CN105989156B - Method, equipment and system for providing search results - Google Patents
Method, equipment and system for providing search results Download PDFInfo
- Publication number
- CN105989156B CN105989156B CN201510094491.5A CN201510094491A CN105989156B CN 105989156 B CN105989156 B CN 105989156B CN 201510094491 A CN201510094491 A CN 201510094491A CN 105989156 B CN105989156 B CN 105989156B
- Authority
- CN
- China
- Prior art keywords
- category
- information
- query sequence
- determining
- search results
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
the application aims to provide a method and a system for providing search results. Compared with the prior art, the method comprises the steps of firstly determining first category optimization information and second category optimization information corresponding to a query sequence provided by user equipment; determining category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information; simultaneously obtaining one or more search results corresponding to the query sequence; and determining the priority information of the search result based on the matching degree information of the category information of the search result and the category prediction information. And displaying the one or more search results according to the priority information of the search results. The method and the device can effectively solve the problem that the search result is far from the intention expressed by the query sequence input by the user.
Description
Technical Field
The present application relates to the field of computers, and more particularly, to a technique for providing search results.
background
The category system of the classified search engine is divided in a mode of gradually expanding from total to divided. The settings of the category systems are also different due to the differences in resource characteristics and user requirements. For example, in an online shopping platform, the category system of its search engine is generally set according to the type of goods. However, when the user inputs the query word "mobile phone" to search, the search engine may query various items in the category of mobile phone and the category of mobile phone charger. From the above, the search results queried under the two categories are completely different. Therefore, it is difficult to distinguish the user's intention only by the correlation between the query words and categories. Therefore, improvements to the query mechanism of existing search engines are needed.
disclosure of Invention
the application aims to provide a method, equipment and a system for providing search results.
according to an aspect of the present application, a method for providing a search result at a first network device is provided, where the method includes:
obtaining one or more search results corresponding to a query sequence sent by user equipment;
Acquiring category prediction information corresponding to the query sequence;
determining priority information of the search result based on matching degree information of category information of the search result and the category prediction information;
Providing at least one of the one or more search results to the user device according to the priority information of the search result.
According to another aspect of the present application, there is also provided a method for providing search results at a user equipment, wherein the method includes:
Sending the query sequence to the corresponding second network equipment;
Receiving category prediction information corresponding to the query sequence returned by the second network device;
Sending the query sequence and the category prediction information to corresponding first network equipment;
and receiving one or more search results corresponding to the query sequence returned by the first network equipment.
According to another aspect of the present application, a method for determining category prediction information corresponding to a query sequence at a second network device is further provided, where the method includes:
Receiving a query sequence sent by user equipment;
Determining first category optimization information and second category optimization information corresponding to the query sequence;
determining category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information;
And sending the category prediction information to the user equipment.
According to still another aspect of the present application, there is also provided a first network device for providing search results, including:
A result obtaining device, configured to obtain one or more search results corresponding to a query sequence sent by a user equipment;
The category acquisition device is used for acquiring category prediction information corresponding to the query sequence;
Priority determining means for determining priority information of the search result based on matching degree information of category information of the search result and the category prediction information;
Providing means for providing at least one of the one or more search results to the user equipment in accordance with the priority information of the search result.
According to yet another aspect of the present application, there is also provided a user equipment for providing search results, including:
The first sending device is used for sending the query sequence to the corresponding second network equipment;
a first receiving device, configured to receive category prediction information corresponding to the query sequence returned by the second network device;
A third sending device, configured to send the query sequence and the category prediction information to a corresponding first network device;
and a third receiving device, configured to receive one or more search results corresponding to the query sequence, where the search results are returned by the first network device.
According to another aspect of the present application, there is provided a second network device for determining category prediction information corresponding to a query sequence, including:
a second receiving device, configured to receive a query sequence sent by a user equipment;
The second determining device is used for determining the first category optimization information and the second category optimization information corresponding to the query sequence;
a fourth determining device, configured to determine category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information;
And the second sending device is used for sending the category prediction information to the user equipment.
According to yet another aspect of the present application, there is also provided a system for providing search results, wherein the system includes at least two of a first network device according to one aspect of the present application, a user device according to another aspect of the present application, and a second network device according to yet another aspect of the present application.
compared with the prior art, the method and the device have the advantages that the category prediction is carried out on the query sequence provided by the user equipment, the priority of each search result is set according to the matching degree of the category to which each search result corresponding to the query sequence belongs and the predicted category, and the problem that the search result is far from the will expressed by the query sequence input by the user can be effectively solved. In addition, according to the mapping relation between the query words and the categories and the category hierarchical structure information corresponding to the query sequence, the first category optimization information corresponding to the query sequence is determined, or according to the mapping relation between the query words and the categories and the modified query sequence corresponding to the query sequence, the second category optimization information corresponding to the query sequence is determined, so that more accurate category prediction information is obtained, the Martian effect caused by category prediction based on a click ratio is effectively inhibited, the information obtaining efficiency of a user is further improved, and the screen utilization rate of user equipment is improved.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:
FIG. 1 illustrates a schematic diagram of a system for providing search results in accordance with an aspect of the subject application;
FIG. 2 is a diagram illustrating a second network device in a system for providing search results in accordance with a preferred embodiment of the present application;
FIG. 3 illustrates a flow diagram of a method for providing search results according to another aspect of the subject application;
FIG. 4 illustrates a flowchart of a method for providing search results in accordance with a preferred embodiment of the present application.
