WO2015109944A1

WO2015109944A1 - Method and device for displaying search results

Info

Publication number: WO2015109944A1
Application number: PCT/CN2015/070303
Authority: WO
Inventors: Yang Liu
Original assignee: Tencent Technology (Shenzhen) Company Limited
Priority date: 2014-01-24
Filing date: 2015-01-07
Publication date: 2015-07-30
Also published as: CN104123332A; CN104123332B

Abstract

A method for displaying search results may include recording historical search data, wherein the historical search data may include historical search words, one or more historical query results associated with each historical search word, and a historical click count of each historical query result; obtaining a current search word inputted by a user; querying the current search word among the historical search data to obtain a set of associated historical search words associated with the current search word; identifying historical query results for the set of associated historical search words; determining the click weight of each of the historical query results for the set of associated search words, according to the historical click count of each historical query result; and displaying, as the search results for the current search word, the historical query results for the set of associated historical search words ordered sequentially according to the click weight.

Description

A METHOD AND A DEVICE FOR DISPLAYING SEARCH RESULTS

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to Chinese Application No. 2014100365585, filed on January 24, 2014, which is hereby incorporated by reference in its entirety.

TECHICAL FIELD

The present disclosure relates to the field of data searches and more particularly to a method and a device for displaying search results.

BACKGROUND

The operating principle of a search engine may include saving gathered search result information into a database； extracting information and creating an index database； a retriever quickly retrieves documents in the index database according to a query keyword inputted by a user, evaluates the relevancy between documents and query, ranks the query results to be outputted, and returns the query results to the user.

Specifically, an existing search engine provides a type of retrieval services. When a user uses a keyword to search for information, the search engine searches the database first. If finding some content desired by the user, the search engine computes the relevancy and ranking of different matched content according to the keywords in webpages, and returns the matched results sequentially to the user according to their relevancy.

This full-text search engine can display the content matching the search results by means of keyword matching. However, each search needs to traverse all matched objects, so the search efficiency is low and the user cannot find the desired content more quickly.

The above search mode based on keyword matching provided in related technologies cannot provide higher search efficiency. As yet, no effective solution has been available to overcome the problem.

BRIEF SUMMARY

The embodiments of the present disclosure provide a method and a device for displaying search results. The present disclosure improves searching efficiency, for example compared to other search technologies based on keyword matching.

In one aspect, a method may be performed in a device that comprises a processor and a device, and the device and/or processor may perform the steps of: recording historical search data, wherein the historical search data includes historical search words, one or more historical query results associated with each of the historical search words, and a historical click count of each historical query result； obtaining a current search word inputted by a user； querying the current search word among the historical search data to obtain a set of associated historical search words that are associated with the current search word； identifying historical query results for the set of associated historical search words； determining a click weight of each of the historical query results for the set of associated historical search words, according to the historical click count of the historical query results for the set of associated historical search words； and displaying, as search results for the current search word, the historical query results for the set of associated historical search words ordered sequentially according to the click weight, through the user interface

In another aspect, a device for displaying search results may include: a saving module or circuitry configured to record historical search data, wherein the historical search data comprises historical search words, one or more historical query results associated with each of the historical search words, and a historical click count of each historical query result； a processing module or circuitry configured to obtain a current search word and query the current search word among the historical search data to obtain a set of associated historical search words that are associated with the current search word； a determining module or circuitry configured to identify historical query results for the set of associated historical search words and determine a click weight of each of the historical query results for the set of associated historical search words, according to the historical click count of the historical query results for the set of associated historical search words； and a display processing module or circuitry configured to display, as search results for the current search word, the historical query results for the set of associated historical search words ordered sequentially according to click weight.

The present disclosure may provide adaptive learning of user search behaviors that may provide more relevant, accurate, or satisfactory search results to a user. By learning the search results and historical search behaviors of a user, the present disclosure can find search results more accurately according to existing historical search data after a user inputs a current search word. The present disclosure may perform a more thorough and more accurate search query based on one or more historical search results (including previously selected/clicked search results from historical search words associated with a current search word) , and provides for determining of a display sequence of the current search results according to the click count of historical search results. Consequently, the search results in a display menu presented to a user may be more accurate and the user can find the query content more conveniently. Accordingly, the present disclosure may lead to improved search efficiency and increased accuracy and relevancy in search results. Further, the present disclosure may overcome the following problem: searches based on keyword matching provided in related technologies cannot provide higher search efficiency.

