WO2014032470A1 - 一种搜索方法、装置和存储介质 - Google Patents

一种搜索方法、装置和存储介质 Download PDF

Info

Publication number
WO2014032470A1
WO2014032470A1 PCT/CN2013/078426 CN2013078426W WO2014032470A1 WO 2014032470 A1 WO2014032470 A1 WO 2014032470A1 CN 2013078426 W CN2013078426 W CN 2013078426W WO 2014032470 A1 WO2014032470 A1 WO 2014032470A1
Authority
WO
WIPO (PCT)
Prior art keywords
search
search keyword
suggested
keyword
suggested search
Prior art date
Application number
PCT/CN2013/078426
Other languages
English (en)
French (fr)
Inventor
曲先洋
耿方圆
禹荣凌
岑植旺
谢玲
Original Assignee
深圳市世纪光速信息技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 深圳市世纪光速信息技术有限公司 filed Critical 深圳市世纪光速信息技术有限公司
Publication of WO2014032470A1 publication Critical patent/WO2014032470A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/02Input arrangements using manually operated switches, e.g. using keyboards or dials
    • G06F3/023Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
    • G06F3/0233Character input methods
    • G06F3/0237Character input methods using prediction or retrieval techniques
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/951Indexing; Web crawling techniques

Definitions

  • search engines as high-impact Internet information processing services, are increasingly being used by users to find the information they need.
  • the main interaction process between the search engine and the user is as follows: the user inputs a search keyword in an input box provided by the search engine; the search engine receives user input, and follows the user The input will be displayed to the user as a search suggestion related to the user input, so as to assist the user to complete the selection and input of the search keyword, the user can input the search keyword by himself or by providing the search engine.
  • the search keyword selects the search keyword to complete the input of the search keyword; when the search engine confirms that the user completes the input of the search keyword, or receives the search instruction issued by the user after completing the input of the search keyword, the search engine
  • a specific computer program is used to collect information such as web pages related to search keywords input by the user on the Internet, and to organize and process the information, and finally display information such as web pages related to the search keywords to the user.
  • a main object of embodiments of the present invention is to provide a search method, apparatus, and storage medium, which can effectively improve information search efficiency.
  • a search method, running on a search device comprising:
  • a search result corresponding to the predicted target search keyword is presented.
  • a search device comprising:
  • a receiving unit configured to receive at least one character input by a user
  • a prediction unit configured to predict, according to the at least one character received by the receiving unit, a target search keyword of the user
  • a display unit configured to display a search result corresponding to the target search keyword predicted by the prediction unit.
  • a search result corresponding to the predicted target search keyword is presented.
  • the search method and apparatus provided in the embodiments of the present invention and the storage medium including the computer executable instructions can actively predict the target search keyword of the user according to at least one character currently input by the user and directly display the target search keyword.
  • the search result therefore, the user can input only one or several characters corresponding to the desired target search keyword, and directly obtain the search result corresponding to the target search keyword, without the user inputting the target search keyword completely.
  • FIG. 1 is a flowchart of a search method according to an embodiment of the present invention.
  • FIG. 2 is a flowchart of a search method according to an embodiment of the present invention.
  • FIG. 3 is a structural block diagram of a search apparatus according to an embodiment of the present invention.
  • FIG. 4 is a structural block diagram of a prediction unit of a search device according to an embodiment of the present invention.
  • FIG. 5 is another structural block diagram of a prediction unit of a search apparatus according to an embodiment of the present invention. detailed description
  • Target search keyword The content that the user wants to input in the input box when searching, and the user wants to use the input content to describe the information he wants to find.
  • Search efficiency Indicates the time it takes for the user to start entering the target search keyword until the desired information is found.
  • the embodiment of the invention provides a search method and device.
  • the user searches for information, and can provide search results for the user without requiring the user to complete the input of the target search keyword, thereby effectively improving the search efficiency.
  • the search method provided by the embodiment of the present invention may be performed by a search device, which may be a device having a search input function, such as a desktop computer, a tablet computer, a mobile phone or a portable laptop computer, which is not limited by the present invention.
  • the search device can also be any device with search engine functionality.
  • the search method provided by the embodiment of the present invention includes the following steps 101-103:
  • the 101 Receives at least one character input by a user.
  • the user When the user needs to search for the desired information, the user will enter the character corresponding to the desired target search keyword in the input box provided by the search device.
  • the user can input the characters of the target search keyword through the input device, and the input device includes: a keyboard, a mouse, a virtual keyboard, a voice input device, or a body-sensing input device, etc., which is not limited by the present invention, and may be any other input device. .
  • the at least one character may be a part of a character or all characters of the target search keyword, and the at least one character may be a Chinese character, a number, an English word, a letter, a symbol, a pinyin, etc., of course, a Chinese character, a number, and an English character.
  • the at least one character may be one or several Chinese characters, one or several numbers, one or several English words, one or several letters, one or several symbols, and may also be Chinese characters, numbers, English words. Any combination of letters, symbols, etc.
  • the at least one character may be "country”, “national day”, “national day”, “national holiday”, etc.; when the target search keyword is "Na t iona l Day hol idays", the at least one character may be “Na ti “, “Na t iona l Day”, “Na t iona l Day hol idays", etc.; when the target search keyword is "10086” "At least one character may be "10" or "10086".
  • the target search keyword is "3G mobile phone
  • the at least one character may be “3", “3G”, “3G mobile phone” or the like.
  • the embodiment of the present invention does not limit at least one character corresponding to the target search keyword, and may be any format, such as the above-mentioned Chinese characters, numbers, English words, letters, symbols, and the like, and any combination thereof. .
  • the user does not have to input all the characters of the target search keyword, but can input only a part of the characters of the target search keyword.
  • the search device Predicts the target search keyword of the user according to at least one character input by the user.
  • the search device receives at least one character input by the user, the prediction of the target search keyword of the user is performed based on the received at least one character.
  • the search device can predict the target search keyword of the user in various ways, for example, by communicating with a server (eg, a search server) to obtain a target search keyword from the server suggesting the search keyword and thereby predicting the user. Or, instead of obtaining the suggested search keyword from the server, the method of locally obtaining the suggested search keyword for prediction is used.
  • the search keyword is suggested to be a word closely related to at least one character input by the user according to at least one character input by the user, and the relationship closely means that at least one character input by the user is part of the suggested search keyword.
  • it is recommended to start the search keyword.
  • the search keyword since there are many words containing at least one character input by the user, it is recommended that the search keyword also meet certain conditions. For example, it is recommended that the search keyword be searched the most frequently. The invention does not limit the conditions it satisfies.
  • search keyword may be "Taobao”, “Taobao Mall", “Taobao Women” and the like.
  • search keyword may also be referred to as a search suggestion.
  • the search device communicates with the server to obtain a way to suggest search keywords and thereby predict the user's target search keywords.
  • this step can be performed as follows:
  • the search device transmits at least one character input by the user to the server, and then the search device receives at least one suggested search keyword sent by the server and sent by the server.
  • the search device transmits the character currently input by the user to the server by communicating with the server.
  • the server obtains at least one suggested search keyword according to the at least one character.
  • the server retrieves the index library according to the at least one character to obtain one or more suggested search keywords corresponding to the at least one character.
  • the server sends the obtained suggested search keyword to the search device, and the recommended search keyword sent to the search device is at least one.
  • the recommended search keyword sent by the server to the search device may be 10 article.
  • the search means selects the suggested search keyword from the received at least one suggested search keyword as the predicted target search keyword of the user.
  • the search device predicts the target search keyword according to the suggested search keyword, and the embodiment of the present invention predicts the target search according to the suggested search keyword.
  • the keywords are not limited, and the search device can use multiple prediction methods.
  • the search device may directly predict the suggested search keyword as the target search keyword. And when the search device receives at least two suggested search keywords from the server:
  • the searching device may predict the first suggested search keyword ranked as the target search keyword according to the order in which the server sends the suggested search keyword.
  • the search device in a case that the search device receives at least two suggested search keywords from the server, the search device separately calculates the reliability of the at least two suggested search keywords. a parameter value, and then, according to the calculated reliability parameter value, selecting a suggested search keyword from the at least two suggested search keywords as the predicted target search keyword of the user.
  • the credibility parameter value of each suggested search keyword is used to reflect the possibility of the suggested search keyword as the target search keyword, that is, the degree of trustworthiness, if a suggested search keyword is trusted
  • the higher the degree parameter value the higher the probability that it is the target search keyword.
  • the reliability parameter value can be calculated according to a preset rule.
  • the confidence parameter value can be a function of the number of searches. The higher the number of searches, the larger the value of the confidence parameter.
  • the search means calculates the reliability parameter value of each of the suggested search keywords based on the number of searches of the at least two suggested search keywords, respectively.
  • the credibility parameter value may be a function of the number of clicks of the search result.