The same or similar reference numbers in the drawings identify the same or similar elements.
Detailed Description
The present application is described in further detail below with reference to the attached figures.
In a typical configuration of the present application, the terminal, the device serving the network, and the trusted party each include one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
the memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The 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 Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, 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 does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.
FIG. 1 illustrates a system for providing search results in accordance with an aspect of the subject application. The system 1 comprises a first network device 13, a second network device 12 and a user device 11. Wherein the first network device 13 comprises: result acquisition means 131, category acquisition means 132, priority determination means 133, providing means 134. The second network device 12 includes: second receiving means 121, second determining means 122, fourth determining means 123, and second transmitting means 124. The user equipment 11 includes: a first transmitting device 111, a first receiving device 112, a third transmitting device 113, and a third receiving device 114.
In particular, the first sending means 111 is configured to send the query sequence to the corresponding second receiving means 121, and the second receiving means 121 delivers the query sequence to the second determining means 122. The second determining device 122 is configured to determine the first category optimization information and the second category optimization information corresponding to the query sequence. Then, the fourth determining device 123 determines the category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information. The second sending device sends the category prediction information to the first receiving device 112, and the first receiving device 112 transfers the category prediction information to the third sending device 113, so that the third sending device 113 sends the query sequence and the category prediction information to the first network device 13. The first network device 13 distributes the query sequence and the category prediction information to the result obtaining device 131 and the category obtaining device 132, respectively. Wherein the result obtaining means 131 further obtains one or more search results corresponding to the query sequence. Then, the priority determining device 133 determines the priority information of the search result based on the matching degree information of the category information of the search result and the category prediction information. Then, the providing device 134 provides at least one of the one or more search results to the third receiving device 114 according to the priority information of the search result, so that the user equipment 11 presents the search result received by the third receiving device 114 to the user.
here, the first network device 13 and the second network device 12 may be the same network device or different network devices based on communication connection. Here, the first network device 13 and the second network device 12 may be implemented by a network host, a single network server, a cloud formed by a plurality of network server sets or a plurality of servers, and the like. Here, the Cloud is made up of a large number of hosts or web servers based on Cloud Computing (Cloud Computing), which is a type of distributed Computing, a super virtual computer consisting of a collection of loosely coupled computers. It should be understood by those skilled in the art that the first network device 13 and the second network device 12 are only examples, and other existing or future network devices may be included within the scope of the present application, as applicable, and are herein incorporated by reference.
here, the user device 11 includes an electronic device capable of automatically performing numerical calculation and information processing according to instructions set or stored in advance, and hardware thereof includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a programmable gate array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.
When a user inputs a query sequence containing query terms and categories in a search bar through a human-computer interaction device and submits the query sequence, the first sending device 111 submits the query sequence to the second receiving device 121 through communication protocols such as http, https and the like, and the query sequence is transmitted to the second determining device 122 through the second receiving device 121, so that first category optimization information and second category optimization information corresponding to the query sequence are determined.
Here, the category refers to a node set in a search engine to help find a search result related to the query term. The search engine comprises a tree-based multi-level system structure category, and the category to which each search result belongs corresponds to a node in the system structure. For example, a node of the category "women's dress" might include the following: the categories of 'one-piece dress', 'sweater', 'coat' and the like are used as sub-nodes; the node of the category "one-piece dress" may further include below: the categories of the thick one-piece dress, the thin one-piece dress and the like are used as sub nodes. The root node in the architecture may be an empty node (i.e., without meaning).
here, the first category optimization information and the second category optimization information may be determined according to a category in the query sequence and a hierarchical structure under the category.
specifically, the second determining device 122 determines the first category optimization information and the second category optimization information corresponding to the received query sequence according to two preset category optimization algorithms. The two category optimization algorithms can determine categories corresponding to the query terms and first category optimization information and second category optimization information of each category under the hierarchical structure of the categories through analysis of search log information.
For example, the second determining device 122 determines each category in the hierarchical structure under the received category according to the category architecture, calculates click rates of search results matching the received query word in the received category and each category under the hierarchical structure in the received category in the search log information, and calculates click probabilities corresponding to each category according to the counted click rates corresponding to each category. Next, a category optimization algorithm executed by the second determining device 122 is: and grouping the obtained probabilities according to a preset probability interval, and homogenizing the probability of each group of categories, thereby obtaining first category optimization information containing the categories and the corresponding homogenized probabilities. Another category optimization algorithm executed by the second determining device 122 is to perform an open-cube process and a normalization process on the obtained probabilities, thereby obtaining second category optimization information including the categories and the corresponding normalized probabilities.
It should be noted that the algorithm for determining the first category optimization information and the second category optimization information is not limited to this. In fact, the determination algorithms of the first category optimization information and the second category optimization information may be used interchangeably. Those skilled in the art will appreciate that the algorithm for determining the first category optimization information and the second category optimization information is merely exemplary, and other existing or future network devices, as may be suitable for the present application, are also included within the scope of the present application and are hereby incorporated by reference.