BRIEF DESCRIPTION OF THE DRAWINGS

The attached drawings as a part of the application further illustrate the present disclosure. The illustrative embodiments of the present disclosure and corresponding descriptions are for the purpose of explanation, and are not intended to limit the present disclosure. Among the attached drawings:

Figure 1 shows an example structure diagram of a device for displaying search results；

Figure 2 shows another example structure diagram of the device for displaying search results；

Figure 3 shows yet another example structure diagram of a device for displaying search results；

Figure 4 shows an example structure diagram of yet another device for displaying search results；

Figure 5 shows an example structure diagram of still another device for displaying search results； and

Figure 6 shows a flowchart of an example method for displaying search results.

DETAILED DESCRIPTION

Note that any portion of the various embodiments, options, and variations presented in this disclosure and corresponding characteristics can be combined as additions or alternatives. The present disclosure is described in detail by reference to the attached drawings and in conjunction with the embodiments, options, and variations.

To help a person skilled in the art to better understand the solutions to the present disclosure, the following section describes the technical solutions one or more of the embodiments of the present disclosure may provide, and in conjunction with the attached drawings. The described embodiments are some, but not all, of the embodiments contemplated by the present disclosure. All other embodiments that a person with ordinary skills in the art may derive from the embodiments herein without creative work should nevertheless fall within the protection scope of the present disclosure.

Note that such terms as “the first” and “the second” used in the Specification, Claims, and above attached drawings of the present disclosure are intended to distinguish objects, but not necessarily describe a specific sequence or precedence order. It should be understood that any element, term, or data qualified by such terms are interchangeable in appropriate circumstances so that embodiments of the present disclosure herein can be implemented in an order other than those illustrated or represented herein. Terms such as “comprise” and “have” and any of their variants are intended to cover non-exclusive inclusions. For example, the processes, methods, systems, products or devices comprising a series of steps or units are not necessarily limited to the steps or units that are clearly listed, but can include additional or alternative steps or units, such as those that are inherent to the processes, methods, systems, products or devices.

Before describing further details of the present disclosure, the following section describes a possible computing device used to carry out the principles of the present disclosure by reference to Figure 1. Unless otherwise noted, the following section describes each embodiment of the present disclosure by reference to the legends of the actions and operations executed by one or multiple computers. Therefore, it can be understood that the actions and operations executed by the computers may include the manipulation of the electrical signals represented in a structured form by the processing units of computers. The manipulation may, for example, convert data or maintain data in a storage system of computers, thereby reconfiguring or changing the operations of computers in a manner comprehensible to any person skilled in the art. The data structure of maintenance data is the physical location of the memory that has specific attributes defined by a data format. The present disclosure is described in the above context, but the description is not restrictive. As understood by a person skilled in the art, various respects of the actions and operations described in the following section can also be implemented by hardware.

Figure 1 shows the structure diagram of an exemplary device for displaying search results. The exemplary device shown in Figure 1 is just an example, and places no restriction upon the usage scope or functions of the present disclosure. In addition, any system implementing the techniques and features disclosed herein should not be deemed to place any reliance or requirement on any component or any combination of components shown in Figure 1.

As shown in Figure 1, the device for displaying search results can comprise modules, logic, or circuitry, such as a saving module 11, a processing module 13, a determining module 15, and a display processing module 17. The modules may be implemented through hardware, circuitry, processor-executable instructions, one or more software programs, or any combination thereof.

The saving module 11 may record historical search data. The historical search data may include data related to searches performed by a search engine, e.g., keyword searches performed through a search website or other online resource. The historical search data may include, as example data elements, historical search words, historical query results associated with each historical search word, and a historical click count of each historical query result. A historical query result may indicate one or more previous search results obtained, displayed, return, selected, or clicked for a particular historical search word. For instance, and as described more below, a history query result may include a user-selected result from a previous search for a historical search word. The saving module 11 may implement, communicate with, or access one or more databases or memories configured to store the historical search data.

The processing module 13 may obtain or identify a current search word, such as a current search word inputted by a user. Upon obtaining a current search word inputted by a user, the processing module 13 may perform a query on the historical search data for the current search word. For instance, the processing module 13 may query the current search word among the historical search data and determine and obtain one or more historical search words associated with the current search word. These associated historical search words may share one or more similarities with the current search word, such as similarities in construction, letters composition, length, text-based similarity, topics or categorizations, or any other configurable similarity criteria. The processing module 13 may also identify historical query results for the associated historical search words.

The determining module 15 may determine the click weight for the historical query results for the associated historical search words, and according to the historical click count of each historical query result respectively for the associated historical search words. The display processing module 17 may present a display menu of search results for the current search word. In particular, the display processing module 17 may display, as the search results, the historical query results for the associated historical search words ordered sequentially according to the click weight, e.g., according to the determined click weight of each of the associated historical query results.