  • the search device will search for the keyword according to the at least two suggestions respectively.
  • Search result clicks calculate the credibility parameter value of each of the suggested search keywords.
  • the credibility parameter value may also be determined by both the number of searches and the number of clicks of the search result, for example, a weighted average of the two, at which time the search device will search for the search keyword based on the at least two suggestions respectively.
  • the number of times and the number of search result clicks of the at least two suggested search keywords, and the credibility parameter value of each of the suggested search keywords is calculated.
  • the number of searches for a word refers to the number of times the word is searched as a search keyword, and specifically refers to the number of times the word is searched by the current user or all users in a certain network range as search keywords within a predetermined time.
  • the search result clicks of a word refer to the number of times the word is clicked in the search result, specifically the number of times the word is clicked by the current user or all users in a certain network range in the search result within a specified time.
  • the present invention does not limit the "predetermined time” and the "certain network range".
  • the "prescribed time” may be half a month, one month, eight months, one year, etc.
  • "a certain network range” may be a server.
  • Network local local area where the user is located Net and so on. For example, when the user searches for "National Day”, that is, when the search keyword is "National Day”, the number of searches for "National Day” will increase once, and in the search results of "National Day”, “National Day Holiday” This information will also be displayed as a result of the above-mentioned "National Day” search. If the user finally clicks and enters the "National Day Holiday” link, the number of clicks on the "National Day Holiday” search result will increase once.
  • the search device can calculate the credibility parameter value according to the number of searches, the number of clicks of the search result, or both, and thereby select a plurality of suggested search keywords to predict one of the predicted search keywords, Increase the accuracy of the forecast.
  • the credibility parameter value can be calculated in combination with the current hotspot and the number of searches. For example, during the Olympics, when the user inputs the character "Olympic", since the Olympics is the current hot content, the suggested search keyword corresponding to "Olympic" In the case of the same number of searches, the "Olympic" can be preferentially predicted as the target search keyword.
  • the search device After calculating the credibility parameter value of the suggested search keyword, the search device selects the suggested search keyword from the received at least two suggested search keywords as the predicted location according to the calculated credibility parameter value.
  • the target search keyword of the user After calculating the credibility parameter value of the suggested search keyword, the search device selects the suggested search keyword from the received at least two suggested search keywords as the predicted location according to the calculated credibility parameter value.
  • the target search keyword of the user After calculating the credibility parameter value of the suggested search keyword, the search device selects the suggested search keyword from the received at least two suggested search keywords as the predicted location according to the calculated credibility parameter value.
  • the search means may predict a suggested search keyword having the highest value of the trust parameter among the at least two suggested search keywords as the target search keyword of the user.
  • the highest credibility parameter value means that the suggested search keyword is the most credible, ie
  • the recommended search keyword is the most likely to be the target search keyword, and the search device predicts the search keyword by using the recommended search keyword with the highest credibility parameter value as the target search keyword of the user, thereby effectively improving the predicted search key.
  • the word is the possibility of the target search keyword, so that the target search keyword provided to the user is more accurate.
  • the recommendation search keyword with the highest credibility parameter value as the target search keyword for example, the difference between the credibility parameter values of the two suggested search keywords is small to near.
  • the suggested search keyword whose trust factor parameter value is slightly smaller but the current user searches more frequently may also be predicted as the target search keyword.
  • a credibility threshold is preset, and after the search device calculates the credibility parameter value of the suggested search keyword, the at least the The credibility parameter values of the suggested search keywords with the highest credibility parameter values in the two suggested search keywords are compared with a predetermined threshold, and the credibility parameters of the suggested search keywords with the highest credibility parameter values are When the value is greater than or equal to the threshold, the suggested search keyword with the highest credibility parameter value is predicted as the target search keyword of the user, otherwise, the search device determines that the suggested search keywords are not the target search of the user. Keywords, the search device will wait for the user's continued input, thereby repeating the above prediction process until the predicted user target search keyword is obtained.
  • the search device By presetting the threshold for judging whether the suggested search keyword can be predicted as the target search keyword, the search device only recommends the highest credibility parameter value when the highest credibility parameter value is greater than or equal to the threshold value.
  • the search keyword prediction is a target search keyword, thereby excluding a search keyword whose credibility parameter value is not high enough, and the reliability is insufficient, so that the accuracy of the predicted target search keyword can be further improved.
  • selecting a suggested search keyword from at least two suggested search keywords as the predicted target search keyword is not limited to the above manner of passing the reliability parameter value.
  • the preference parameter value is used to reflect the value of the user's preference for a certain preference, which may be a user's preference or habit of sports, music, news, games, and the like.
  • the preference search parameter value is used to predict the target search keyword
  • the user's favorite preference parameter values are set in advance
  • the sports preference parameter value is 5
  • the music preference parameter value is 3
  • the suggested search keywords include sports and music keywords
  • the search keyword is predicted as the target search keyword.
  • multiple sports search keywords may be further filtered from other target search keyword prediction methods, such as the above-mentioned credibility. The way the parameter value is.
  • the prediction of the target search keyword is not limited to the above manner of communicating with the server to obtain the suggested search keyword and predicting the target search keyword of the user, and may also use the local target search keyword prediction.
  • Way such as historical search records recorded from local devices (Also referred to as a search log, the historical search record includes a historical search record of the current user or a historical search record of all local users within a predetermined time, etc.), the suggested search keyword is retrieved, and the search keyword is searched according to the retrieved suggestion , predict the current user's target search keyword.
  • the specific target search keyword prediction manner may be the same as the foregoing embodiment.
  • the target search keyword is predicted by comparing the reliability parameter values of the plurality of suggested search keywords, and the details are not described herein.
  • the search device may collect webpage information related to the target search keyword on the Internet, organize the webpage information as a search result, and directly display the webpage information to the user by using an output device.
  • the search device can also obtain the search service based on the target search keyword provided by the server by means of the server, so that the search result corresponding to the target search keyword can be finally obtained through the service, and then the search result is obtained by using an output device or the like. Directly presented to the user without the user having to complete the input of the target keyword, and the user is not required to give an indication of the search.
  • the output device includes: a liquid crystal display, a cathode ray tube display, etc., which is not limited by the present invention, and may be any other output device.
  • the search method provided by the embodiment of the present invention does not require a large amount of time and effort on the process of selecting, inputting, and the like of the search keyword as in the past, but can input only the target search keyword corresponding to the desired target.
  • the search result corresponding to the target search keyword can be directly obtained, and the search device does not need to wait for the user to completely input the target search keyword, and does not need to wait for the user to release the target search keyword input after completion.
  • the search instruction can provide the search result to the user if the user does not completely input the target search keyword, and therefore, the search efficiency can be effectively improved.
  • the search device may return to the step. 101. Continue receiving the user's input, and then predict the target search keyword according to the user's continued input and the previously input character in step 102 until the predicted target search keyword is obtained in step 102.
  • the searching method includes the following steps 201-21 3: 201.
  • the search device receives the character input by the user.
  • the characters may be one or more of the first time input by the user, or one or more characters corresponding to the target search keyword sequentially input by the user.
  • the character input by the user is the first character input by the user for the first time in this step, in this case, for example, when the user
  • the target search keyword is "National Day Holiday Arrangement", at which point the character received by the search device is "Country”.
  • the searching device sends the received character input by the user to the server.
  • the server receives the character sent by the search device, and performs pre-processing on the character input by the user.
  • the server performs pre-processing on the first character. It can be understood that if the server has received the character input by the user before the current reception, the server may only pre-process the character received this time or the server may input the character previously input by the user and the current input. The characters are combined and the pre-processing is performed on the combined character combination.
  • the pre-processing is used to unify the format of characters to improve the recognition rate of the characters, for example, including string normalization, processing of Chinese phonetic transcription, and the like.
  • string normalization refers to the encoding format of the Unicode, for example, the English case is unified;
  • Chinese phonetic transcription refers to the corresponding pinyin letters of the Chinese character.
  • the server retrieves the index library according to the preprocessed characters to obtain a recommended search keyword corresponding to the character, for example, 10 suggested search keywords.