Preferably, the second network device 12 further comprises: statistics apparatus 125 (shown in fig. 2). The statistical device 125 is configured to perform statistical processing on the search log information to obtain a mapping relationship between the query term and the category. Correspondingly, the second determining device 122 is configured to determine the first category optimization information and the second category optimization information corresponding to the query sequence according to the mapping relationship between the query term and the category.
specifically, the statistical device 125 performs statistical processing according to the query term in the search log information, the clicked search result, and the category to which the clicked search result belongs, so as to obtain a mapping relationship between the query term and at least one category, and store the mapping relationship. When the second determining device 122 obtains a query sequence input by a user, the second determining device 122 determines a category corresponding to a query term in the query sequence according to the mapping relationship provided by the statistical device 125, and determines first category optimization information and second category optimization information corresponding to the query sequence according to the category determined by the mapping relationship and the category in the query sequence.
Here, the manner of determining the first category optimization information and the second category optimization information corresponding to the query sequence by the second determining device 122 according to the obtained categories may be as described above. Preferably, the second determining means 122 includes: the first determining unit is used for determining the first category optimization information or the second determining unit is used for determining the second category optimization information. (none are shown in the drawings)
The first determining unit is configured to determine first category optimization information corresponding to the query sequence according to the mapping relationship between the query term and the category hierarchical structure information corresponding to the query sequence.
it should be noted that the first determining unit may determine the first category optimization information corresponding to the query sequence by using the described algorithms. The first category optimization information may also be determined in the following manner.
It should also be noted that the second determining means 122 may only provide the first category optimization information to the fourth determining means 123 when the user inputs the query sequence for the first time. When the user modifies the query term, the second determining unit is used to determine second category optimization information.
specifically, the first determining unit determines a category corresponding to the query term according to the mapping relationship, and determines category hierarchical structure information of the corresponding category according to a category architecture. Then, the first determining unit determines a category corresponding to the query term, a category in the query sequence, and the number of clicks of each category in the hierarchical structure information of each category from a search log. And a score for each category is obtained using equation 1.
Wherein, query is a query word, catx is a category obtained according to the mapping relationship or a category (which may be a root node category) in the query sequence, and c is a category in the category hierarchical structure information. After a user specifies query or even catx during searching, the set of categories to which the search result belongs is clicked: ca (query, catx), where the number of clicks of category c is click.
then, the first determination unit calculates the ranking score f from the leaf category to the root node category hierarchically in the category architecturecact_refine(query, catx, c) is shown in equation 2.
Equation 2
wherein the parameter beta controls the influence of the click ratio between the parent and child categories on the final effect. The sub-set of catx is subcatx, and i is one of the sub-sets.
thus, the first determining unit obtains the first category optimization information including the ranking scores of the categories corresponding to the query terms.
and the second determining unit is used for determining second category optimization information corresponding to the query sequence according to the mapping relation between the query terms and the categories and the modified query sequence corresponding to the query sequence.
here, when the user is not expected to face the current search result, the query term may be modified for optimization. The modified query sequence corresponds to a query sequence of query terms and/or categories, etc. modified by the user.
for example, when a user searches for "matchsticks" (which are expected to be an outdoor brand), if matchsticks for ignition are returned, the user will change the query word to "matchstick outdoor".
When the modified query term queryb is a child intention string of the query term querya before modification, then let the set of child intention strings of querya be: sub _ querya. The second determining unit determines second category optimization information of the query sequence including the query word querya using equation 3.
Wherein the parameter gamma controls the influence of the click ratio between the parent and child categories on the final effect. Subcatx is a subset of catx.
Thus, the second determining unit obtains second category optimization information including the ranking scores of the categories corresponding to the query terms.
Then, the second determining means 122 provides the obtained first category optimization information and second category optimization information to the fourth determining means 123.
The fourth determining device 123 determines the category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information.
specifically, the fourth determining device 123 may determine the category prediction information of the query sequence by comprehensively evaluating the probability or score of each category in the first category optimization information and the second category optimization information.
for example, the probability of the category a1 in the first category optimization information is p1, the probability of the category a2 in the first category optimization information is p2, the probability of the category a1 in the second category optimization information is p1 ', and the probability of the category a2 in the second category optimization information is p 2'. The fourth determining device 123 first normalizes the probabilities p1, p2, p1 ', and p 2' according to the number of categories, and thus obtains the category prediction information of the query sequence, including: category a1 and the corresponding probability are:Category a2 and the corresponding probability are:
preferably, the fourth determining device 123 updates the category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information in a weighting manner.
for example, the fourth determining device 123 uses equation 4 to perform weighted update on the first category optimization information and the second category optimization information, and obtains category prediction information corresponding to the query sequence.
and the parameters a and b respectively control the influence of the first category optimization information and the second category optimization information on the final prediction effect.