The saving module 11 may record historical search data that track or indicate historical search behaviors of a user, and the present disclosure may determine search results that account for previous searches and previously clicked/selected search results. By learning, adapting to, and using the historical search results and historical search behaviors of a user, the present disclosure may provide more accurate and/or personalized search results based on the historical search behavior of the user. Thus, the disclosure may return search results with increased accuracy for a current search word according to the existing historical search data of the user. In the above process, the present disclosure may perform a deeper and more accurate query of the current search word based on one or more query results, and can determine the display sequence of the current search results according to the click count of historical search results. Consequently, the search results in the display menu are more accurate and the user can find the query content more conveniently, which may lead to improved search efficiency and more accurate search results. Further, the present disclosure may overcome lower search efficiency problems for the search modes based on keyword matching provided by other technologies.

In the above embodiment of the present disclosure, the historical search data can be saved as a keyword mapping table, for example as shown in Table 1 below.

Table 1

In the exemplary historical search data shown in Table 1, as a result of automatic learning of the system, the historical search data can include: historical search words, historical query results associated with each historical search word, and a historical click count of each historical query result. The historical search data in Table 1 may be query results that are obtained by means of keyword matching for a current search word. In that regard, the device may determine previous/historical search words “al” , “an” , “a” , and/or “xiaolong” from a current search word, and Table 1 may include previously selected search results as the historical query results for these particular historical search words. Additional or alternative features are discussed in Figure 2.

Figure 2 shows another example structure diagram of a device for displaying search results. In the example shown in Figure 2, a processing module 13 can include a match module 131 and a querying module 132, which the device may implement in hardware, software, or as a hardware-software combination.

The match module 131 may match a current search word with one or more historical search words among the historical search data. For instance, the match module 131 may match the current search word with a set of associated historical search words among the historical search data, wherein the set of associated historical search words comprises at least one historical search word of the current search word. Thus, the match module 131 may obtain a set of historical search words associated with the current search word.

The querying module 132 may access and use the historical query results associated with each historical search word in the set of associated historical search words as a set of search results associated with the current search word.

Thus, the device may obtain search results associated with a current search word among the historical search data, and according to the historical search behaviors of the user. To illustrate, when the user inputs a current search word, the match module 131 matches the current search word with associated historical search words among the historical search data, and obtains the historical search words matching the current search word, thereby constituting a set of historical search words associated with the current search word.

In some variations, the match module 131 matches the current search word with one or more historical search words by querying the historical search data to identify the historical search words that include or contain the current search word. For instance, the match module may identify, among the historical search data, historical search words that are consistent with the current search word or historical search words whose initial one or more characters match the current search word, e.g., completely match, or match up to a certain threshold percentage. To illustrate, take the historical search words in Table 1 as an example. If the current search word is “a” , the matching module 131 may identify associated historical search words that match the current search word to include: “al” , “an” , and “a” .

Upon obtaining a set of associated historical search words matching the current search word, the device may determine search results associated with the current search word. In particular, the device may determine the search results for the current search word as the historical query results for the set of associated historical search words. Take the historical search words in Table 1 as an example again. Upon matching the historical search words “al” , “an” , and “a” as being associated with a current search word “a” , the querying module 132 may obtain historical query results respectively for the associated historical search words, e.g., the historical query results for the historical search word “al” , the historical query results for the historical search word “an” , and the historical query results for the historical search word “a” . Accordingly, the query module 132 may obtain a group of matched historical query results that include “all” , “angela” , and “andie” from Table 1. The device may return this group of historical query results as the search results for the current search word. Note that both the historical query results for the historical search word “a” and “al” both include the previously selected search result “all” .

Figure 3 shows yet another example structure diagram for a device that display search results. In Figure 3, a determining module 15 may include a reading module 151 and an acquisition module 152.

The reading module 151 may determine, from the historical search data, a historical click count of one or more historical query results in a set of associated historical search words, e.g., the associated historical search words for a current search word. To illustrate, take the historical search words in Table 1 as an example. The reading module 151 may identify the historical click count of the historical query results “all” , “angela” , and “andie” associated with the current search word “a” , and in particular as historical click count values of 4, 2, and 1 respectively. Note that the historical search words that match a current search word of "a" may include “al” and “a” . In this example, both historical search words “al” and “a” include a historical query result of “all” , and the reading module 151 may determine the historical click counts for the historical query result of “all” to include the click count attributed to historical search word “al” and “a” . In other words, when historical query results for two, different historical search words include an identical previously selected result (e.g., the historical query result of “all” in the example above) , then the reading module 151 may account for the historical click count arising from both of these historical search words for the identical historical query result. That is, the reading module 151 may determine the historical click count of the historical query result “all” as the sum of the historical click count for the query result “all” arising from previous searches for the historical search word “al” and similarly for the historical click count arising from previous searches for the historical search word “a” .