  • the recommended search keyword is only an example, and the recommended search keyword is not limited to 10, and the number of the suggested search keywords can be arbitrarily set by a person skilled in the art in combination with the need.
  • the server searches the index library according to the pre-processed characters to query a plurality of suggested search keywords, and preferably, the server may further perform the weighting processing on the suggested search keywords that are queried, so that the queried suggestions are obtained.
  • the suggested search keywords returned to the search device are selected. In this step, the recommended search keywords returned are 10.
  • the weighting process is for excluding a plurality of duplicated suggested search keywords so that only one piece is saved, thereby ensuring the diversity of the suggested search keywords remaining.
  • “National Day holiday” and “country “The holiday of the holiday” is basically the same. Therefore, the “National Day holiday” is excluded from the exclusion process, and only the “National Day holiday” is retained.
  • the server sends the obtained 10 recommended search keywords to the search device.
  • the search device receives the above 10 suggested search keywords, and requests the server to send information about the number of searches for the 10 suggested search keywords and the number of clicks of the search results.
  • the server receives the request, and sends the information about the number of searches and the number of clicks to the search device.
  • the server may directly send the information of the number of searches and the number of clicks to the search device in the step 205, and send the information of the search times and the number of clicks to the server by the server in the embodiment of the present invention.
  • the method is not limited, and may be transmitted together with other information sent to the search device, or may be separately generated and transmitted.
  • the search device receives the information of the number of searches and the number of clicks of the search result, and calculates the reliability parameter value of each suggested search keyword according to both the number of searches and the number of clicks of the search result.
  • a weighted average value of the search times and the click times may be calculated, and the weighted coefficients of the two are respectively set, and the combined reliability is calculated by combining the two. Parameter value.
  • the search device selects a suggested search keyword with the highest credibility parameter value, compares the credibility parameter value with a preset threshold value, and determines whether the suggested search keyword is predicted as the target search keyword, and if so, Go to step 210, if no, go back to step 201.
  • the searching device compares the size of the credibility parameter values of the ten suggested search keywords, and selects the recommended search keyword with the highest credibility parameter value to compare with the threshold, when the credibility parameter value is greater than When the value is equal to the threshold value, the recommended search keyword with the highest value of the credibility parameter is predicted as the target search keyword. Otherwise, the search device determines that the 10 recommended search keywords are not the target search keyword, and the search device returns to Step 201: Wait for the user to continue inputting, for example, receiving the second character input by the user, thereby repeating the above prediction process until the predicted target search keyword of the user is obtained.
  • the user's target search keyword is "National Day holiday arrangement" as an example.
  • the predicted search keyword is not obtained.
  • the search device returns to step 201, Receiving the second character "Qing" input by the user, repeating the above steps for the "National Day” input at this time, if the predicted target search keyword is not yet obtained, the search device will return to step 201 again, receiving the third input by the user. Characters "sections" until the predicted target search keyword is obtained.
  • the search device requests the server to provide a search result corresponding to the predicted target search keyword.
  • the user's target search keyword is "National Day holiday arrangement" as an example, in the steps
  • the target search keyword obtained in 209 is "National Day Holiday Arrangement", and the search device requests the server to provide search results corresponding to "National Day Holiday Arrangement”.
  • the server receives a request of the search device to retrieve a search result corresponding to the target search keyword.
  • the server sends the retrieved search result to the search device.
  • the search device receives the above search result and presents it to the user.
  • the user's target search keyword is "National Day Holiday Arrangement" as an example, and the search device presents the search result corresponding to the "National Day Holiday Arrangement" to the user.
  • the search device calculates the credibility parameter value of the suggested search keyword by combining the search count and the search result, and further sets a threshold value compared with the credibility parameter value, Only the suggested search keywords larger than the threshold are predicted as the target search keywords, thereby improving the accuracy of the search device predicting target search keywords.
  • an embodiment of the present invention further provides a search apparatus, including:
  • the receiving unit 10 is configured to receive at least one character input by the user
  • the prediction unit 11 is configured to predict a target search keyword of the user according to the at least one character input by the user received by the receiving unit 10;
  • the display unit 12 is configured to display a search result corresponding to the target search keyword predicted by the prediction unit 1 1 .
  • the search device after acquiring the character input by the user, can actively predict the target search keyword of the user according to at least one character currently input by the user, and directly display the search result of the target search keyword. Therefore, the user can directly input one or several characters corresponding to the desired target search keyword, and directly obtain the search result corresponding to the target search keyword, that is, the search device does not need to wait for the user to completely input the target search. Keywords, no need Waiting for the user to perform the search for the search after the target search keyword input is completed, the search structure can be provided to the user if the user does not completely input the target search keyword, and therefore, the search efficiency can be effectively improved.
  • the prediction unit may include: a character sending module 1001, configured to send at least one character input by the user to a server;
  • the module 1002 is configured to receive, by the server, at least one suggested search keyword that is obtained by the server according to the at least one character input by the user sent by the character sending module 1001;
  • the target search keyword prediction module 1003 is configured to select a suggested search keyword from the at least one suggested search keyword received by the suggested search keyword acquisition module 1002 as the predicted target search keyword of the user.
  • the search keyword generated by the search keyword acquiring module 1002 is at least two suggested search keywords sent by the server;
  • the target search keyword prediction module 1003 may be specifically configured to: separately calculate a reliability parameter value of the at least two suggested search keywords, and according to the calculated reliability parameter value, from the at least two Among the suggested search keywords, the suggested search keyword is selected as the predicted target search keyword of the user.
  • the target search keyword prediction module 003 may calculate a credibility parameter value of each of the suggested search keywords according to the search times of the at least two suggested search keywords, where the recommended search is performed. The more the number of search times of the keyword, the higher the value of the reliability parameter; the target search keyword prediction module 1003 may further calculate each of the clicks according to the number of search results of the at least two suggested search keywords.
  • the target search keyword prediction module 1003 may also be respectively configured according to The number of searches of the at least two suggested search keywords and the search result clicks of the at least two suggested search keywords, and the credibility parameter values of each of the suggested search keywords are calculated.
  • the target search keyword prediction module 1003 may be specifically configured to: predict a suggested search keyword with the highest value of the trust parameter among the at least two suggested search keywords as the user's target Or searching for a keyword of the recommended search keyword having the highest value of the reliability parameter among the at least two suggested search keywords; and comparing the value of the reliability parameter with the predetermined threshold; When the credibility parameter value of the highest recommended search keyword is greater than or equal to the threshold value, the recommended search keyword with the highest credibility parameter value is predicted as the target search keyword of the user.
  • the predicting unit may further include: a suggested search keyword acquiring module 2001, configured to learn from the local history according to at least one character input by the user Searching for at least one suggested search keyword in the search record;
  • the target search keyword prediction module 2002 is configured to select a suggested search keyword from the retrieved at least one suggested search keyword as the predicted target search keyword of the user.
  • the specific target search keyword prediction manner may be the same as the embodiment described above with reference to FIG. 4, for example, by comparing the reliability parameter values of the plurality of suggested search keywords to predict the target search keyword, and details are not described herein again.
  • the search device provided by the embodiment of the present invention may be a computer, a tablet computer, a mobile phone or a portable mobile computer, etc., which is not limited by the present invention.
  • the embodiment of the present invention further provides a storage medium containing computer executable instructions for executing a search method when executed by a computer processor, the method comprising:
  • a search result corresponding to the predicted target search keyword is presented.
  • the present invention can be implemented by software and necessary general hardware, and can also be implemented by hardware, but in many cases, the former is a better implementation. .
  • the technical solution of the present invention which is essential or contributes to the prior art, may be embodied in the form of a software product, which may be stored in a computer readable storage medium, such as a floppy disk of a computer.
  • Read-Only Memory (ROM), Random Access Memory (RAM), Flash memory (FLASH), hard disk or optical disk, etc. includes instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to perform the methods described in various embodiments of the present invention.
  • each unit and module included is only divided according to functional logic, but is not limited to the above division, as long as the corresponding function can be implemented;
  • the specific names of the units are also for convenience of distinction from each other and are not intended to limit the scope of the present invention.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Human Computer Interaction (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本发明公开了一种搜索方法、装置和存储介质,所述搜索方法包括:接收用户输入的至少一个字符;根据所述用户输入的至少一个字符,预测所述用户的目标搜索关键词;展现与所述预测的目标搜索关键词对应的搜索结果。本发明能够根据用户当前输入的至少一个字符对用户的目标搜索关键词主动地予以预测并直接展现该目标搜索关键词的搜索结果,因此,可以有效提高搜索效率。