With equation 4, the obtaining of the category prediction information by the fourth determining means 123 includes: each category in the first category optimization information and the second category optimization information and the corresponding weighted score ffinal(query,catx,c)。
The fourth determining means 123 passes the obtained category prediction information to the second transmitting means and transmits it to the first receiving means 112. The first receiving device 112 then sends the category prediction information and the query sequence to the first network device 13 through the third sending device 113. The first network device 13 queries at least one search result corresponding to the query sequence, provides each search result to the result obtaining device 131, and provides the obtained category prediction information to the category obtaining device 132. The result acquisition means 131 and the category acquisition means 132 supply the acquired search result and the category prediction information to the priority determination means 133, respectively.
It should be noted that, the fourth determining device 123 may also directly send the category prediction information to the category obtaining device 132 in the first network device 13.
The priority determining means 133 determines the priority information of the search result based on the matching degree information of the category information of the search result and the category prediction information.
Specifically, the priority determining device 133 matches the category information of each search result with the categories in the category prediction information, determines the priority information of the category information of each search result according to complete agreement, partial agreement, and disagreement, and associates the determined priority information with each search result.
preferably, the priority determining device 133 determines the priority information of the search result based on the matching degree information and the content relevance information of the search result and the query sequence.
Specifically, the priority determining device 133 performs weighting processing on preset matching degree information and content relevance information to determine priority information of the search result.
For example, the search results include: search result b1, search result b2, and search result b 3. The category corresponding to the search results b1 and b2 is category c1, and the category corresponding to the search result b3 is c 2. Categories in the category prediction information include: c1, c 3. The query sequence includes: the query term "aabb".
The priority determining means 133 obtains the priority information Y1 of the search results b1 and b2 and the priority information Y2 of the search result b3 by matching categories, wherein Y1> Y2. Meanwhile, character matching is carried out to obtain a search result b1 containing 'aabb', and if complete matching is carried out, the content correlation information is determined to be Y3; if the search result b2 includes "aa", partial matching determines that the content relevance information is Y4, and if the search result b3 includes "bb", partial matching determines that the content relevance information is Y4, wherein Y3> Y4.
The priority determining device 133 obtains the priority information of the search result b1 as (a 1% × Y1+ a 2% × Y3), the priority information of the search result b2 as (a 1% × Y1+ a 2% × Y4), and the priority information of the search result b3 as (a 1% × Y2+ a 2% × Y4) according to the weight of the preset category matching degree as a 1% and the weight of the content relevance as a 2%, where a1> a 2.
The priority determining means 133 may also determine the priority information of each search result according to the score or probability of each category in the category prediction information as a weight.
Continuing with the above example, the priority determining means 133 determines that the priority information of the search result b1 is ffinal_c1(a 1% Y1+ a 2% Y3), the priority information of the search result b2 being ffinal_c1(a 1% Y1+ a 2% Y4), the priority information of the search result b3 being ffinal_c3(a 1% Y2+ a 2% Y4). Wherein f isfinal_c1Predicting a score, f, for category c1 in the information for the categoryfinal_c3Scores of the categories that are not matched with the categories in the category prediction information, which are preset for the priority determination means 133.
It should be noted that the manner of determining the priority information of the search results is only an example, and in fact, the priority determining device 133 may also determine the priority information of each search result by means of weighted average and the like. In addition, other existing or future manners of determining priority information for the search results, as applicable to the present application, are also intended to be encompassed within the scope of the present application and are hereby incorporated by reference.
After determining the priority information of each search result, the priority determining device 133 provides each search result and the corresponding priority information to the third receiving device 114 through the providing device 134, and the third receiving device 114 presents at least one search result to the user.
Here, the providing device 134 sends all the search results and the corresponding priority information to the third receiving device 114, and the third receiving device 114 displays the search results in a whole or in a paginated order according to the screen size of the user equipment 11 where the search results are located.
Preferably, the providing device 134 performs a sorting process on at least one of the one or more search results according to the priority information of the search result; and then provides the corresponding sorting result to the user equipment 11.
Specifically, the providing device 134 provides the one or more search results to the third receiving device 114 in the order of the priority information of each search result from high to low.
Here, the providing device 134 may provide all search results in one page to the third receiving device 114. Partial search results may also be provided to the third receiving device 114 through a paging technique.
FIG. 3 illustrates a method for providing search results, the method being performed primarily by the system shown in FIG. 1, according to one aspect of the subject application. The system includes a first network device, a second network device, and a user device. Wherein the first network device performs steps S6, S7, S8. The second network device performs steps S2, S3, S4. The user equipment performs steps S1, S5, S9.
Specifically, in step S1, the user equipment sends a query sequence to the second network device. In step S2, the second network device determines the first category optimization information and the second category optimization information corresponding to the query sequence. Next, in step S3, the second network device determines category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information. In step S4, the second network device sends the category prediction information to the user device. In step S5, the ue sends the query sequence and the category prediction information to the first network device. In step S6, the first network device also obtains one or more search results corresponding to the query sequence. In step S7, the first network device determines priority information of the search result based on matching degree information of category information of the search result and the category prediction information. In step S8, the first network device provides at least one of the one or more search results to the user device according to the priority information of the search result. In step S9, the user device presents the received search results to the user.
here, the first network device and the second network device may be the same network device or different network devices based on communication connection. Here, the first network device and the second network device may be implemented by a network host, a single network server, a cloud formed by a plurality of network server sets or a plurality of servers, and the like. Here, the cloud is made up of a large number of hosts or web servers based on cloud computing (CloudComputing), which is a type of distributed computing, a super virtual computer consisting of a collection of loosely coupled computers. Those skilled in the art should understand that the first network device and the second network device are only examples, and other existing or future network devices may be applicable to the present application, and are included in the scope of the present application and are incorporated by reference herein.