The acquisition module 152 may determine a click weight attributable to one or more historical query results, for example by applying a click weight function. As one example, the acquisition module 152 may multiply the historical click count of each historical query result in the set of associated historical search words by an associated click weight coefficient. In this way, the acquisition module 152 may obtain a click weight of each search result (e.g., historical query result) associated with the current search word. To illustrate, take the historical search words in Table 1 as an example. In one example, the acquisition module may apply a click weight coefficient of 3, which may be configurable or preset. In this example, the acquisition module may determine the click weights of the three historical query results of “all” , “angela” , and “andie” that associated with a current search word “a” as 12, 6, and 3 respectively.

As discussed above, the reading module 151 and the acquisition module 152 may collaborate to determine a click count and click weight for search results for a current search word, wherein the search results include historical query results for historical search words matched to the current search word. In doing so, the device may establish, determine, attribute, or identify a relevancy between the search keywords of the user’s current search behavior and the historical click behaviors. One product of such a determination may include an indexed list of a current search word, for example as described in Table 2.

Table 2

The device may, for example generate Table 2 as part of displaying search results for a current search word. In particular, Table 2 illustrates a current search word of “a” , and the resulting matched historical query results that the device may determine as search results. Table 2 also includes a historical click count and click weight the device may determine for the matched historical query results. Thus, for current search word of “a” , the device may determine search results that include historical query results “all” , “angela” , and “andie” , as well as their respective click weights (12, 6, and 3 respectively in Table 2) . As seen in Table 2, if the device identifies an identical historical query result associated with the multiple historical search words (e.g., historical search words “al” and “a” matching the current search word “a” both include the historical query result “all” ) , the device may combine (e.g., sum, or apply some other combinative function) the historical click counts of the identical historical query result associated with multiple historical search words.

Figure 4 shows another example of a structure diagram for a device that displays search results. As shown in Figure 4, the above display processing module 17 may include a ranking module 171 and a displaying module 172.

The ranking module 171 may rank search results determined by the device, for example in a descending order according to the click weight. Thus, the ranking module 171 may obtain a set of descending search results, the descending search results may be indicative of historical user behavior in searching, which may provide search results with increased efficiency and accuracy.

The displaying module 172 may present the search results, for example through a user interface. The displaying module 172 may selectively present the search results, utilizing various portions of a display screen to emphasize particular search results. For example, the displaying module 172 may display a first N number of search results in the set of search results sequentially in a first display area of a display menu, and display remaining search results (e.g., other than the first N search results) sequentially in a second display area of the display menu. In particular variations, the first display area can be the commonly accessed, viewed, or used area of the user, e.g., a top of the screen or prioritized side bar of a display screen. The number N may be configurable by the device, and may include the first 3, 5, or 10 search results from a set of descending search results for example.

The display module 172 may present the first N search results in a high priority or selected portion of a user interface, as these first N search results may represent the search results (including historical query results) that the device learns according to historical search data that the user clicks most frequently, e.g., over a predetermined period of time. As examples, the first display area can be located in a pull-down menu of the search input box, and display the first N top-ranked search results (for example, 3 search results) in an eye-catching position. Other search results aside from the first N search results can be placed in the second display area, and may reflect historical query results that the system learns according to historical search data that the user clicks less frequently in comparison. The second display area can be appropriately located according to the layout of the display screen.

The device may adjust the click weight of one or more historical query results. For example, the device can include a setting module that may reset the click weight of a particular search result. For instance, the setting module may reduce the click weight of a first search result (e.g., to 0, or by a configurable reduction value) upon detecting that a count of times where the first search result is displayed in a first display area and not clicked exceeds a particular threshold. That is, the device may learn the behavioral aspect of a user that a particular search result has not being clicked/selected by a user more than a threshold number of times when being displayed in a (e.g., more prominent) first display area, even though the particular search result may have been selected by the user more in the past. Accordingly, the device may reduce the click weight of this particular search result, allowing for another search result to be displayed in the first display area that may be more relevant or accurate with respect to the current search word for the user, and may thus increase the likelihood of being selected by the user when presented in the first display area.

Accordingly, the device may adaptively learn search results according to historical search data, and the device may attempt to present search results displayed in the first display area that are more frequently clicked search results among historical records (e.g., from past searches) . Additionally, the device may vary search results according the user’s search habits and behaviors. The device may record the click count of the search results displayed in the first display area. If the device detects that a count of times when a search result displayed in the first display area has not been clicked exceeds a non-click threshold (for example, 5) , the device may determine that the current search result is no longer a preferred search result of the user, and can then reset its click weight to 0. For example, the device may identify that a particular search result has consecutively not been clicked more than a non-click threshold number of times, and specifically for when the particular search results was presented in the first display area for a current search word. After reducing the click count or click weight of the particular search word, and during a subsequent search of the user, the device may operate such that first display area no longer displays the particular search result. The device can place the particular search result in the second display area instead, for example.

Figure 5 shows yet another example structure diagram for a device that displays search results. As shown in Figure 5, the device can include a detecting module 191 and a computing module 192.

The detecting module 191 may detect when one or more search results displayed in a display menu are clicked.

The computing module 192 may, upon detecting that a search result in the display menu is clicked, adjust a click weight of the clicked search result. For example, the computing module 192 may increment the click weight (or count) of the clicked search result by a first weight and/or decrement the click weight (or count) of one or more non-clicked search results in the display menu by a second weight. In some variations, the above first weight for incrementing a clicked search result may be set to 3, and the second weight for decrementing unclicked search results may be set to 1, though the values of these two weights can be otherwise configured or adjusted.

Optionally, the saving module 11 can include a first saving submodule that may record a current search word inputted by a user as a historical search word； a second saving submodule that may record the historical query results of the historical search word； and a third saving submodule that may record the click count of the history query results upon detecting that historical query results are clicked.

The following section provides a method variation for performing any combination of the capabilities or functionality of the devices described above. Note that the method steps can be executed on a computer system, for example through a set of computer-executable instructions. Moreover, the method may be performed in any varying sequence of steps.

Figure 6 shows a flowchart of an example method for displaying search results. As shown in Figure 6, the method for displaying search results can include the following steps:

Step S60: Recording historical search data, wherein the historical search data includes historical search words, historical query results associated with each of the historical search words, and a historical click count of each historical query result. As noted above The historical search data can be search data recorded by a local terminal over a period of time or historical search data imported by from external device, e.g., in a template-based format.

Step S62: Obtaining a current search word inputted by a user and querying the current search word among the historical search data to obtain one or more associated historical search words, e.g., historical search words that match or are associated with the current search word as discussed above. S62 may also include identifying the historical query results for the associated historical search words, which may provide insight as to previously selected search results for the associated historical search words.

Step S62 may help determine search results by using the search data of the user’s past search behaviors. In that regard, the method may provide search results based on historical search data, e.g., without a full-text match process which may improve search efficiency.

Step S64: Determining the click weight of each of the historical query results for the associated historical search words, according to the historical click count of each historical query result for the associated historical search words.

Step S66: Displaying, as the search results for the current search word, the historical query results for the associated historical search words ordered sequentially according to the click weight of each search result, e.g., through a user interface or via a display menu.

Steps S64 to S66 may include determining a display sequence for the search results. Determining the display sequence may include accessing the click counts of historical query results, which may be indicative of past search behaviors relevant to the current search word. Thus, the method may return more relevant, desired search results to the user and associate specific content relevant to the user more quickly.

The exemplary method described herein may record historical search data that track or indicate historical search behaviors of a user, and the present disclosure may determine search results that account for previous searches and previously clicked/selected search results. By learning, adapting to, and using historical search results and historical search behaviors of a user, the present disclosure may provide search results with increased relevancy, personalization, and accuracy, e.g., as compared to other search methods. In the above process, the present disclosure may perform a deeper, more thorough, and more accurate search of the current search word based on one or more historical query/search results, and can determine the display sequence of the current search results according to the click count of historical search results. Consequently, the search results in the display menu may be more accurate and the user can find the query content more conveniently, so the technical result of more accurate search results and increased efficiency is achieved. Further, the present disclosure overcome search efficiency problems for search modes based on keyword matching provided by other technologies.

In the above embodiment of the present disclosure, the historical search data can be saved as a keyword mapping table, as shown in Table 1 above and reproduced below.

Table 1 (Reproduced from above)

In the exemplary historical search data shown in Table 1, as a result of automatic learning of a system or method, the historical search data can include: historical search words, historical query results associated with each historical search word, a historical click count of each historical query result, or any combination thereof. The historical search words identified in the historical search data shown in Table 1 may, for example, be the query results that are obtained by means of keyword matching for a current search word.

Step S62 (querying the current search word among the historical search data and obtaining a plurality of search results associated with the current search word) may be include one or more of the following sub-steps:

Step S621: Matching the current search word with one or more historical search words among the historical search data. Matching may include matching the current search word with a set of associated historical search words among the historical search data, wherein the set of associated historical search words comprises at least one historical search word of the current search word, and thus obtaining a set of historical search words associated with the current search word.

Step S623: Using or accessing historical query results associated with each historical search word in the set of associated historical search words as a set of search results associated with or for the current search word. In that regard, the search results may include previously clicked or selected search results from previous searches performed for the historical search words associated with the current search word.