Description

一种搜索方法、 装置和存储介质 本专利申请要求于 2012年 8月 28 日提交的、 申请号为 201210310396.0、 申请人为腾讯科技(深圳)有限公司、 发明名称为 "一种搜索方法和装置" 的 中国专利申请的优先权, 该申请的全文以引用的方式并入本申请中。 技术领域 本发明涉及网络技术领域, 尤其涉及一种搜索方法、 装置和存储介质。 背景技术
随着互联网的发展, 搜索引擎作为高渗透率的互联网信息处理服务, 开始 越来越多地被用户用以查找需要的信息。
现有技术中, 当用户通过搜索引擎进行搜索时, 搜索引擎和用户之间的主 要交互过程如下: 用户在搜索引擎提供的输入框中输入搜索关键词; 搜索引擎 接收用户输入, 并随着用户的输入将与用户输入相关的建议搜索关键词作为搜 索建议来展示给用户, 以此辅助用户完成搜索关键词的选取和输入, 用户可以 自行将搜索关键词输入完整, 也可以通过在搜索引擎提供的搜索建议中选取搜 索关键词从而完成搜索关键词的输入; 在搜索引擎确认用户完成了搜索关键词 的输入时, 或者收到用户在完成搜索关键词的输入后下达的搜索指示时, 搜索 引擎按照既定策略、 运用特定的计算机程序搜集互联网上与用户输入的搜索关 键词相关的网页等信息, 并对其进行组织和加工, 最终把与搜索关键词相关的 网页等信息显示给用户。
但是, 在上述交互过程中, 用户对搜索关键词的选取、 输入等过程都要耗 费用户的时间和精力, 从而导致搜索效率较低。 发明内容
本发明实施例的主要目的在于, 提供一种搜索方法、 装置和存储介质, 能 够有效提高信息搜索效率。
本发明实施例釆用如下技术方案: 一种搜索方法, 运行于搜索装置, 包括:
接收用户输入的至少一个字符;
根据所述至少一个字符, 预测所述用户的目标搜索关键词;
展现与所述预测的目标搜索关键词对应的搜索结果。
一种搜索装置, 包括:
接收单元, 用于接收用户输入的至少一个字符;
预测单元, 用于根据所述接收单元接收的所述至少一个字符, 预测所述用 户的目标搜索关键词;
显示单元, 用于展现与所述预测单元预测的目标搜索关键词对应的搜索结 果。
一种包含计算机可执行指令的存储介质, 所述计算机可执行指令在由计算 机处理器执行时用于执行一种搜索方法, 该方法包括:
接收用户输入的至少一个字符;
根据所述至少一个字符, 预测所述用户的目标搜索关键词;
展现与所述预测的目标搜索关键词对应的搜索结果。
本发明实施例中提供的搜索方法、 装置和包含计算机可执行指令的存储介 质, 能够根据用户当前输入的至少一个字符对用户的目标搜索关键词主动地予 以预测并直接展现关于该目标搜索关键词的搜索结果, 因此, 用户能够只输入 与希望的目标搜索关键词对应的一个或几个字符, 即可直接获得与该目标搜索 关键词对应的搜索结果, 不需要用户完整输入目标搜索关键词, 也不需要用户 在目标搜索关键词输入完成后下达的进行搜索的指示, 也就是说, 能够在用户 未完整输入目标搜索关键词的情况下即可为用户提供搜索结果, 因此, 可以有 效提高搜索效率。
附图说明
为了更清楚地说明本发明实施例中的技术方案, 下面将对实施例描述中所 需使用的附图作简单地介绍, 当然, 以下描述中的附图仅仅是本发明的一些实 施例, 对于本领域普通技术人员来讲, 在不付出创造性劳动的前提下, 还可以 对这些附图进行修改和替换。
图 1为本发明实施例的一种搜索方法的流程图;
图 2为本发明实施例的一种搜索方法的流程图;
图 3为本发明实施例的一种搜索装置的结构框图;
图 4为本发明实施例的一种搜索装置的预测单元的结构框图;
图 5为本发明实施例的搜索装置的预测单元的另一种结构框图。 具体实施方式
下面将结合附图对本发明实施例中的技术方案进行清楚、 完整地描述, 显 然, 所描述的实施例是本发明的一部分实施例, 而不是全部的实施例, 是为了 阐述本发明的原理, 而不是要将本发明限制于这些具体的实施例。 基于本发明 中的实施例, 本领域普通技术人员在没有做出创造性劳动前提下所获得的所有 其他实施例, 都属于本发明保护的范围。
为了使本领域的技术人员更好的理解本发明的技术方案, 首先对本发明实 施例中所涉及的部分技术术语进行简要介绍, 具体如下:
目标搜索关键词: 用户进行搜索时希望在输入框中输入的内容, 用户希望 使用该输入内容来描述其想要查找的信息。
搜索效率: 表示从用户开始输入目标搜索关键词直至查找到所需信息, 所 花费的时间。
基于上述介绍, 以下对本发明实施例进行说明。
本发明实施例提供一种搜索方法和装置, 由用户进行信息搜索, 能够在不 需要用户完成目标搜索关键词的输入的情况下, 为用户提供搜索结果, 从而有 效提高搜索效率。
本发明实施例提供的搜索方法可由搜索装置执行, 所述搜索装置可以为台 式计算机、 平板电脑、 手机或便携式手提电脑等具有搜索输入功能的装置, 本 发明对此不做限定。 所述搜索装置也可以是任意具有搜索引擎功能的装置。
如图 1所示, 本发明实施例提供的搜索方法包括以下步骤 101-103 :
101 , 接收用户输入的至少一个字符。 当用户需要搜索想要的信息时, 用户将在搜索装置提供的输入框中输入与 其希望的目标搜索关键词对应的字符。 用户可通过输入装置输入目标搜索关键 词的字符, 所述输入装置包括: 键盘、 鼠标、 虚拟键盘、 声控输入装置或者体 感输入装置等, 本发明对此不做限定, 还可以是其它任意输入装置。
本步骤中, 所述至少一个字符可以是目标搜索关键词的一部分字符或全部 字符, 该至少一个字符可以为汉字、 数字、 英文单词、 字母、 符号、 拼音等, 当然还可以汉字、 数字、 英文单词、 字母、符号、拼音等中的至少两种的组合。 具体而言, 该至少一个字符可以是一个或几个汉字、 一个或几个数字、 一个或 几个英文单字、 一个或几个字母、 一个或几个符号, 还可以由汉字、 数字、 英 文单词、 字母、 符号等任意组合而成。
举例而言, 例如当目标搜索关键词为 "国庆节假期" 时, 所述至少一个字 符可以是"国"、 "国庆"、 "国庆节 " "国庆节假"等;当目标搜索关键词为 "Na t iona l Day hol idays " 时, 所述至少一个字符可以是 "Na t i " 、 "Na t iona l Day" 、 "Na t iona l Day hol idays "等; 当目标搜索关键词为 "10086 " 时, 所述至少 一个字符可以是 "10" 、 "10086 " 当目标搜索关键词为 "3G手机" 时, 所述至 少一个字符可以是 "3 " 、 "3G " , "3G手机" 等。
可以理解的是, 本发明实施例对所述目标搜索关键词对应的至少一个字符 不做限定, 其可以是任何格式, 例如上述的汉字、 数字、 英文单词、 字母、 符 号等以及它们的任意组合。
本步骤中, 用户可不必输入目标搜索关键词的全部字符, 而是可以仅输入 目标搜索关键词的一部分字符。