Here, the user equipment includes an electronic device capable of automatically performing numerical calculation and information processing according to instructions set or stored in advance, and hardware thereof includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a programmable gate array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.
When a user inputs a query sequence containing query words and categories in a search bar through a human-computer interaction device and submits the query sequence, the user equipment submits the query sequence to the second network equipment through communication protocols such as http, https and the like. And the second network equipment determines the first category optimization information and the second category optimization information corresponding to the query sequence.
Here, the category refers to a node set in a search engine to help find a search result related to the query term. The search engine comprises a tree-based multi-level system structure category, and the category to which each search result belongs corresponds to a node in the system structure. For example, a node of the category "women's dress" might include the following: the categories of 'one-piece dress', 'sweater', 'coat' and the like are used as sub-nodes; the node of the category "one-piece dress" may further include below: the categories of the thick one-piece dress, the thin one-piece dress and the like are used as sub nodes. The root node in the architecture may be an empty node (i.e., without meaning).
here, the first category optimization information and the second category optimization information may be determined according to a category in the query sequence and a hierarchical structure under the category.
Specifically, the second network device determines first category optimization information and second category optimization information corresponding to the received query sequence according to two preset category optimization algorithms. The two category optimization algorithms can determine categories corresponding to the query terms and first category optimization information and second category optimization information of each category under the hierarchical structure of the categories through analysis of search log information.
for example, the second network device determines each category in the hierarchical structure under the received category according to the category architecture, calculates click rates of search results, which are matched with the received query word, in the received category and each category under the hierarchical structure thereof in the search log information, and calculates click probabilities corresponding to each category according to the counted click rates corresponding to each category. Next, a category optimization algorithm executed by the second network device is: and grouping the obtained probabilities according to a preset probability interval, and homogenizing the probability of each group of categories, thereby obtaining first category optimization information containing the categories and the corresponding homogenized probabilities. And performing cubic processing and normalization processing on the obtained probabilities to obtain second category optimization information containing the categories and the corresponding probabilities after the normalization processing.
It should be noted that the algorithm for determining the first category optimization information and the second category optimization information is not limited to this. In fact, the determination algorithms of the first category optimization information and the second category optimization information may be used interchangeably. Those skilled in the art will appreciate that the algorithm for determining the first category optimization information and the second category optimization information is merely exemplary, and other existing or future network devices, as may be suitable for the present application, are also included within the scope of the present application and are hereby incorporated by reference.
preferably, the second network device further performs step S10 (shown in fig. 4) before performing step S2. In step S10, the second network device performs statistical processing on the search log information to obtain a mapping relationship between the query term and the category. And then, the second network equipment determines the first category optimization information and the second category optimization information corresponding to the query sequence according to the mapping relation between the query word and the category.
Specifically, the second network device performs statistical processing according to the query word, the clicked search result, and the category to which the clicked search result belongs in the search log information, thereby obtaining a mapping relationship between the query word and at least one category, and storing the mapping relationship. When the second network device obtains a query sequence input by a user, determining a category corresponding to a query word in the query sequence according to the mapping relation, and determining first category optimization information and second category optimization information corresponding to the query sequence according to the category determined through the mapping relation and the category in the query sequence.
here, the manner of determining, by the second network device, the first category optimization information and the second category optimization information corresponding to the query sequence according to the obtained categories may be as described above. Preferably, the step S2 includes: step S21 or step S22. (none are shown in the drawings)
in step S21, the second network device determines, according to the mapping relationship between the query term and the category hierarchical structure information corresponding to the query sequence, first category optimization information corresponding to the query sequence.
It should be noted that the second network device may determine the first category optimization information corresponding to the query sequence by using the described algorithms. The first category optimization information may also be determined in the following manner.
It should be further noted that the second network device may determine only the first category optimization information when the user inputs the query sequence for the first time. When the user modifies the query term, the second category optimization information is determined using the step S22.
Specifically, the second network device determines a category corresponding to the query term according to the mapping relationship, and determines category hierarchical structure information of the corresponding category according to a category architecture. Then, the second network device determines the category corresponding to the query word, the category in the query sequence, and the click times of the categories in the hierarchical structure information of the categories from the search log. And a score for each category is obtained using equation 1.
wherein, query is a query word, catx is a category obtained according to the mapping relationship or a category (which may be a root node category) in the query sequence, and c is a category in the category hierarchical structure information. After a user specifies query or even catx during searching, the set of categories to which the search result belongs is clicked: ca (query, catx), where the number of clicks of category c is click.