Thus, the method may include obtaining search results associated with a current search word among the historical search data, and according to the historical search behaviors of the user. To illustrate, when the user inputs a current search word, the present disclosure may match the current search word with associated historical search words among the historical search data at Step S621, and obtain the historical search words matching the current search word, thereby constituting a set of historical search words associated with the current search word.

In some variations, matching the current search word with one or more historical search words may include applying a matching condition. The matching condition may specify historical search words as matching the current search word when, for example, the historical search word contains the current search word. Here, “containing” can include when the current search word is completely consistent with a historical search word or is located in the beginning of a historical search word, e.g., the initial characters of the historical search word is the current search word. Take the historical search words in Table 1 as an example. If the current search word is “a” , the matching condition may specify identifying the historical search words matching the current search word to include: “al” , “an” , and “a” as these historical search words start with the current search word “a” .

Upon obtaining a set of associated historical search words matching the current search word, the present disclosure may obtain a set of search results associated with the current search word. The search results may include the historical query results for the set of associated historical search words. Take the historical search words in Table 1 as an example again. Upon identifying or matching the historical search words “al” , “an” , and “a” as being associated with the current search word, the method may include obtaining a group of matched historical query results that include the words “all” , “angela” , and “andie” from Table 1. This group of historical query results may be included as part of the search results of the current search word.

Step S64 (determining the click weight of each search result according to the historical click count of each historical query result) may include the following sub-steps:

Step S641: Reading or determining, from the historical search data, a historical click count of one or more historical query results in the set of associated historical search words. To illustrate, take the historical search words in Table 1 as an example. The historical click count of the historical query results “all” , “angela” , and “andie” associated with the current search word “a” may be determined to be 4, 2, and 1 respectively. Note that the historical search words matching with the current search word "a" include the historical search words “al” and “a” . In this example, both historical search words “al” and “a” include a historical query result of “all” , so the determined historical click count of the historical query result “all” may include click counts from previous searches for both historical search words “al” and “a” . In other words, when historical query results for two, different historical search words include an identical previously selected result (e.g., the historical query result of “all” in the example above) , then determining the historical click count for this historical query result may account for the historical click count arising from both of these historical search words.

Step S643: Determining a click weight attributable to one or more historical query results, for example by applying a click weight function. The click weight function may include multiplying the historical click count of each historical query result in the set of associated historical search words by an associated click weight coefficient, and thus obtaining the click weight of each search result (e.g., historical query result) associated with the current search word.

Take the historical search words in Table 1 as an example. The click weight coefficients of these historical search words may be preset to 3, as one example. Therefore, the click weights of the three historical query results “all” , “angela” , and “andie” which associated with the current search word “a” may be determined as 12, 6, and 3 respectively.

Specifically, in the above embodiment, Step S641 and Step S643 may jointly determine the click weight and click count for search results for the current search word, where the search results may be indicative of previously selected/clicked results from searches for historical search words matching the current search words. In this way, steps S641, S643, or both may identify, leverage, or establish a relevancy between the search keywords of the user’s current search behavior and the historical click behaviors from previous searches. One possible result from this process i the creating of an index list of the current search word, one example of which is described in Table 2 above and reproduced below.

Table 2 (reproduced from above)

Table 2 may be generated as part of displaying the search results for a current search word, for example. In particular, Table 2 illustrates a current search word of “a” , and the resulting matched historical query results that the device may determine as search results, which may include “all” , “angela” , and “andie” . Table 2 also includes a historical click count and click weight which may be determined for the matched historical query results. Thus, for current search word of “a” , the determined search results may include “all” , “angela” , and “andie” . The click weights determined for these search results may be 12, 6, and 3 respectively. If it is identified that an identical, common, or consistent historical query result is associated with multiple historical search words (e.g., “al” and “a”) , determining the click weight for this historical query result may include combining (e.g., adding up) the historical click counts of this historical query result arising from the multiple historical search words.

Step S66 (displaying a plurality of search results associated with the current search word in a display menu sequentially according to the click weight of each search result) can include the following sub-steps:

Step S661: Ranking search results in a descending order according to click weight, and thus obtaining a set of descending search results.

Step S663: Selectively displaying the search results, such as displaying the first N search results in the set of descending search results sequentially in a first display area of a display menu, and displaying other search results (e.g., other than the first N search results) sequentially in a second display area of a display menu.

Optionally, if it is detected that the number of times when a particular search result displayed in the first display area is presented and not clicked exceeds a non-click threshold, the click weight of this particular search result may be lowered or reset to 0.

In some variations, the first display area can be a commonly used area of a user. For instance, the first display area can be located in a pull-down menu of the search input box, and that displays the top-ranked search results (for example, the top 3 search results according to click weight) . Other search results can be placed in the second display area, which may include the search results that the method or system learns according to historical search data that the user clicks infrequently, e.g., less relevant. The second display area can be appropriately located according to the layout of the display screen.

