the first sequence sub-module, comprising:

Keyword according to user's input obtains described Search Results, calculates the correlativity of each object search and keyword in Search Results based on described first order models, using the relevance values obtained as the first sequence point.

8. system as claimed in claims 6 or 7, is characterized in that classifying module comprises:

One or more relevance threshold is set, by this first sequence point corresponding described relevance threshold, is divided into multiple interval;

Each object search, according to its first sequence point affiliated interval, is referred in this affiliated interval corresponding object search set.

9. the system as described in claim 6-8, is characterized in that,

Described have the object search presetting mark, and comprising: based on extend information and the record relevant to extend information of this object search, described default mark comprises described extend information for identifying this object search;

Sequence adjusting module, comprising: the Search Results completing sequence adjustment is returned to user, meanwhile, the described extend information with the object search presetting mark is returned to user.

10. system as claimed in claim 9, is characterized in that,

Second sequence sub-module, comprising: described second order models utilizes described record, has to described object search calculating second sequence point presetting mark;

Sequence adjusting module, comprising: in each interval corresponding object search set, has the object search presetting mark and utilizes its second sequence point, determine its new sorting position, to adjust the sorting position of object searches all in this object search set.