Then, the second network device re-calculates the ranking score f from leaf category to root node category hierarchically in the category architecturecact_refine(query, catx, c) is shown in equation 2.
wherein the parameter beta controls the influence of the click ratio between the parent and child categories on the final effect. The sub-set of catx is subcatx, and i is one of the sub-sets.
thus, the second network device obtains the first category optimization information containing the ranking scores of the categories corresponding to the query terms.
in step S22, the second network device determines second category optimization information corresponding to the query sequence according to the mapping relationship between the query term and the category and the modified query sequence corresponding to the query sequence.
Here, when the user is not expected to face the current search result, the query term may be modified for optimization. The modified query sequence corresponds to a query sequence of query terms and/or categories, etc. modified by the user.
For example, when a user searches for "matchsticks" (which are expected to be an outdoor brand), if matchsticks for ignition are returned, the user will change the query word to "matchstick outdoor".
When the modified query term queryb is a child intention string of the query term querya before modification, then let the set of child intention strings of querya be: sub _ querya. The second network device determines second category optimization information for a query sequence containing the query term querya using equation 3.
Wherein the parameter gamma controls the influence of the click ratio between the parent and child categories on the final effect. Subcatx is a subset of catx.
Thus, the second network device obtains second category optimization information including the ranking scores of the categories corresponding to the query terms.
Next, in step S3, the second network device determines category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information.
Specifically, the second network device may determine the category prediction information of the query sequence by comprehensively evaluating probabilities or scores of various categories in the first category optimization information and the second category optimization information.
For example, the probability of the category a1 in the first category optimization information is p1, the probability of the category a2 in the first category optimization information is p2, and the probability of the category a1 in the second category optimization information is p 1', and the probability of the category a2 in the second category optimization information is p1is p 2'. The second network device firstly normalizes the probabilities p1, p2, p1 'and p 2' according to the number of categories, and thus obtaining the category prediction information of the query sequence comprises the following steps: category a1 and the corresponding probability are:Category a2 and the corresponding probability are:
Preferably, the second network device updates, by weighting, the category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information.
For example, the second network device performs weighted update on the first category optimization information and the second category optimization information by using formula 4, and obtains category prediction information corresponding to the query sequence.
And the parameters a and b respectively control the influence of the first category optimization information and the second category optimization information on the final prediction effect.
With formula 4, the obtaining, by the second network device, category prediction information includes: each category in the first category optimization information and the second category optimization information and the corresponding weighted score ffinal(query,catx,c)。
then, the second network device transmits the obtained category prediction information to the user equipment. And the user equipment sends the category prediction information and the query sequence to the first network equipment. And the first network equipment queries at least one search result corresponding to the query sequence. When the first network device inquires the search result, step S7 is executed.
It should be noted that, the two network devices may also directly send the category prediction information to the first network device.
in step S7, the first network device determines priority information of the search result based on matching degree information of category information of the search result and the category prediction information.
Specifically, the first network device matches category information of each search result with categories in the category prediction information, determines priority information of the category information of each search result according to complete agreement, partial agreement, and disagreement, and corresponds the determined priority information to each search result.
Preferably, the first network device determines the priority information of the search result based on the matching degree information and the content relevance information of the search result and the query sequence.
Specifically, the first network device performs weighting processing on preset matching degree information and content correlation information to determine priority information of the search result.
for example, the search results include: search result b1, search result b2, and search result b 3. The category corresponding to the search results b1 and b2 is category c1, and the category corresponding to the search result b3 is c 2. Categories in the category prediction information include: c1, c 3. The query sequence includes: the query term "aabb".
the first network device obtains the priority information Y1 of the search results b1 and b2 and the priority information Y2 of the search result b3 through matching categories, wherein Y1> Y2. Meanwhile, character matching is carried out to obtain a search result b1 containing 'aabb', and if complete matching is carried out, the content correlation information is determined to be Y3; if the search result b2 includes "aa", partial matching determines that the content relevance information is Y4, and if the search result b3 includes "bb", partial matching determines that the content relevance information is Y4, wherein Y3> Y4.
The weight value of the first network device according to the preset category matching degree is a 1%, the weight value of the content relevance is a 2%, wherein a1> a2, the priority information of the search result b1 is (a 1% × Y1+ a 2% × Y3), the priority information of the search result b2 is (a 1% × Y1+ a 2% × Y4), and the priority information of the search result b3 is (a 1% × Y2+ a 2% × Y4).
The first network device may also determine priority information of each search result according to a score or probability of each category in the category prediction information as a weight.
Continuing with the above example, the first network device determines that the priority information of the search result b1 is ffinal_c1(a 1% Y1+ a 2% Y3), the priority information of the search result b2 being ffinal_c1(a 1% Y1+ a 2% Y4), the priority information of the search result b3 being ffinal_c3(a 1% Y2+ a 2% Y4). Wherein f isfinal_c1predicting a score, f, for category c1 in the information for the categoryfinal_c3And scoring the categories of the first network equipment which are not matched with the categories in the category prediction information.
It should be noted that, as those skilled in the art should understand, the manner of determining the priority information of the search results is only an example, and in fact, the first network device may also determine the priority information of each search result by a weighted average and the like manner. In addition, other existing or future manners of determining priority information for the search results, as applicable to the present application, are also intended to be encompassed within the scope of the present application and are hereby incorporated by reference.