102 ,根据所述用户输入的至少一个字符,预测所述用户的目标搜索关键词。 当搜索装置接收到用户输入的至少一个字符时, 将根据所接收的至少一个 字符, 进行用户的目标搜索关键词的预测。 具体而言, 搜索装置可以釆用多种 方式对用户的目标搜索关键词予以预测, 例如, 与服务器 (例如搜索服务器) 通信以获得来自服务器建议搜索关键词并以此预测用户的目标搜索关键词; 或 者, 不从服务器获取建议搜索关键词, 而釆用本地获取建议搜索关键词进行预 测的方式等。 其中, 顾名思义, 建议搜索关键词是根据用户输入的至少一个字符而获得 的与用户输入的至少一个字符关系紧密的词语, 所述关系紧密是指用户输入的 至少一个字符是建议搜索关键词的一部分,通常是建议搜索关键词的开始字符, 而且, 由于包含用户输入的至少一个字符的词语非常多, 建议搜索关键词还需 满足一定的条件, 例如, 建议搜索关键词被搜索的次数最多, 本发明对其满足 的条件不做限定。 举例而言, 如果用户输入的至少一个字符为 "淘", 建议搜索 关键词可以是 "淘宝网"、 "淘宝商城"、 "淘宝网女装" 等。 下文中, 建议搜索 关键词也可称为搜索建议。
可选的, 在本发明的一个实施例中, 搜索装置釆用与服务器通信来获得建 议搜索关键词并以此预测用户的目标搜索关键词的方式。 在这种情况下, 本步 骤可按照如下方式执行:
首先, 搜索装置将所述用户输入的至少一个字符发送给服务器, 然后, 搜 索装置接收该服务器发送的、 所述服务器获取的至少一条建议搜索关键词。
具体而言, 搜索装置通过与服务器的通信, 将用户当前输入的字符发送给 服务器。 服务器接收到用户输入的至少一个字符后, 根据该至少一个字符而获 取至少一条建议搜索关键词。 具体的, 服务器根据该至少一个字符, 检索索引 库来获取对应于该至少一个字符的一条或多条建议搜索关键词。 服务器将获取 的建议搜索关键词发送给搜索装置, 发送给搜索装置的建议搜索关键词为至少 一条, 例如, 在本发明的一个实施例中, 服务器发送给搜索装置的建议搜索关 键词可以为 10条。
接下来, 搜索装置将从所述接收的至少一条建议搜索关键词中选择建议搜 索关键词作为预测的所述用户的目标搜索关键词。
具体而言, 在搜索装置接收到所述服务器发送的建议搜索关键词之后, 搜 索装置将根据建议搜索关键词来预测目标搜索关键词, 本发明实施例对如何根 据建议搜索关键词来预测目标搜索关键词不做限定, 搜索装置可以釆用多种预 测方式。
举例而言, 当搜索装置从服务器接收的建议搜索关键词仅为一条时, 搜索 装置可以直接地将该建议搜索关键词预测为目标搜索关键词。 而当搜索装置从服务器接收的建议搜索关键词为至少两条时:
可选的, 在本发明的一个实施例中, 搜索装置可以按照服务器发送建议搜 索关键词的排列顺序, 将排列在前的第一条建议搜索关键词预测为目标搜索关 键词。
可选的, 在本发明的另一个实施例中, 在搜索装置从服务器接收的建议搜 索关键词为至少两条的情况下, 搜索装置分别计算所述至少两条建议搜索关键 词的可信度参数值, 然后, 根据所述计算的可信度参数值, 从所述至少两条建 议搜索关键词中选择建议搜索关键词作为预测的所述用户的目标搜索关键词。
其中, 每条建议搜索关键词的可信度参数值用于反映该建议搜索关键词作 为所述目标搜索关键词的可能性的大小, 即可以信赖的程度, 如果一条建议搜 索关键词的可信度参数值越高, 即表明其是目标搜索关键词的可能性越高。 可 信度参数值可以根据预先设定的规则计算出。 例如, 可信度参数值可以是搜索 次数的函数, 搜索次数越高, 可信度参数值越大。 此时, 搜索装置将分别根据 所述至少两条建议搜索关键词的搜索次数, 计算每条所述建议搜索关键词的可 信度参数值。 再例如, 可信度参数值可以是搜索结果点击次数的函数, 搜索结 果点击次数越高, 可信度参数值越高, 此时, 搜索装置将分别根据所述至少两 条建议搜索关键词的搜索结果点击次数, 计算每条所述建议搜索关键词的可信 度参数值。 当然, 可信度参数值还可以由搜索次数和搜索结果点击次数两者而 决定, 例如, 两者的加权平均值, 此时, 搜索装置将分别根据所述至少两条建 议搜索关键词的搜索次数和所述至少两条建议搜索关键词的搜索结果点击次数, 计算每条所述建议搜索关键词的可信度参数值。
需要说明的是, 一个词语的搜索次数是指该词语被作为搜索关键词进行搜 索的次数, 具体是指规定时间内该词语被当前用户或者一定网络范围内所有用 户作为搜索关键词进行搜索的次数; 而一个词语的搜索结果点击次数是指该词 语在搜索结果中被点击的次数, 具体是指规定时间内该词语在搜索结果中被当 前用户或者一定网络范围内所有用户点击的次数。其中,本发明对"规定时间" 以及 "一定网络范围" 不做限定, 例如 "规定时间" 可以是半个月、 一个月、 八个月、 及一年等, "一定网络范围" 可以是服务器网络、 用户所在的本地局域 网等。举例说明, 当用户搜索 "国庆节" 时, 即搜索关键词为 "国庆节"时, "国 庆节" 的搜索次数将增加一次, 而在 "国庆节" 的搜索结果中, "国庆节假期" 这条信息也会被作为上述 "国庆节" 的搜索结果而展现, 如果此时用户最终点 击并进入了 "国庆节假期" 这个链接, 则 "国庆节假期" 的搜索结果点击次数 将增加一次。
由于搜索装置可以根据搜索次数、 搜索结果点击次数或结合这二者来计算 可信度参数值, 并由此对多条建议搜索关键词加以 选以将其中一条预测为目 标搜索关键词, 因此能够增大预测的准确性。
而且, 应当理解, 本领域技术人员可以选择多种计算建议搜索关键词的可 信度参数值的方式, 并不限于该实施例釆用的单独地根据上述搜索次数、 搜索 结果点击次数或结合这二者的方式, 本发明对此不做限定。 举例而言, 可以结 合时下热点和搜索次数计算可信度参数值,例如在奥运会期间,当用户输入 "奥" 这个字符时, 由于奥运会为时下热点内容, 在 "奥" 对应的建议搜索关键词的 搜索次数相同的情况下, 可优先将 "奥运会" 预测为目标搜索关键词。
在计算出建议搜索关键词的可信度参数值后 , 搜索装置将根据所述计算的 可信度参数值, 从接收到的至少两条建议搜索关键词中选择建议搜索关键词作 为预测的所述用户的目标搜索关键词。
例如, 搜索装置可以将所述至少两条建议搜索关键词中可信度参数值最高 的建议搜索关键词预测为所述用户的目标搜索关键词。 如前所述, 由于建议搜 索关键词的可信度参数值反映了该建议搜索关键词作为目标搜索关键词的可能 性, 可信度参数值最高意味着该建议搜索关键词最可信, 即该建议搜索关键词 为目标搜索关键词的可能性最高, 搜索装置通过将可信度参数值最高的建议搜 索关键词预测为用户的目标搜索关键词, 由此, 可以有效地提高预测的搜索关 键词为目标搜索关键词的可能性,从而使提供给用户的目标搜索关键词更准确。
当然, 可以理解的是, 不必一定将可信度参数值最高的建议搜索关键词预 测为目标搜索关键词, 例如在两条建议搜索关键词的可信度参数值的差值很小 到几近可忽略的情况下, 也可以将可信度参数值稍小但当前用户近期较常搜索 的建议搜索关键词预测为目标搜索关键词。 进一步的, 在本发明的一个实施例中, 为了提高预测的准确性, 预先设定 了可信度阔值, 搜索装置在计算出建议搜索关键词的可信度参数值后, 将所述 至少两条建议搜索关键词中可信度参数值最高的建议搜索关键词的可信度参数 值与预定的阈值进行比较, 当所述可信度参数值最高的建议搜索关键词的可信 度参数值大于等于所述阔值时, 将所述可信度参数值最高的建议搜索关键词预 测为所述用户的目标搜索关键词, 否则, 搜索装置判断这些建议搜索关键词均 不是用户的目标搜索关键词, 搜索装置将等待用户的继续输入, 从而重复以上 预测过程, 直到获得预测的用户目标搜索关键词为止。
通过预先设定用于判断是否可以将建议搜索关键词预测为目标搜索关键词 的阔值, 搜索装置仅在最高的可信度参数值大于等于阔值时将该可信度参数值 最高的建议搜索关键词预测为目标搜索关键词, 由此排除可信度参数值不够高、 即可信度不够的搜索关键词, 因此可以进一步提高预测的目标搜索关键词的准 确性。
可以理解的是, 从至少两条建议搜索关键词中选择建议搜索关键词作为预 测的目标搜索关键词, 即对多条建议搜索关键词的筛选, 并不限于上述通过可 信度参数值的方式, 例如, 基于用户喜好度参数值的方式。 举例而言, 所述喜 好度参数值用于反映用户对某种喜好的喜爱程度的值, 所述喜好可以是体育、 音乐、 新闻、 游戏等用户的嗜好或习惯。 举例而言, 当使用喜好度参数值来预 测目标搜索关键词时, 通过预先设定用户的各喜好的喜好度参数值, 例如, 体 育的喜好度参数值为 5、 音乐的喜好度参数值为 3 , 在建议搜索关键词中包括体 育方面和音乐方面的关键词时, 由于体育方面的搜索关键词的喜好度参数值高 于音乐方面的搜索关键词的喜好度参数值, 由此将体育方面的搜索关键词预测 为目标搜索关键词。 而当存在多个相同的喜好度参数值的建议搜索关键词时, 例如多个体育方面的搜索关键词, 可以结合其他的目标搜索关键词预测方式进 一步从中进行筛选, 例如上述的通过可信度参数值的方式。
可以理解的是, 本步骤中, 目标搜索关键词的预测不限于上述与服务器通 信以获得建议搜索关键词并以此预测用户的目标搜索关键词的方式, 也可以釆 用本地目标搜索关键词预测的方式, 例如从本地设备中所记录的历史搜索记录 (也可称为搜索日志, 该历史搜索记录包括当前用户的历史搜索记录或者规定 时间内所有本地用户的历史搜索记录等) 中检索出建议搜索关键词, 并根据所 检索出的建议搜索关键词, 预测当前用户的目标搜索关键词。 具体的目标搜索 关键词预测方式可以与前述实施例相同, 例如通过比较多条建议搜索关键词的 可信度参数值等方式来预测目标搜索关键词, 这里不再赘述。
103 , 展现与所述预测的目标搜索关键词对应的搜索结果。
具体而言, 搜索装置在预测目标搜索关键词后, 可以自行搜集互联网上的 与目标搜索关键词相关的网页信息, 将这些网页信息进行组织加工后作为搜索 结果, 利用输出装置直接展现给用户。 当然, 搜索装置也可以借助于服务器, 即获得服务器提供的基于该目标搜索关键词的搜索服务, 从而可通过该服务最 终获得与目标搜索关键词对应的搜索结果, 然后利用输出装置等将搜索结果直 接展现给用户, 而不需要用户完成目标关键词的输入, 也不要用户下达进行搜 索的指示。 其中, 所述输出装置包括: 液晶显示器、 阴极射线管显示器等, 本 发明对此不做限定, 还可以是其它任意输出装置。
综上可知, 本发明实施例提供的搜索方法, 用户不会如以往那样在搜索关 键词的选取、 输入等过程上耗费大量的时间和精力, 而是能够只输入与希望的 目标搜索关键词对应的一个或几个字符, 即可直接获得与该目标搜索关键词对 应的搜索结果, 搜索装置不需要等待用户完整输入目标搜索关键词, 也不需要 等待用户在目标搜索关键词输入完成后下达的进行搜索的指示, 能够在用户未 完整输入目标搜索关键词的情况下即可为用户提供搜索结果, 因此, 可以有效 提高搜索效率。
需要说明的是, 在本发明实施例提供的搜索方法中, 如果在步骤 102 中, 搜索装置通过步骤 101 中接收的用户输入的至少一个字符无法预测目标搜索关 键词时, 搜索装置可返回至步骤 101 , 继续接收用户的输入, 然后在步骤 102 中根据用户继续输入的和在先输入的字符, 预测目标搜索关键词, 直到在步骤 102中获得预测的目标搜索关键词。
下面参照图 2 ,通过如下实施例对本发明提供的搜索方法进行进一步的详细 说明。 本实施例中, 所述搜索方法包括以下步骤 201-21 3: 201、 搜索装置接收到用户输入的的字符。
具体而言, 所述字符可以是用户首次输入的一个或多个、 也可以是用户接 下来顺序输入的一个或多个对应于目标搜索关键词的字符。
本实施例中, 为了使本领域技术人员更好的理解本实施例的技术方案, 假 设本步骤中, 用户输入的字符为用户首次输入的第一个字符, 这种情况下, 例 如, 当用户的目标搜索关键词是 "国庆节假期安排", 此时搜索装置接收的字符 为 "国"。
202、 搜索装置将接收的用户输入的字符发送给服务器。
203、 服务器接收搜索装置发送的字符, 并对用户输入的字符执行预处理。 本步骤中, 由于在本发明实施例中, 用户输入的字符为与期望的目标搜索 关键词对应的首个字符, 因此, 服务器即对该首个字符执行预处理。 可以理解 的是, 如果在本次接收前, 服务器已经接收过用户在先输入的字符, 服务器可 仅对本次接收的字符进行预处理或者服务器可将用户在先输入的字符和本次接 收的字符结合并对结合后的字符组合执行预处理。
具体而言, 所述预处理用于使字符的格式统一以提高对该字符的识别率, 例如包括字符串归一化、 及对汉字注音的处理等。 具体而言, 字符串归一化是 指统一字符的编码格式, 例如英文的大小写统一等; 对汉字注音是指对汉字标 注其对应的拼音字母。
204、 服务器根据经预处理后的字符, 检索索引库以获取对应于该字符的建 议搜索关键词, 例如 10条建议搜索关键词。
可以理解的是, 建议搜索关键词为 10条仅为示例, 建议搜索关键词不限于 10条, 本领域技术人员可结合需要任意地设定该建议搜索关键词的条数。
具体而言, 服务器根据预处理后的字符检索索引库从而从中查询出多条建 议搜索关键词, 优选的, 服务器可进一步对查询到的建议搜索关键词进行排重 处理, 从而在查询到的建议搜索关键词中选取出返回给搜索装置的建议搜索关 键词, 本步骤中, 返回的建议搜索关键词为 10条。
所述排重处理用于排除多条内容重复的建议搜索关键词使其仅存一条, 以 此来保证存留的建议搜索关键词具有多样性。 举例而言, "国庆节放假" 和 "国 庆节假期" 这两个词意思基本完全一致, 因此在该排重处理排除这二者之中的 "国庆节放假 ", 仅保留有 "国庆节假期"。
205、 服务器将获取的 10条建议搜索关键词发送给搜索装置。
206、 搜索装置接收上述 10条建议搜索关键词, 并请求该服务器发送关于 该 10条建议搜索关键词的搜索次数和搜索结果点击次数的信息。
207、 服务器接收上述请求, 向搜索装置发送上述搜索次数和点击次数的信 息。
当然, 可以理解的是, 也可以省略上述步骤 206、 207。 此时, 例如可使服 务器在上述步骤 205 中直接将搜索次数和点击次数的信息与该建议搜索关键词 同时共同发送给搜索装置, 本发明实施例对服务器对搜索次数和点击次数的信 息的发送方式不做限定, 其可以与其他发送给搜索装置的信息共同发送, 也可 以单独生成信息来进行发送。
208、 搜索装置接收搜索次数和搜索结果点击次数的信息, 根据搜索次数和 搜索结果点击次数两者计算出各建议搜索关键词的可信度参数值。
具体而言, 在本发明实施例中, 可计算上述搜索次数和点击次数二者的加 权平均值, 通过分别设置该二者的加权系数, 来结合这二者计算出综合后的可 信度参数值。
209、 搜索装置选择可信度参数值最大的建议搜索关键词, 将其可信度参数 值与预设的阔值比较, 判断是否将该建议搜索关键词预测为目标搜索关键词, 如果是, 执行步骤 210, 如果否, 返回至步骤 201。
具体而言,搜索装置比较上述 10条建议搜索关键词的可信度参数值的大小, 选出可信度参数值最大的建议搜索关键词与阔值进行比较, 当该可信度参数值 大于等于阔值时, 将该条可信度参数值最高的建议搜索关键词预测为目标搜索 关键词, 否则, 搜索装置判断这 10条建议搜索关键词均不是目标搜索关键词, 搜索装置将返回至步骤 201 , 等待用户的继续输入, 例如,接收用户输入的第二 个字符,从而重复上述预测过程,直到获得预测的用户的目标搜索关键词为止。
同样以用户的目标搜索关键词为 "国庆节假期安排" 为例, 经过上述步骤 202-209没有获得预测的搜索关键词, 该情况下, 搜索装置返回至步骤 201 , 接 收用户输入的第二个字符 "庆", 针对此时输入的 "国庆" 重复上述步骤, 如还 不能获得预测的目标搜索关键词,搜索装置将再次返回至步骤 201 ,接收用户输 入的第三个字符 "节", 直到获得预测的目标搜索关键词。
210、搜索装置请求该服务器提供与预测的目标搜索关键词对应的搜索结果。 例如, 同样以用户的目标搜索关键词为 "国庆节假期安排" 为例, 在步骤
209中获得预测的目标搜索关键词为 "国庆节假期安排", 搜索装置请求该服务 器提供与 "国庆节假期安排" 对应的搜索结果。
211、 服务器接收搜索装置的请求检索与目标搜索关键词对应的搜索结果。
212、 服务器将检索到的搜索结果发送给搜索装置。
21 3、 搜索装置接收上述搜索结果, 并将其展现给用户。
例如, 同样以用户的目标搜索关键词为 "国庆节假期安排" 为例, 搜索装 置将对应于 "国庆节假期安排" 的搜索结果展现给用户。
根据本发明的该实施例的搜索方法, 搜索装置通过结合搜索次数和搜索结 果来计算出建议搜索关键词的可信度参数值, 并进一步设置了与该可信度参数 值比较的阔值, 仅将大于该阔值的建议搜索关键词预测为目标搜索关键词, 由 此提高了搜索装置预测目标搜索关键词的准确性。
与前文的搜索方法的实施例相对应, 如图 3所示, 本发明的实施例还提供 了一种搜索装置, 包括:
接收单元 10 , 用于接收用户输入的至少一个字符;
预测单元 11 ,用于根据所述接收单元 10接收的所述用户输入的至少一个字 符, 预测所述用户的目标搜索关键词;
显示单元 12 ,用于展现与所述预测单元 1 1预测的目标搜索关键词对应的搜 索结果。
本发明实施例提供的搜索装置, 在获取了用户输入的字符后, 能够根据用 户当前输入的至少一个字符对用户的目标搜索关键词主动地予以预测并直接展 现该目标搜索关键词的搜索结果, 因此, 用户能够只输入与希望的目标搜索关 键词对应的一个或几个字符, 即可直接获得与该目标搜索关键词对应的搜索结 果, 也就是说, 搜索装置不需要等待用户完整输入目标搜索关键词, 也不需要 等待用户在目标搜索关键词输入完成后下达的进行搜索的指示, 能够在用户未 完整输入目标搜索关键词的情况下即可为用户提供搜索结构, 因此, 可以有效 提高搜索效率。
具体的, 在本发明的一个实施例中, 如图 4所示, 所述预测单元可包括: 字符发送模块 1001 , 用于将所述用户输入的至少一个字符发送给服务器; 建议搜索关键词获取模块 1002 , 用于接收所述服务器发送的、 所述服务器 根据所述字符发送模块 1001发送的所述用户输入的至少一个字符而获取的至少 一条建议搜索关键词;
目标搜索关键词预测模块 1003 ,用于从所述建议搜索关键词获取模块 1002 接收的至少一条建议搜索关键词中选择建议搜索关键词作为预测的所述用户的 目标搜索关键词。
可选的, 在本发明的一个实施例中, 建议搜索关键词获取模块 1002所接收 的、 所述服务器发送的建议搜索关键词为至少两条;
此时, 目标搜索关键词预测模块 1003可具体用于: 分别计算所述至少两条 建议搜索关键词的可信度参数值, 并根据所述计算的可信度参数值, 从所述至 少两条建议搜索关键词中选择建议搜索关键词作为预测的所述用户的目标搜索 关键词。
可选的, 所述目标搜索关键词预测模块 1 003可以分别根据所述至少两条建 议搜索关键词的搜索次数, 计算每条所述建议搜索关键词的可信度参数值, 所 述建议搜索关键词的搜索次数越多则其可信度参数值越高; 所述目标搜索关键 词预测模块 1003还可以分别根据所述至少两条建议搜索关键词的搜索结果点击 次数, 计算每条所述建议搜索关键词的可信度参数值, 所述建议搜索关键词的 搜索结果点击次数越多则所述可信度参数值越高; 当然, 所述目标搜索关键词 预测模块 1003还可以分别根据所述至少两条建议搜索关键词的搜索次数和所述 至少两条建议搜索关键词的搜索结果点击次数, 计算每条所述建议搜索关键词 的可信度参数值。
进一步的, 所述目标搜索关键词预测模块 1003可具体用于: 将所述至少两 条建议搜索关键词中可信度参数值最高的建议搜索关键词预测为所述用户的目 标搜索关键词; 或者, 将所述至少两条建议搜索关键词中可信度参数值最高的 建议搜索关键词的可信度参数值与预定的阈值进行比较; 当所述可信度参数值 最高的建议搜索关键词的可信度参数值大于等于所述阔值时, 将所述可信度参 数值最高的建议搜索关键词预测为所述用户的目标搜索关键词。
可选的,在本发明的一个实施例中,如图 5所示,所述预测单元还可包括: 建议搜索关键词获取模块 2001 , 用于根据所述用户输入的至少一个字符, 从本地历史搜索记录中检索出至少一条建议搜索关键词;
目标搜索关键词预测模块 2002 , 用于从所述检索出的至少一条建议搜索关 键词中选择建议搜索关键词作为预测的所述用户的目标搜索关键词。
具体的目标搜索关键词预测方式可以与前结合图 4所述实施例相同, 例如 通过比较多条建议搜索关键词的可信度参数值等方式来预测目标搜索关键词, 这里不再赘述。
本发明实施例提供的搜索装置可以为电脑、 平板电脑、 手机或便携式手提 电脑等, 本发明对此不做限定。
本发明实施例中各单元的更具体的工作方式可参见前文描述的搜索方法, 这里不再赘述。
本发明实施例还提供一种包含计算机可执行指令的存储介质, 所述计算机 可执行指令在由计算机处理器执行时用于执行一种搜索方法, 其特征在于, 该 方法包括:
接收用户输入的至少一个字符;
根据所述至少一个字符, 预测所述用户的目标搜索关键词;
展现与所述预测的目标搜索关键词对应的搜索结果。
通过以上关于实施方式的描述, 所属领域的技术人员可以清楚地了解到, 本发明可借助软件及必需的通用硬件来实现, 当然也可以通过硬件实现, 但很 多情况下前者是更佳的实施方式。 基于这样的理解, 本发明的技术方案本质上 或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来, 该计算机 软件产品可以存储在计算机可读存储介质中, 如计算机的软盘、 只读存储器 ( Read-Only Memory, ROM )、随机存取存 4诸器( Random Access Memory, RAM )、 闪存(FLASH )、 硬盘或光盘等, 包括若干指令用以使得一台计算机设备(可以 是个人计算机,服务器,或者网络设备等)执行本发明各个实施例所述的方法。
值得注意的是, 上述搜索装置的实施例中, 所包括的各个单元和模块只是 按照功能逻辑进行划分的, 但并不局限于上述的划分, 只要能够实现相应的功 能即可; 另外, 各功能单元的具体名称也只是为了便于相互区分, 并不用于限 制本发明的保护范围。
以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围并不局限于 此, 任何熟悉本技术领域的技术人员在本发明披露的技术范围内, 可轻易想到 的变化或替换, 都应涵盖在本发明的保护范围之内。 因此, 本发明的保护范围 应以所述权利要求的保护范围为准。

Claims

WO 2014/032470 权 利 要 书 PCT/CN2013/078426
1、 一种搜索方法, 运行于搜索装置, 其特征在于, 包括:
接收用户输入的至少一个字符;
根据所述至少一个字符, 预测所述用户的目标搜索关键词;
展现与所述预测的目标搜索关键词对应的搜索结果。
2、 根据权利要求 1所述的方法, 其特征在于,
所述根据所述至少一个字符, 预测所述用户的目标搜索关键词包括: 将所述至少一个字符发送给服务器;
接收所述服务器发送的、 所述服务器根据所述至少一个字符而获取的至少 一条建议搜索关键词;
从所述接收的至少一条建议搜索关键词中选择其一作为所述预测的目标 搜索关键词。
3、 根据权利要求 1所述的方法, 其特征在于,
所述根据所述至少一个字符, 预测所述用户的目标搜索关键词包括: 根据所述至少一个字符, 从本地历史搜索记录中检索出至少一条建议搜索 关键词;
从检索出的至少一条建议搜索关键词中选择其一作为所述预测的目标搜 索关键词。
4、 根据权利要求 2或 3所述的方法, 其特征在于,
所述接收的至少一条建议搜索关键词或所述检索出的至少一条建议检索 关键词为至少两条建议搜索关键词;
所述从所述接收的至少一条建议搜索关键词中选择其一作为预测的目标 搜索关键词或者从所述检索出的至少一条建议搜索关键词中选择其一作为所 述预测的目标搜索关键词包括:
分别计算所述至少两条建议搜索关键词的可信度参数值;
根据计算的可信度参数值, 从所述至少两条建议搜索关键词中选择其一作 为所述预测的目标搜索关键词。
5、 根据权利要求 4所述的方法, 其特征在于,
所述根据计算的可信度参数值, 从所述至少两条建议搜索关键词中选择其 一作为所述预测的目标搜索关键词包括:
将所述至少两条建议搜索关键词中可信度参数值最高的建议搜索关键词 作为所述预测的目标搜索关键词; 或者
将所述至少两条建议搜索关键词中可信度参数值最高的建议搜索关键词 的可信度参数值与预定的阔值进行比较; 当所述可信度参数值最高的建议搜索 关键词的可信度参数值大于等于所述阔值时, 将所述可信度参数值最高的建议 搜索关键词作为所述预测的目标搜索关键词。
6、 根据权利要求 4或 5所述的方法, 其特征在于,
所述分别计算所述至少两条建议搜索关键词的可信度参数值包括: 分别根据所述至少两条建议搜索关键词的搜索次数, 计算每条所述建议搜 索关键词的可信度参数值, 所述建议搜索关键词的搜索次数越多则其可信度参 数值越高; 或者
分别根据所述至少两条建议搜索关键词的搜索结果点击次数, 计算每条所 述建议搜索关键词的可信度参数值, 所述建议搜索关键词的搜索结果点击次数 越多则其可信度参数值越高; 或者
分别根据所述至少两条建议搜索关键词的搜索次数和所述至少两条建议 搜索关键词的搜索结果点击次数, 计算每条所述建议搜索关键词的可信度参数 值。
7、 一种搜索装置, 其特征在于, 包括:
接收单元, 用于接收用户输入的至少一个字符;
预测单元, 用于根据所述接收单元接收的所述至少一个字符, 预测所述用 户的目标搜索关键词;
显示单元, 用于展现与所述预测单元预测的目标搜索关键词对应的搜索结 果。
8、 根据权利要求 7所述的装置, 其特征在于,
所述预测单元包括:
字符发送模块, 用于将所述至少一个字符发送给服务器;
建议搜索关键词获取模块, 用于接收所述服务器发送的、 所述服务器根据 所述字符发送模块发送的所述至少一个字符而获取的至少一条建议搜索关键 词;
目标搜索关键词预测模块, 用于从所述建议搜索关键词获取模块获取的至 少一条建议搜索关键词中选择其一作为所述预测的目标搜索关键词。
9、 根据权利要求 7所述的装置, 其特征在于,
所述预测单元包括:
建议搜索关键词获取模块, 用于根据所述至少一个字符, 从本地历史搜索 记录中检索出至少一条建议搜索关键词;
目标搜索关键词预测模块, 用于从检索出的至少一条建议搜索关键词中选 择其一作为所述预测的目标搜索关键词。
10、 根据权利要求 8或 9所述的装置, 其特征在于,
所述接收的至少一条建议搜索关键词或所述检索出的至少一条建议检索 关键词为至少两条建议搜索关键词;
所述目标搜索关键词预测模块用于:
分别计算所述至少两条建议搜索关键词的可信度参数值;
根据计算的可信度参数值, 从所述至少两条建议搜索关键词中选择其一作 为所述预测的目标搜索关键词。
11、 根据权利要求 10所述的装置, 其特征在于,
所述目标搜索关键词预测模块用于:
将所述至少两条建议搜索关键词中可信度参数值最高的建议搜索关键词 作为所述预测的目标搜索关键词; 或者
将所述至少两条建议搜索关键词中可信度参数值最高的建议搜索关键词 的可信度参数值与预定的阔值进行比较; 当所述可信度参数值最高的建议搜索 关键词的可信度参数值大于等于所述阔值时, 将所述可信度参数值最高的建议 搜索关键词作为所述预测的目标搜索关键词。
12、 根据权利要求 10或 11所述的装置, 其特征在于,
所述目标搜索关键词预测模块用于:
分别根据所述至少两条建议搜索关键词的搜索次数, 计算每条所述建议搜 索关键词的可信度参数值, 所述建议搜索关键词的搜索次数越多则其可信度参 数值越高; 或者
分别根据所述至少两条建议搜索关键词的搜索结果点击次数, 计算每条所 述建议搜索关键词的可信度参数值, 所述建议搜索关键词的搜索结果点击次数 越多则所述可信度参数值越高; 或者
分别根据所述至少两条建议搜索关键词的搜索次数和所述至少两条建议 搜索关键词的搜索结果点击次数, 计算每条所述建议搜索关键词的可信度参数 值。
13、 一种包含计算机可执行指令的存储介质, 所述计算机可执行指令在由 计算机处理器执行时用于执行一种搜索方法, 其特征在于, 该方法包括: 接收用户输入的至少一个字符;
根据所述至少一个字符, 预测所述用户的目标搜索关键词;
展现与所述预测的目标搜索关键词对应的搜索结果。
PCT/CN2013/078426 2012-08-28 2013-06-28 一种搜索方法、装置和存储介质 WO2014032470A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201210310396.0 2012-08-28
CN201210310396.0A CN103631826A (zh) 2012-08-28 2012-08-28 一种搜索方法和装置

Publications (1)

Publication Number Publication Date
WO2014032470A1 true WO2014032470A1 (zh) 2014-03-06

Family

ID=50182455

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/078426 WO2014032470A1 (zh) 2012-08-28 2013-06-28 一种搜索方法、装置和存储介质

Country Status (2)

Country Link
CN (1) CN103631826A (zh)
WO (1) WO2014032470A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106611029B (zh) * 2015-10-27 2020-03-03 北京国双科技有限公司 提高网站站内搜索效率的方法和装置
CN107193818A (zh) * 2016-03-14 2017-09-22 百度在线网络技术(北京)有限公司 一种搜索方法、装置和终端设备
CN105824975A (zh) * 2016-04-19 2016-08-03 百度在线网络技术(北京)有限公司 搜索推荐方法和装置
CN110502699B (zh) * 2019-08-27 2022-06-03 百度在线网络技术(北京)有限公司 用于获取信息的方法和装置
CN113468235B (zh) * 2021-05-31 2023-05-09 北京达佳互联信息技术有限公司 信息获取方法、装置、服务器及存储介质

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100169341A1 (en) * 2008-12-30 2010-07-01 Ebay Inc. Predictive algorithm for search box auto-complete
CN102368252A (zh) * 2010-09-30 2012-03-07 微软公司 将搜索查询应用到内容集

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6564213B1 (en) * 2000-04-18 2003-05-13 Amazon.Com, Inc. Search query autocompletion
US7836044B2 (en) * 2004-06-22 2010-11-16 Google Inc. Anticipated query generation and processing in a search engine
CN1588367A (zh) * 2004-09-02 2005-03-02 魏新成 搜索栏输入字母提示搜索词和网名

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100169341A1 (en) * 2008-12-30 2010-07-01 Ebay Inc. Predictive algorithm for search box auto-complete
CN102368252A (zh) * 2010-09-30 2012-03-07 微软公司 将搜索查询应用到内容集

Also Published As

Publication number Publication date
CN103631826A (zh) 2014-03-12

Similar Documents

Publication Publication Date Title
JP5745627B2 (ja) 予測的なクエリ提案のキャッシュ
RU2618910C2 (ru) Способ и устройство для отображения информации
TWI492075B (zh) 在對話期間提供建議的方法,使用所述方法的電子裝置,以及電腦程式產品
US8825694B2 (en) Mobile device retrieval and navigation
US11494427B2 (en) Generating a personal database entry for a user based on natural language user interface input of the user and generating output based on the entry in response to further natural language user interface input of the user
US11693863B1 (en) Query completions
US9910932B2 (en) System and method for completing a user query and for providing a query response
US20140040741A1 (en) Smart Auto-Completion
CN106462613B (zh) 基于用户属性来对建议进行排名
TW201229790A (en) Information providing method and device
US20120059814A1 (en) Methods and apparatus for selecting a search engine to which to provide a search query
WO2014032470A1 (zh) 一种搜索方法、装置和存储介质
WO2014014374A1 (en) Search query suggestions based in part on a prior search
US20130339380A1 (en) Providing query suggestions
JP2012008770A (ja) 情報処理装置および情報処理方法
KR101681851B1 (ko) 웹 브라우저 동작 방법 및 시스템
US9576077B2 (en) Generating and displaying media content search results on a computing device
RU2589855C2 (ru) Способ, система и компьютерная программа для управления информацией из социальных сетей
CN110659406B (zh) 搜索方法及装置
RU2744111C2 (ru) Способ и система для формирования подсказок по расширению поисковых запросов в поисковой системе
JP2012194783A (ja) アプリケーションマーケットにおいて使用されるサーバ、通信端末、システム及びgui決定方法
JP5777663B2 (ja) 検索支援装置及び検索支援プログラム
EP4354323A1 (en) Method and apparatus for searching point of information
US20150154681A1 (en) Enriching product catalog with product name keywords
KR20240104206A (ko) 비도큐멘트 아이템의 지원형 검색

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13833541

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13833541

Country of ref document: EP

Kind code of ref document: A1