Step S66 (displaying a plurality of search results associated with the current search word in a display menu sequentially according to the click weight of each search result) can be followed by, preceded by, or include the following sub-steps:

Step S71: Detecting whether a search result displayed in a display menu is clicked, and executing Step S73 upon detection.

Step S73: Adjusting the click weight of a clicked search result, a non-clicked search result, or both. Adjusting may include incrementing the click weight of the clicked search result displayed in the display menu by a first weight (for example, incrementing by 3) and/or decrementing the click weight of a non-clicked search results displayed in the display menu by a second weight (for example, decrementing by 1) .

Note that Step S60 (recording historical search data) can include any of the following sub-steps:

Step S601: Recording a historical search word inputted by a user, for example by recording a current search word inputted by the user as a historical search word in the historical search data.

Step S602: Recording the historical query results associated with each historical search word upon obtaining historical query results according to historical search words. To illustrate, when recording the current search word as a historical search word, recording the historical query results may include recording the search results for the current search word as the historical query results for the current search word.

Step S603: If detecting that history query results are clicked, recording the click count of the historical query results.

Step S601 to Step S603 may jointly implement the learning of historical search data by the system, e.g., by storing previous search words, search results for the previous search words, and search result selections from the previous searches, which may be subsequently accessed to determine search results for a current search word.

Note that each method embodiment described above is represented as a series of action combinations for the convenience of description. However, a person skilled in the art should know that the present disclosure is not restricted by the sequence of the described actions, because certain steps can be executed in other sequences or simultaneously according to the present disclosure. In addition, a person skilled in the art should also know that the embodiments described in the Specification are all preferred embodiments, and the involved actions and modules are not necessarily required for the present disclosure.

By describing the above embodiment, a person skilled in the art can clearly understand that the method based on the above embodiment can be carried out by means of software plus a necessary general hardware platform, or by means of hardware alone. In many cases, the former mode is preferred. Accordingly, the technical solution to the present disclosure or the part contributing to the prior art can be embodied in the form of a software product. This computerized software product is stored in a storage medium (for example, a ROM/RAM, a disk, or a CD-ROM) and comprises a certain number of instructions that instruct a terminal device (for example, a mobile phone, a PC, a server, or network equipment) to execute the method described in each embodiment of the present disclosure.

The present disclosure can further provide a mobile terminal used to run any of the above methods for displaying search results. In other words, both the method embodiment and device embodiment provided by the present disclosure can be run on a mobile terminal, and can, as an application, improve the accuracy of the search results associated with the user’s current search word by using historical search data, thereby offering good user experience.

The foregoing serial numbers of embodiments of the present disclosure are only used for a descriptive purpose, but do not stand for any superiority or inferiority of the embodiments.

The descriptions of each foregoing embodiment place particular emphasis on different aspects. For the part not detailed in an embodiment, refer to the related descriptions in other embodiments.

It should be understood that the clients disclosed in the embodiments provided by the application can be implemented in other ways. The device embodiments described above are just used for an illustrative purpose. For example, definition of units is only a definition of logical functions, which can be defined in other modes in practice； a plurality of units or components can be combined or integrated into another system, or some features may be ignored or not be executed. In addition, the mutual coupling, direct coupling, or communication connection displayed or discussed can be carried out through some interfaces, units, or modules, and can exist in electrical or other forms.

The units described as separate parts may be or may not be physically separated, and parts displayed as units may be or may not be physical units, namely, they can be located at the same place or distributed on multiple network units. Some or all of the units can be selected as actually required to achieve the objectives of the present embodiment solution.

Besides, various functional units in each embodiment of the present disclosure can be integrated into one processing unit or physically exist separately from each other. Alternatively, two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software function units.

If said integrated unit is implemented in the form of a software function unit and marketed or used as an independent product, it can be stored in a computer readable storage medium. Based on the above understanding, the essence of the technical solution to the present disclosure, a part of the technical solution contributing to the prior art, or all or part of the technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and comprises a certain number of instructions that instruct a computer device (such as a PC, a server, or network equipment) to execute all or part of the steps in the method described in each embodiment of the present disclosure. The aforesaid storage medium can be any type of medium that can store program codes, such as U disk, Read-Only Memory (ROM) , Random Access Memory (RAM) , mobile hard drive, magnetic disk, or CD-ROM.

he above only describe preferred embodiments of the present disclosure. Note that without departing from the principles of the present disclosure, a person ordinary skilled in the art can make improvements and polishments, which should nevertheless fall within the protection scope of the present disclosure.