After determining the priority information of each search result, the first network device provides each search result and corresponding priority information to the user equipment, and the user equipment displays at least one search result to a user.
The first network device sends all the search results and corresponding priority information to the user equipment, and the user equipment displays the search results in a whole or in a paginated manner according to the priority from high to low according to the screen size of the user equipment where the user equipment is located.
Preferably, the first network device performs ranking processing on at least one of the one or more search results according to priority information of the search result; and then providing the corresponding sorting result to the user equipment.
Specifically, the first network device provides the one or more search results to the user equipment according to the priority information of the search results in the order from high to low in priority.
Here, the first network device may provide all search results to the user device in one page. Partial search results may also be provided to the user device through a paging technique.
In summary, the method and system for providing search results of the present application can effectively solve the problem that the search results are far from the will expressed by the query sequence input by the user by predicting the categories of the query sequence provided by the user equipment and setting the priority of each search result according to the matching degree between the category to which each search result corresponding to the query sequence belongs and the predicted category; in addition, the predicted tendency of the categories is adjusted by collecting the query words before and after the adjustment of the user, so that the Martian effect generated in the searching process is effectively reduced; in addition, the priority information of each search result is determined according to the matching degree of the category information corresponding to each search result and the category prediction information and the content correlation of the query sequence and each search result, so that the search result which best meets the user's intention can be preferentially displayed to the user, the user can obtain the expected search result as desired, and the use experience of the user is effectively improved. Therefore, the application effectively overcomes various defects in the prior art and has high industrial utilization value.
it will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the apparatus claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Claims (21)
1. A method at a first network device for providing search results, wherein the method comprises:
Obtaining one or more search results corresponding to a query sequence sent by user equipment;
acquiring category prediction information corresponding to the query sequence;
determining priority information of the search result based on matching degree information of category information of the search result and the category prediction information;
Providing at least one of the one or more search results to the user device according to the priority information of the search result.
2. the method of claim 1, wherein the obtaining category prediction information corresponding to the query sequence comprises:
And receiving category prediction information corresponding to the query sequence sent by the user equipment.
3. The method of claim 1 or 2, wherein the determining priority information for the search results comprises:
And determining the priority information of the search result based on the matching degree information and the content relevance information of the search result and the query sequence.
4. The method of claim 1, wherein the providing at least one of the one or more search results to the user device according to the priority information of the search result comprises:
at least one of the one or more search results is subjected to sorting processing according to the priority information of the search results;
Providing the corresponding sorting result to the user equipment.
5. a method at a user equipment for providing search results, wherein the method comprises:
Sending the query sequence to the corresponding second network equipment;
Receiving category prediction information corresponding to the query sequence returned by the second network device;
Sending the query sequence and the category prediction information to corresponding first network equipment;
and receiving one or more search results corresponding to the query sequence returned by the first network equipment.
6. The method of claim 5, wherein the one or more search results are ordered by priority information for the search results, wherein the priority information is determined based on matching degree information of category information of the search results with the category prediction information.
7. a method at a second network device for determining category prediction information corresponding to a query sequence, wherein the method includes:
Receiving a query sequence sent by user equipment;
Determining first category optimization information and second category optimization information corresponding to the query sequence;
determining category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information;
And sending the category prediction information to the user equipment.
8. The method of claim 7, wherein the determining category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information comprises:
and updating the category prediction information corresponding to the query sequence in a weighting manner according to the first category optimization information and the second category optimization information.
9. The method of claim 7 or 8, wherein the method further comprises:
carrying out statistical processing on the search log information to obtain a mapping relation between the query words and the categories;
wherein the determining the first category optimization information and the second category optimization information corresponding to the query sequence includes:
And determining first category optimization information and second category optimization information corresponding to the query sequence according to the mapping relation between the query word and the category.
10. The method of claim 9, wherein the determining the first category optimization information and the second category optimization information corresponding to the query sequence comprises:
Determining first category optimization information corresponding to the query sequence according to the mapping relation between the query terms and the categories and the category hierarchical structure information corresponding to the query sequence; or
And determining second category optimization information corresponding to the query sequence according to the mapping relation between the query terms and the categories and the modified query sequence corresponding to the query sequence.
11. a first network device for providing search results, wherein the first network device comprises:
A result obtaining device, configured to obtain one or more search results corresponding to a query sequence sent by a user equipment;
The category acquisition device is used for acquiring category prediction information corresponding to the query sequence;
Priority determining means for determining priority information of the search result based on matching degree information of category information of the search result and the category prediction information;
Providing means for providing at least one of the one or more search results to the user equipment in accordance with the priority information of the search result.
12. The first network device of claim 11, wherein the category obtaining means is configured to:
and receiving category prediction information corresponding to the query sequence sent by the user equipment.
13. The first network device of claim 11 or 12, wherein the priority determining means is configured to:
And determining the priority information of the search result based on the matching degree information and the content relevance information of the search result and the query sequence.
14. The first network device of claim 11, wherein the providing means is configured to:
At least one of the one or more search results is subjected to sorting processing according to the priority information of the search results;
Providing the corresponding sorting result to the user equipment.