Claims

A method for displaying search results in a device comprising a user interface, the method comprising:

recording historical search data, wherein the historical search data includes historical search words, one or more historical query results associated with each of the historical search words, and a historical click count of each historical query result；

obtaining a current search word inputted by a user；

querying the current search word among the historical search data to obtain a set of associated historical search words that are associated with the current search word；

identifying historical query results for the set of associated historical search words；

determining a click weight of each of the historical query results for the set of associated historical search words, according to the historical click count of the historical query results for the set of associated historical search words；

displaying, through the user interface, as search results for the current search word, the historical query results for the set of associated historical search words ordered sequentially according to the click weight.
The method according to claim 1, wherein querying the current search word among the historical search data comprises:

matching the current search word with the set of associated historical search words among the historical search data, wherein the set of associated historical search words comprises at least one historical search word of the current search word； and

using the historical query results associated with each of the associated historical search words as the search results associated with the current search word.
The method according to claim 1, wherein determining the click weight of each of the associated historical search words according to the historical click count of each historical query result comprises:

accessing the historical click count of each historical query result for the associated historical search words from the historical search data； and

multiplying the historical click count of each historical query result by an associated click weight coefficient, and thus obtaining the click weight of each associated historical search word associated with the current search word.
The method according to claim 1, wherein displaying the historical query results for the set of associated historical search words ordered sequentially according to the click weight comprises:

ranking said each associated historical query result for the set of associated historical search words in descending order according to the click weight, and thus obtaining a set of descending search results；

displaying the first N number of the descending search results sequentially in a first display area of the user interface；

displaying, sequentially, other search results besides the first N number of the descending search results in a second display area of the user interface.
The method according to claim 4, further comprising:

detecting that a particular search result in the first N number of descending search results displayed in the first area has not been clicked more n times, and in response:

resetting the click weight of the particular search result to 0.
The method according to claim 1, further comprising, after displaying the historical query results for the set of associated historical search words ordered sequentially according to the click weight as the search results for the current search word:

detecting that one of the search results is clicked, and in response:

incrementing the click weight of the clicked search result by a first weight； and

decrementing the click weight of one or more non-clicked search results displayed in the user interface by a second weight.
The method according to claim 1, wherein recording the historical search data comprises:

recording the current search word as a historical search word；

upon obtaining the search results for current search word, recording the search results as historical query results for the current search word；

if detecting that any of the search results are clicked, recording the click count of a clicked search result in the corresponding historical query result for the current search word.
A device for displaying search results, comprising:

a saving module configured to record historical search data, wherein the historical search data comprises historical search words, one or more historical query results associated with each of the historical search words, and a historical click count of each historical query result；

a processing module configured to:

obtain a current search word； and

query the current search word among the historical search data to obtain a set of associated historical search words that are associated with the current search word； and

a determining module configured to:

identify historical query results for the set of associated historical search words； and

determine a click weight of each of the historical query results for the set of associated historical search words, according to the historical click count of the historical query results for the set of associated historical search words； and

a display processing module configured to display, as search results for the current search word, the historical query results for the set of associated historical search words ordered sequentially according to click weight.
The device according to claim 8, wherein the processing module comprises:

a match module configured to match the current search word with the set of historical search words among said historical search data according to a matching condition；

a querying module configured to use the historical query results associated with each of the associated historical search words as search results associated with the current search word.
The device according to claim 9, wherein said determining module comprises:

a reading module configured to access the historical click count of each historical query result for the set of associated historical search words from the historical search data； and

an acquisition module configured to multiply the historical click count of each historical query result by an associated click weight coefficient to obtain the click weight of each historical query result associated with said current search word.
The device according to claim 8, wherein said display processing module comprises:

a ranking module configured to rank each historical query result associated with the current search word in descending order according to the click weight, and thus obtain a set of descending search results；

a displaying module configured to display the first N number of the descending search results sequentially in a first display area of a display menu, and display other search results besides the first N number of the descending search results in a second display area of the display menu.
The device according to claim 1, further comprising:

a setting module configured to:

detect that a particular search result in the first N number of the descending search results has been displayed in the first display area and not clicked more than a non-click threshold number of times； and, in response:

reduce the click weight of the particular search result.
The device according to claim 8, further comprising:

a detecting module configured to detect whether one of the search results displayed is clicked； and

a computing module configure to, upon detecting that one of the search results is clicked, incrementing the click weight of the clicked search results by a first weight and decrementing the click weight of one or more non-clicked search results by a second weight.
The device according to claim 8, wherein said saving module comprises:

a first saving submodule configured to record the current search word inputted by the user as a historical search word；

a second saving submodule configured to record the search results for the current search word as historical query results associated with the current search word；

a third saving submodule configured to record the click count of a clicked search result in the corresponding historical query result upon detecting that any of the search results are clicked.