15. a user device for providing search results, wherein the user device comprises:
The first sending device is used for sending the query sequence to the corresponding second network equipment;
A first receiving device, configured to receive category prediction information corresponding to the query sequence returned by the second network device;
A third sending device, configured to send the query sequence and the category prediction information to a corresponding first network device;
And a third receiving device, configured to receive one or more search results corresponding to the query sequence, where the search results are returned by the first network device.
16. The user device of claim 15, wherein the one or more search results are ordered by priority information for the search results, wherein the priority information is determined based on matching degree information of category information for the search results with the category prediction information.
17. A second network device for determining category prediction information corresponding to a query sequence, wherein the second network device comprises:
A second receiving device, configured to receive a query sequence sent by a user equipment;
The second determining device is used for determining the first category optimization information and the second category optimization information corresponding to the query sequence;
a fourth determining device, configured to determine category prediction information corresponding to the query sequence according to the first category optimization information and the second category optimization information;
And the second sending device is used for sending the category prediction information to the user equipment.
18. The second network device of claim 17, wherein the fourth determining means is configured to:
and updating the category prediction information corresponding to the query sequence in a weighting manner according to the first category optimization information and the second category optimization information.
19. The second network device of claim 17 or 18, wherein the second network device further comprises:
the statistical device is used for carrying out statistical processing on the search log information so as to obtain the mapping relation between the query words and the categories;
wherein the second determining means is for:
and determining first category optimization information and second category optimization information corresponding to the query sequence according to the mapping relation between the query word and the category.
20. the second network device of claim 19, wherein the second determining means comprises:
A first determining unit, configured to determine, according to the mapping relationship between the query term and the category hierarchical structure information corresponding to the query sequence, first category optimization information corresponding to the query sequence; or
And the second determining unit is used for determining second category optimization information corresponding to the query sequence according to the mapping relation between the query terms and the categories and the modified query sequence corresponding to the query sequence.
21. A system for providing search results, wherein the system comprises at least two of a first network device according to any one of claims 11 to 14, a user device according to claim 15 or 16 and a second network device according to any one of claims 17 to 20.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510094491.5A CN105989156B (en) | 2015-03-03 | 2015-03-03 | Method, equipment and system for providing search results |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510094491.5A CN105989156B (en) | 2015-03-03 | 2015-03-03 | Method, equipment and system for providing search results |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105989156A CN105989156A (en) | 2016-10-05 |
| CN105989156B true CN105989156B (en) | 2019-12-17 |
Family
ID=57038240
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510094491.5A Active CN105989156B (en) | 2015-03-03 | 2015-03-03 | Method, equipment and system for providing search results |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105989156B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887437A (en) * | 2009-05-12 | 2010-11-17 | 阿里巴巴集团控股有限公司 | Search result generating method and information search system |
| CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
| CN102591890A (en) * | 2011-01-17 | 2012-07-18 | 腾讯科技（深圳）有限公司 | Method for displaying search information and search information display device |
-
2015
- 2015-03-03 CN CN201510094491.5A patent/CN105989156B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887437A (en) * | 2009-05-12 | 2010-11-17 | 阿里巴巴集团控股有限公司 | Search result generating method and information search system |
| CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
| CN102591890A (en) * | 2011-01-17 | 2012-07-18 | 腾讯科技（深圳）有限公司 | Method for displaying search information and search information display device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105989156A (en) | 2016-10-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107346326B (en) | Method and system for information retrieval | |
| US10180967B2 (en) | Performing application searches | |
| US9818142B2 (en) | Ranking product search results | |
| US10489448B2 (en) | Method and system for dynamically ranking images to be matched with content in response to a search query | |
| JP5984917B2 (en) | Method and apparatus for providing suggested words | |
| CN103455531B (en) | A kind of parallel index method supporting high dimensional data to have inquiry partially in real time | |
| WO2017096892A1 (en) | Index construction method, search method, and corresponding device, apparatus, and computer storage medium | |
| WO2012039760A1 (en) | Processing of categorized product information cross reference to other applications | |
| JP2014512614A (en) | Method and apparatus for determining product category information | |
| EP2724267A1 (en) | Search method and apparatus | |
| CA3085463A1 (en) | Search engine for identifying analogies | |
| CN105574030A (en) | Information search method and device | |
| CN108140022B (en) | Data query method and database system | |
| US9300712B2 (en) | Stream processing with context data affinity | |
| JP2021068448A (en) | Method, apparatus, and system for data mapping | |
| CN110929218A (en) | Difference minimization random grouping method and system | |
| WO2019179208A1 (en) | Drop-down suggestion list | |
| CN104615620A (en) | Map search type identification method and device and map search method and system | |
| WO2022268089A1 (en) | Data processing method and system and related device | |
| CN104077555A (en) | Method and device for identifying badcase in image search | |
| US11379535B2 (en) | Accelerated large-scale similarity calculation | |
| CN105989156B (en) | Method, equipment and system for providing search results | |
| CN107622048B (en) | Text mode recognition method and system | |
| CN107273509B (en) | Neural network data memory, data storage method and data search method | |
| CN112464101A (en) | Electronic book sorting recommendation method, electronic device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1229487 Country of ref document: HK |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |