WO2015055094A1 - 筛选条件提供方法、装置和搜索方法、装置 - Google Patents

筛选条件提供方法、装置和搜索方法、装置 Download PDF

Info

Publication number
WO2015055094A1
WO2015055094A1 PCT/CN2014/088306 CN2014088306W WO2015055094A1 WO 2015055094 A1 WO2015055094 A1 WO 2015055094A1 CN 2014088306 W CN2014088306 W CN 2014088306W WO 2015055094 A1 WO2015055094 A1 WO 2015055094A1
Authority
WO
WIPO (PCT)
Prior art keywords
target object
user
screening
information
screening condition
Prior art date
Application number
PCT/CN2014/088306
Other languages
English (en)
French (fr)
Inventor
韩博
Original Assignee
北京百度网讯科技有限公司
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 北京百度网讯科技有限公司 filed Critical 北京百度网讯科技有限公司
Publication of WO2015055094A1 publication Critical patent/WO2015055094A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/951—Indexing; Web crawling techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation

Definitions

  • the present invention relates to the field of Internet technologies, and in particular, to a screening condition providing method, apparatus, and searching method and apparatus.
  • the inventors have found that the prior art has at least the following problems: since the user's demand for goods or services is personalized, the common screening conditions are mostly fixed parameters of goods or services, and cannot be provided with users.
  • the real demand has a sufficiently relevant screening condition, so it is difficult to obtain a screening result that meets the user's real needs.
  • the user does not even know which features of the goods or services should be searched and screened, resulting in screening conditions based on fixed parameter settings that create obstacles for the user to obtain the selected objects. Therefore, the existing fixed mode screening conditions cannot match the user's personalized requirements, the screening is relatively inefficient, and the user experience is poor.
  • the present invention aims to solve at least one of the above technical problems.
  • a first object of the present invention is to provide a screening condition providing method.
  • the method can provide personalized screening conditions for various commodities or services, thereby allowing users to perform more accurate and quick screening from the perspective of requirements, so that the screening results are more personalized and the user experience is improved.
  • a second object of the present invention is to provide a screening condition providing device.
  • a third object of the present invention is to propose a search method.
  • a fourth object of the present invention is to propose a search device.
  • the screening condition providing method of the first aspect of the present invention includes the following steps: obtaining Correlation information corresponding to the target object; establishing a focus point of the target object according to the association information; generating at least one filter condition corresponding to the target object according to the attention point, and providing at least a part of the at least one filter condition to the user.
  • the screening condition providing method establishes a focus point of the target object by acquiring associated information corresponding to the target object, and generates a screening condition to be provided to the user for selection, thereby allowing the user to filter the condition based on the personalized design from the perspective of the requirement.
  • the correlation between the screening conditions and the real needs of the users is increased, the screening is more precise and faster, the screening results are more personalized, and the user experience is improved.
  • the screening condition providing apparatus of the second aspect of the present invention includes: an association information acquiring module, configured to acquire association information corresponding to the target object, and a point of interest establishing module, configured to establish a focus point of the target object according to the association information.
  • a screening condition generating module configured to generate at least one screening condition corresponding to the target object according to the point of interest, and provide at least a part of the at least one screening condition to the user.
  • the screening condition providing apparatus establishes a focus point of the target object by acquiring the associated information corresponding to the target object, and generates a screening condition to be provided to the user for selection, so that the screening condition is closer to the real needs of the user, so that the screening is performed. More accurate and faster, the screening results are more personalized and enhance the user experience.
  • a search method of an embodiment of the third aspect of the present invention includes the steps of: receiving a search request; generating at least one filter condition associated with a target object for which the search request is targeted according to the search request; and generating at least one filter condition Part of it is provided to the user.
  • the search method of the embodiment of the present invention acquires the target object information according to the user's search request to establish a corresponding focus point, and generates a filter condition for the user to perform a search selection, thereby allowing the user to design the search condition based on the personalized view from the demand perspective. Make more accurate and fast search, make the search more personalized and enhance the user experience.
  • the search apparatus of the fourth aspect of the present invention includes: a search module, configured to receive a search request; and a generating module, configured to generate, according to the search request, at least one screening condition associated with the target object that the search request is associated with, And providing at least a portion of the at least one screening condition to the user.
  • the searching device of the embodiment of the present invention acquires the target object information according to the user's search request to establish a corresponding point of interest, and generates a filter condition for the user to perform a search selection, thereby allowing the user to design the search condition based on the personalized view from the demand perspective. Make more accurate and fast search, make the search more personalized and enhance the user experience.
  • FIG. 1 is a flow chart of a method for providing a screening condition according to an embodiment of the present invention
  • FIG. 2(a) is a schematic diagram of a screening subject and screening conditions corresponding thereto according to an embodiment of the present invention
  • FIG. 2(b) is a schematic diagram of a screening subject and screening conditions corresponding thereto according to still another embodiment of the present invention.
  • FIG. 2(c) is a schematic diagram showing a screening subject and screening conditions corresponding thereto according to another embodiment of the present invention.
  • FIG. 2(d) is a schematic diagram of a screening subject and screening conditions corresponding thereto according to still another embodiment of the present invention.
  • FIG. 3 is a flow chart of a method for providing a screening condition according to still another embodiment of the present invention.
  • FIG. 4 is a flow chart of a method for providing a screening condition according to another embodiment of the present invention.
  • Figure 5 is a block diagram showing the structure of a filter condition providing apparatus according to an embodiment of the present invention.
  • FIG. 6 is a block diagram showing the structure of a screening condition providing apparatus according to still another embodiment of the present invention.
  • FIG. 7 is a flow chart of a search method in accordance with one embodiment of the present invention.
  • FIG. 8 is a flowchart of a search method according to still another embodiment of the present invention.
  • FIG. 9(a) is a flowchart of establishing a point of interest of a target object based on association information, according to an embodiment of the present invention.
  • FIG. 9(b) is a flowchart of generating at least one filter condition corresponding to a target object according to a point of interest, according to an embodiment of the present invention.
  • FIG. 10 is a structural block diagram of a search apparatus according to an embodiment of the present invention.
  • FIG. 11 is a block diagram showing the structure of a search device according to still another embodiment of the present invention.
  • the current screening conditions are mainly based on the screening mode of product classification. Based on the fixed product model, price range and other basic product descriptions, the screening conditions are set, and the user's individualized requirements cannot be matched, which is relatively inefficient and has a poor user experience.
  • the present invention provides a screening condition providing method, including the steps of: acquiring association information corresponding to a target object; establishing a focus point of the target object according to the association information; and generating at least one screening condition corresponding to the target object according to the attention point, And providing at least a portion of the at least one screening condition to the user.
  • FIG. 1 is a flow chart of a method of providing screening conditions in accordance with one embodiment of the present invention.
  • the screening condition providing method includes:
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service.
  • the target object and its associated category can be obtained by any existing or future method or device capable of obtaining user instruction information.
  • the related information corresponding to the target object may be collected through social media such as Baidu, Baidu Post, or other means capable of acquiring the target object and related information.
  • the related information may be a comment information of a user's description, rating, comment, sharing experience, etc. of a certain target object, or may be statistical information such as number of evaluations, number of times of scoring, number of times of sharing, and other types of related information such as users. Feedback after searching, etc.
  • the semantic information may be analyzed by a semantic mining method such as natural language processing or any method capable of implementing semantic mining in the future, and then the analysis result may be filtered to obtain a key related to the characteristics of the target object. Words to extract the user's focus on the target object within the category.
  • a semantic mining method such as natural language processing or any method capable of implementing semantic mining in the future
  • the points of interest are clustered according to the relevance, the heat, the source, and the like of the point of interest, forming a screening theme and a screening condition corresponding thereto, and providing at least one screening condition to the user, so that the user according to at least one screening condition Search.
  • Figures 2(a), 2(b), and 2(c) are respectively three possible embodiments for providing screening conditions in accordance with the method of the present invention.
  • comfort, durability and cleanliness are the screening themes.
  • Each screening theme corresponds to at least one specific screening condition for the user to select.
  • it is light and continuous. Good brushing is the corresponding specific screening conditions.
  • the screening condition providing method of the present invention can also be applied to a search engine, and any scheme, method and apparatus having search and screening requirements, such as prompts for providing search conditions when a user performs a search.
  • personalized screening conditions can be provided while displaying goods or services to the user, such as good pronunciation quality, windshield projection, power saving, and resource occupation. Small.
  • the screening condition providing method of the embodiment of the present invention establishes a focus point of the target object by acquiring related information corresponding to the target object, and generates a screening condition to be provided to the user for selection, thereby allowing the user to filter the condition based on the personalized design from the perspective of the requirement.
  • the correlation between the screening conditions and the real needs of the users is increased, the screening is more accurate and faster, the screening results are more personalized, and the user experience is improved.
  • FIG. 3 is a flow chart of a method for providing a screening condition according to still another embodiment of the present invention.
  • the related items are filtered to establish a focus point of the target object, and the attention points are sorted according to the frequency of occurrence of the attention point, thereby generating a screening condition, and querying the search database according to the filtering condition selected by the user to obtain the query result.
  • the screening condition providing method includes the following steps:
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service.
  • the brand list corresponding to each category and the corresponding specific target object model list are established according to the target object, for example, the basketball shoes include brands such as Nike, Adidas, and the corresponding specific basketball shoe model list.
  • the related information corresponding to the target object may be collected through social media such as Baidu, Baidu Post, or other means capable of acquiring the target object and related information.
  • the related information includes comment information, etc., and may be a comment information of a user's description, rating, comment, sharing experience, etc. of a certain target object, or statistical information such as number of evaluations, number of times of scoring, number of shares, and other types. Relevant information such as feedback information after user search.
  • the related item is a comment information such as a word, a short sentence, and the like related to the target object attribute, and the related information may include a plurality of related items. Specifically, each of the related items in the related information is segmented, the language in the related item is decomposed into independent words, and then the part of the separated words is labeled, for example, the words "noun” and "adjective" are marked.
  • S304 Filter the related items according to the part of speech of the plurality of participles in the related item after the word segmentation, and establish a focus point of the target object according to the frequency of occurrence of the related items after the screening.
  • the related items are filtered according to a specific part of speech matching mode, for example, the word segmentation is selected according to the matching mode of “noun + adjective”, and the words that do not meet the requirements are removed.
  • the combined association result, the obtained screening result is the related item of the phrase containing the form of "noun + adjective", and then
  • the phrase in the selected association item is categorized by synonymous synonym, such as "large screen” and "large screen size” can be classified into the same type description.
  • the frequency statistics of the classified words are obtained, and the high-frequency attention points of the target object are obtained according to the statistical result. For example, the frequency of appearance of the similar words can be sorted from high to low, and the first 10 items are preset as the focus of the target object. .
  • the search database includes points of interest corresponding to multiple brands and multiple brands.
  • the points of interest corresponding to the plurality of brands include the points of interest of the target objects of the respective models corresponding to the plurality of brands, so as to filter out the target objects of the corresponding models by searching for the points of interest.
  • the frequency of occurrence of each point of interest in all related items is counted, wherein the frequency of occurrence of the point of interest is the main basis for evaluating the priorities of different points of interest.
  • S307 Sort the points of interest according to the frequency of occurrence of the point of interest.
  • the points of interest may be sorted according to the frequency of occurrence from high to low.
  • the points of interest are clustered according to the correlation, heat, source, and the like of the attention points of the preset number of bits before the sorting, and the screening subject and the corresponding filtering condition are formed, and at least one screening is performed.
  • Conditions are provided to the user to enable the user to search based on at least one filter condition.
  • Figure 2 (b) shows a possible way to provide screening conditions for the user when selecting a notebook computer, wherein the display, function, experience, and appearance are screening topics, and each screening subject corresponds to at least one specific screening condition. For the user to select, for example, the resolution in FIG. 2(b) is high, the game is smooth, the noise is small, and the fineness is respectively corresponding to the specific screening conditions.
  • the user selects one or several submissions that are required according to the screening theme and the corresponding filtering conditions, and then the receiving end receives the screening condition selected by the user.
  • the focused point is searched in the search database, and the retrieved eligible points of interest are arranged in a preset manner and order, and then the target object corresponding to the retrieved focused point is Return to the user.
  • the screening condition providing method of the embodiment of the present invention generates a screening condition by classifying and analyzing the related items in the related information to filter the user attention points of the target object, and sorting the attention points according to the frequency of occurrence of the attention points, thereby generating a screening condition.
  • the search database is queried according to the filtering condition selected by the user to obtain the corresponding target object, so that the searched target object is more in line with the real needs of the user, and the search process is more intelligent and improves the user experience.
  • FIG. 4 is a flow chart of a method for providing a screening condition according to another embodiment of the present invention.
  • the weight of the corresponding focus point is generated according to the source information of the associated information, so as to improve the sorting priority of the target object with the corresponding comment according to the frequency of occurrence and weight of the attention point, and combine multiple screenings with the user's historical record.
  • the conditions are sorted. Specifically, as shown in FIG. 4, the screening condition providing method includes the following steps.
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service.
  • a list of brands corresponding to each category and a corresponding list of specific target object models are established, such as basketball shoes including brands such as Nike, Adidas, and corresponding specific basketball shoe model lists.
  • related information corresponding to the target object of the product may be collected through social media such as Baidu, Baidu Post Bar, or other vertical industry user comment information capable of acquiring the target object and related information.
  • the related information includes user comment information and/or search information comment information, etc., and may be user's comment information in the form of description, rating, comment, sharing experience, etc. of the target object of the product, or may be the number of evaluations, the number of times of scoring, Statistics such as sharing times, and other types of related information such as feedback after user search.
  • the related item is a comment information such as a word, a short sentence, and the like related to the target object attribute, and the related information may include a plurality of related items. Specifically, each of the related items in the related information is segmented, the language in the related item is decomposed into independent words, and then the part of the separated words is labeled, for example, the words "noun” and "adjective" are marked.
  • S404 Filter the related items according to the part of speech of the plurality of participles in the related item after the word segmentation, and establish a focus point of the target object according to the frequency of occurrence of the related items after the screening.
  • the related items are filtered according to a specific part of speech matching mode, for example, the word segmentation is selected according to the matching mode of “noun + adjective”, and the words that do not meet the requirements are removed.
  • the combined association result, the obtained filtering result is the related item of the phrase containing the form of "noun + adjective", and then the synonymous synonym classification of the phrase in the selected related item, such as "screen large” and "screen size” Big” can be classified into the same type of description and so on.
  • the frequency statistics of the classified words are obtained, and the high-frequency attention points of the target object are obtained according to the statistical result. For example, the frequency of appearance of the similar words can be sorted from high to low, and the first 10 items are preset as the focus of the target object. .
  • the search database includes points of interest corresponding to multiple brands and multiple brands.
  • the points of interest corresponding to the plurality of brands include the points of interest of the target objects of the respective models corresponding to the plurality of brands, so as to filter out the target objects of the corresponding models by searching for the points of interest.
  • the frequency of occurrence of each point of interest in all related items is counted, wherein the frequency of occurrence of the point of interest is the main basis for evaluating the priorities of different points of interest.
  • the source information of the related information is obtained, and the source list of the related information of the user in the vertical industry is established, and basketball shoes are taken as an example.
  • Baidu knows, Baidu Post Bar, e-commerce website and other media accumulate a large number of basketball shoe related information. origin of.
  • S408 Generate a weight corresponding to the attention point according to the source information of the related information, so as to be sorted according to the appearance frequency and the weight of the attention point.
  • the weight of the corresponding point of interest is generated according to the source information of the associated information, and is sorted according to the frequency of occurrence of the point of interest and the comprehensive evaluation of the weights.
  • the ranking priority of the target object with the corresponding point of interest in the review information is enhanced in the search result.
  • the points of interest are clustered according to the correlation, heat, source, and the like of the attention points of the preset number of bits before the sorting, and the screening subject and the corresponding filtering condition are formed, and at least one screening is performed.
  • Conditions are provided to the user to enable the user to search based on at least one filter condition.
  • Figure 2 (b) shows a possible way to provide screening conditions for the user when selecting a notebook computer, wherein the display, function, experience, and appearance are screening topics, and each screening subject corresponds to at least one specific screening condition. For the user to select, for example, the resolution in FIG. 2(b) is high, the game is smooth, the noise is small, and the fineness is respectively corresponding to the specific screening conditions.
  • S410 Acquire a history record of the user, sort the plurality of filter conditions according to the history record of the user, and provide at least a part of the plurality of filter conditions after the sorting to the user.
  • the screening condition when the screening condition is multiple, the history of the user may be obtained, and multiple screening conditions are sorted according to the user's historical record (such as used filtering conditions, browsed target objects, etc.), and after sorting Multiple screening criteria are provided to the user.
  • a part of the screening condition may be selected and provided to the user, and the unprovided part may be directly deleted or hidden in the interactive element in the screening condition providing page, and may be needed by the user (eg, through interaction). Displayed when the operation sends a display command, etc.).
  • S411 Receive a screening condition selected by a user.
  • the user selects one or several submissions that are required according to the screening theme and the corresponding filtering conditions, and then the receiving end receives the screening condition selected by the user.
  • the focus is searched in the search database, and the retrieved The eligible points of interest are arranged in a preset manner and order, and then the target object corresponding to the retrieved point of interest is returned to the user.
  • the screening condition providing method of the embodiment of the present invention establishes a target point of interest and a search database by acquiring related information corresponding to the target object, sorts the attention point according to the frequency of occurrence, the source, and the like, and generates a corresponding target object by the sorting result.
  • Conditions, combined with the user's history to sort multiple filter conditions, provide users to search, so as to provide users with more effective screening conditions, the screening results are more personalized, improve user experience.
  • the present invention also proposes a screening condition providing device.
  • a screening condition providing apparatus includes: an association information acquiring module, configured to acquire association information corresponding to the target object; a point of interest establishing module, configured to establish a focus point of the target object according to the association information; and a screening condition generating module, configured to The point generates at least one filter condition corresponding to the target object, and provides at least a portion of the at least one filter condition to the user.
  • FIG. 5 is a block diagram showing the structure of a filter condition providing apparatus according to an embodiment of the present invention.
  • the screening condition providing apparatus includes: an association information acquiring module 100, a point of interest establishing module 200, and a screening condition generating module 300.
  • the association information acquisition module 100 is configured to acquire association information corresponding to the target object.
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service.
  • the target object and its associated category can be obtained by any existing or future method or device capable of obtaining user instruction information.
  • the associated information obtaining module 100 may collect related information corresponding to the target object through social media such as Baidu, Baidu Post, or other means capable of acquiring the target object and related information.
  • the related information includes comment information, etc., and may be a review information of a user's rating, comment, sharing experience, etc. of a certain target object, or may be statistical information such as number of ratings, number of times of scoring, number of shares, and other types of correlation. Information such as feedback after the user searches.
  • the point of interest establishing module 200 is configured to establish a point of interest of the target object according to the associated information. More specifically, the point of interest establishing module 200 may perform semantic analysis on the related information through a semantic mining method such as natural language processing or a semantic mining method that may occur in the future according to the related information, and then filter the analysis result to obtain a target object attribute. Keyword to extract the user's focus on the target object within the category.
  • a semantic mining method such as natural language processing or a semantic mining method that may occur in the future according to the related information
  • the screening condition generating module 300 is configured to generate at least one screening condition corresponding to the target object according to the attention point, and provide at least a part of the at least one screening condition to the user.
  • the screening condition generating module 300 is further configured to perform statistics on the frequency of occurrence of the attention point, and sort the attention points according to the frequency of occurrence of the attention point, and the attention point according to the preset number of bits before the sorting. Generate at least one filter condition corresponding to the target object.
  • the module 300 clusters the points of interest according to the correlation, the heat, the source, and the like of the points of interest, forms a screening theme and a screening condition corresponding thereto, and provides at least one screening condition to the user, so that the user performs according to at least one screening condition. search for.
  • Figures 2(a), 2(b), and 2(c) are respectively three possible embodiments for providing screening conditions in accordance with the method of the present invention.
  • comfort, durability and cleanliness are the screening themes.
  • Each screening theme corresponds to at least one specific screening condition for the user to select.
  • it is light and continuous. Good brushing is the corresponding specific screening conditions.
  • the screening condition providing apparatus of the embodiment of the present invention establishes a focus point of the target object by acquiring the related information corresponding to the target object, and generates a screening condition to be provided to the user for selection, thereby making the search condition closer to the real needs of the user, and making the screening more Accurate and fast, the screening results are more personalized and enhance the user experience.
  • Figure 6 is a block diagram showing the structure of a filter condition providing apparatus according to still another embodiment of the present invention.
  • the screening condition providing apparatus includes: an associated information acquiring module 100, a focused point establishing module 200, a screening condition generating module 300, a search database establishing module 400, a searching module 500, a word segmentation sub-module 210, a screening sub-module 220, and A sub-module 230 is generated.
  • the focus point building module 200 includes a word segmentation sub-module 210, a screening sub-module 220, and a generation sub-module 230.
  • the search database establishing module 400 is configured to establish a brand list, wherein the brand list includes a plurality of brands, and a search database is separately established for the plurality of brands, and the search database includes a plurality of brands and a plurality of brands corresponding points of interest.
  • the search database establishing module 400 establishes a list of brands corresponding to each category and a corresponding list of specific target object models, such as basketball shoes including brands such as Nike, Adidas, and corresponding specific basketball shoe model lists.
  • the searching module 500 is configured to receive a screening condition selected by the user, and query the search database according to the filtering condition selected by the user to obtain the query result, and provide the query result to the user. More specifically, the user selects one or several submissions that he/she needs according to the screening theme and the corresponding filtering conditions, and then the search module 500 receives the filtering conditions selected by the user, and according to the filtering conditions selected by the user, in the search database. The point of interest is searched, and the retrieved eligible points of interest are arranged in a preset manner and order, and then the target object corresponding to the retrieved point of interest is returned to the user.
  • the screening condition generating module 300 is further configured to acquire source information of the related information, and generate a weight of the corresponding attention point according to the source information of the related information, so as to focus on the frequency of interest according to the appearance frequency and weight of the attention point. Sort.
  • the screening condition generating module 300 collects the frequency of occurrence of the attention point, obtains the source information of the related information, and establishes a source list of the related information of the user in the category, and takes basketball shoes as an example, Baidu knows, Baidu Post Bar, and electricity Media such as the business website are sources of accumulated information on basketball shoes.
  • the weights of the corresponding points of interest are generated according to the source information of the associated information, and are sorted according to the frequency of occurrence of the points of interest and the comprehensive evaluation of the weights.
  • the evaluation information in the search results contains corresponding The sorting priority of the target object of the point of interest.
  • the filter condition generating module 300 is further configured to acquire a history record of the user, sort the plurality of filter conditions according to the history record of the user, and provide at least a part of the plurality of filter conditions after the sorting to the user.
  • the screening condition is multiple, the screening condition generating module 300 acquires the history record of the user, sorts the plurality of filtering conditions according to the user's historical record (such as the used filtering condition, the browsed target object, etc.), and sorts the A number of subsequent screening conditions are provided to the user.
  • a part of the screening condition may be selected and provided to the user, and the unprovided part may be directly deleted or hidden in the interactive element in the screening condition providing page, and may be needed by the user (eg, through interaction). Displayed when the operation sends a display command, etc.).
  • the word segmentation sub-module 210 is configured to perform word segmentation on each of the association items in the association information, and perform part-of-speech tagging on the plurality of word segments in the association item after the segmentation.
  • the related item is a comment information such as a word, a short sentence, and the like related to the target object attribute, and the related information may include a plurality of related items. More specifically, the word segmentation sub-module 210 classifies each related item in the related information, decomposes the language in the related item into independent words, and then labels the part of the separated independent words, for example, "noun", " The adjective "is part of the word.
  • the screening sub-module 220 is configured to filter the related items according to the part of speech of the plurality of participles in the related item after the word segmentation. More specifically, the screening sub-module 220 filters the related items according to the specific part-of-speech matching mode according to the part of speech of the plurality of participles in the related item after the word segmentation, for example, according to the "noun + adjective" mode, the participle is filtered, and the excluding The associated item of the required combination of words, the resulting screening result is the associated item of the phrase containing the form of "noun + adjective".
  • the generating sub-module 230 is configured to establish a target point of the target object according to the frequency of occurrence of the associated item after the screening. Specifically, the generating sub-module 230 performs synonymous synonym categorization on the phrases in the selected related items, such as “large screen” and “large screen size” may be classified into the same type description. Then, the frequency statistics of the classified words are obtained, and the high-frequency attention points of the target object are obtained according to the statistical result. For example, the frequency of appearance of the similar words can be sorted from high to low, and the first 10 items are preset as the focus of the target object. .
  • the screening condition providing device of the embodiment of the present invention establishes a target point of interest and a search database by acquiring related information corresponding to the target object, and sorts the attention point according to the frequency of occurrence, the source, and the like to generate a screening condition, and combines the history of the user.
  • the record sorts multiple screening conditions, provides the user for searching, and then searches the search database according to the filtering condition selected by the user to obtain the corresponding target object, thereby providing the user with screening conditions that are closer to the real needs of the user, and the screening result is more personalized. Improve the user experience.
  • the present invention also proposes a search method.
  • a search method comprising the steps of: receiving a search request; generating, according to the search request, at least one screening condition associated with the target object for which the search request is directed; and providing at least a portion of the at least one filter condition to the user.
  • FIG. 7 is a flow chart of a search method in accordance with one embodiment of the present invention.
  • the search method includes:
  • a search request sent by a user may be received by a search engine or device.
  • the search request should include information such as the target object and its category.
  • S702. Generate at least one screening condition associated with the target object for which the search request is targeted according to the search request.
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service.
  • the target object in the search request and the category thereof are analyzed and processed, the focus of the target object is established, and the points of interest are clustered according to the correlation, heat, source and other information of the point of interest, forming and searching.
  • the filtering subject associated with the target object and the corresponding filtering criteria are requested.
  • Figures 2(a), 2(b), and 2(c) are respectively three possible embodiments for providing search conditions in accordance with the method of the present invention.
  • comfort, durability and cleanliness are the screening themes.
  • Each screening theme corresponds to at least one specific screening condition for the user to select.
  • it is light and continuous.
  • Good brushing is the corresponding specific screening conditions.
  • a part of the screening conditions may be selected and provided to the user according to factors such as the frequency of interest of the screening conditions.
  • the search method of the present invention can be applied to a search engine, as well as any scheme, method and apparatus having search and screening requirements, such as prompts for providing search conditions when a user performs a search.
  • a part of the screening conditions may be selected and provided to the user according to factors such as the frequency of interest of the screening conditions, and the Some of the filter criteria can be hidden in the interaction elements within the filter criteria page and displayed when the user needs (such as sending display instructions via interactive actions).
  • personalized screening conditions can be provided while displaying goods or services to the user, such as good pronunciation quality, windshield projection, power saving, and resource occupation. Small.
  • the search method of the embodiment of the present invention acquires the target object information according to the user's search request to establish a corresponding focus point, and generates a filter condition for the user to perform a search selection, thereby allowing the user to design the search condition based on the personalized view from the demand perspective. Make more accurate and fast search, make the search more personalized and enhance the user experience.
  • FIG. 8 is a flow chart of a search method in accordance with yet another embodiment of the present invention.
  • the search method includes the following steps:
  • a search request sent by a user may be received by a search engine or device.
  • the search request should include information such as the target object and its category.
  • the related information corresponding to the target object may be collected through social media such as Baidu, Baidu Post, or other means capable of acquiring the target object and related information.
  • the related information may be a comment information of a user's description, rating, comment, sharing experience, etc. of a certain target object, or may be statistical information such as number of evaluations, number of times of scoring, number of times of sharing, and other types of related information such as users. Feedback after searching, etc.
  • the semantic information may be analyzed by a semantic mining method such as natural language processing or any method capable of implementing semantic mining in the future, and then the analysis result may be filtered to obtain a key related to the characteristics of the target object. Words to extract the user's focus on the target object within the category. As shown in FIG. 9(a), the focus of establishing the target object according to the associated information further includes:
  • the related item is a comment information such as a word, a short sentence, and the like related to the target object attribute, and the related information may include a plurality of related items. Specifically, each of the related items in the related information is segmented, the language in the related item is decomposed into independent words, and then the part of the separated words is labeled, for example, the words "noun” and "adjective" are marked.
  • S8032 Filter the related items according to the part of speech of the plurality of participles in the related item after the word segmentation, and establish a focus point of the target object according to the frequency of occurrence of the related items after the screening.
  • the related items are filtered according to a specific part of speech matching mode, for example, the word segmentation is selected according to the matching mode of “noun + adjective”, and the words that do not meet the requirements are removed.
  • the combined association result, the obtained filtering result is the related item of the phrase containing the form of "noun + adjective", and then the synonymous synonym classification of the phrase in the selected related item, such as "screen large” and "screen size” Big” can be classified into the same type of description and so on.
  • the frequency statistics of the classified words are obtained, and the high-frequency attention points of the target object are obtained according to the statistical result. For example, the frequency of appearance of the similar words can be sorted from high to low, and the first 10 items are preset as the focus of the target object. .
  • S804 Generate at least one screening condition corresponding to the target object according to the attention point, and provide the user.
  • the search device clusters the points of interest according to the correlation, the heat, the source, and the like of the point of interest to form a screening theme and a screening condition corresponding thereto.
  • generating at least one screening condition corresponding to the target object according to the attention point further includes:
  • the frequency of occurrence of each point of interest in all related items is counted, wherein the frequency of occurrence of the point of interest is the main basis for evaluating the priorities of different points of interest. For example, depending on the frequency of occurrence of the point of interest, the focus is on the focus. Sort.
  • S8042 Acquire source information of the associated information.
  • the source information of the related information is obtained, and the source list of the related information of the user in the category is established, and basketball shoes are taken as an example.
  • Baidu knows, Baidu Post Bar, e-commerce website and other media accumulate a large number of basketball shoe related information. origin of.
  • S8043 Generate a weight corresponding to the attention point according to the source information of the related information, to sort the attention points according to the appearance frequency and the weight of the attention point.
  • the weight of the corresponding point of interest is generated according to the source information of the associated information, and is sorted according to the frequency of occurrence of the point of interest and the comprehensive evaluation of the weights.
  • the ranking priority of the target object with the corresponding point of interest in the review information is enhanced in the search result.
  • S8044 Generate at least one screening condition corresponding to the target object according to the attention point of the preset number of bits before sorting.
  • the points of interest are clustered according to the correlation, heat, source, and the like of the attention points of the preset number of bits before the sorting, and the screening subject and the corresponding filtering condition are formed, and at least one screening is performed.
  • Conditions are provided to the user to enable the user to search based on at least one filter condition.
  • Figure 2 (b) shows a possible way to provide screening conditions for the user when selecting a notebook computer, wherein the display, function, experience, and appearance are screening topics, and each screening subject corresponds to at least one specific screening condition. For the user to select, for example, the resolution in FIG. 2(b) is high, the game is smooth, the noise is small, and the fineness is respectively corresponding to the specific screening conditions.
  • S8045 Acquire a history record of the user, sort the plurality of filter conditions according to the history record of the user, and provide at least a part of the plurality of filter conditions after the sorting to the user.
  • the screening condition is multiple
  • the history of the user may be obtained, and multiple screening conditions are sorted according to the user's historical record (such as used filtering conditions, browsed target objects, etc.), and after sorting Multiple screening criteria are provided to the user.
  • a part of the screening condition may be selected and provided to the user, and the unprovided part may be directly deleted or hidden in the interactive element in the screening condition providing page, and may be needed by the user (eg, through interaction). Displayed when the operation sends a display command, etc.).
  • the search method of the embodiment of the present invention establishes a focus point of the target object by acquiring related information corresponding to the target object, sorts the focus point according to the frequency of occurrence, source, and the like, and generates a screening condition corresponding to the target object by the sort result, and combines the user with the user.
  • the history records a plurality of filter conditions and provides them to the user for searching, thereby providing users with screening conditions that are more in line with actual needs, making the search more intelligent, the search results more personalized, and improving the user experience.
  • the present invention also proposes a search device.
  • a search device comprising: a search module, configured to receive a search request; and a generating module, configured to generate at least one filter condition associated with the target object for which the search request is targeted according to the search request, and provide at least a portion of the at least one filter condition To the user.
  • Figure 10 is a block diagram showing the structure of a search device in accordance with one embodiment of the present invention.
  • the search device includes a search module 100 and a generation module 200.
  • the search module 100 is configured to receive a search request.
  • the search request sent by the user may be received by the search module 100.
  • the search request should include information such as the target object and its category.
  • the generating module 200 is configured to generate, according to the search request, at least one screening condition associated with the target object for which the search request is targeted, and provide the at least one screening condition to the user.
  • the target object may be a commodity, or may be another category object that can be filtered, such as a service. More specifically, the generating module 200 analyzes and processes information such as a target object and its belonging category in the search request, establishes a focus point of the target object, and clusters the attention points according to information such as relevance, heat, and source of the attention point. Forming a filter subject associated with the target object for which the search request is associated and a filter condition corresponding thereto, and providing at least one filter condition to the user.
  • the searching device of the embodiment of the present invention acquires the target object information according to the user's search request to establish a corresponding point of interest, and generates a filter condition for the user to perform a search selection, thereby allowing the user to design the search condition based on the personalized view from the demand perspective. Make more accurate and fast search, make the search more personalized and enhance the user experience.
  • FIG. 11 is a block diagram showing the structure of a search device according to still another embodiment of the present invention.
  • the search apparatus includes: a search module 100, a generation module 200, an acquisition module 300, an establishment module 400, a word segmentation sub-module 410, a screening sub-module 420, and a generation sub-module 430.
  • the establishing module 400 includes a word segmentation sub-module 410, a filtering sub-module 420, and a generating sub-module 430.
  • the obtaining module 300 is configured to acquire association information corresponding to the target object.
  • the obtaining module 300 may collect related information corresponding to the target object through social media such as Baidu, Baidu Post, or other means capable of acquiring the target object and related information.
  • the related information may be a comment information of a user's description, rating, comment, sharing experience, etc. of a certain target object, or may be statistical information such as number of evaluations, number of times of scoring, number of times of sharing, and other types of related information such as users. Feedback after searching, etc.
  • the establishing module 400 is configured to establish a focus of the target object according to the associated information. More specifically, according to the association information, the establishing module 400 may perform semantic analysis on the related information through a semantic mining method such as natural language processing or any method that can be implemented in the future, and then filter the analysis result to obtain a target object. Feature-related keywords to extract the user's focus on the target object within the category.
  • a semantic mining method such as natural language processing or any method that can be implemented in the future
  • the generating module 200 is further configured to generate at least one screening condition corresponding to the target object according to the attention point. More specifically, the generating module 200 can be related to information such as relevance, heat, source, and the like of the point of interest. The points are clustered to form a screening theme and corresponding screening conditions.
  • the generating module 200 is further configured to perform statistics on the frequency of occurrence of the attention point, and sort the attention points according to the frequency of occurrence of the attention point, and generate at least one screening condition corresponding to the target object according to the attention point of the preset number of bits before the sorting. More specifically, the generating module 200 may count the frequency of occurrence of each point of interest in all associated items, wherein the frequency of occurrence of the point of interest is the primary basis for evaluating the priorities of the different points of interest.
  • the generating module 200 is further configured to obtain source information of the related information, and generate a weight of the corresponding attention point according to the source information of the related information, to sort according to the appearance frequency and the weight of the attention point.
  • the generating module 200 acquires the source information of the related information, and establishes a source list of the related information of the user in the category, and takes basketball shoes as an example. Baidu knows, Baidu Post Bar, e-commerce website and other media accumulate a large number of basketball shoes. The source of the associated information.
  • the generating module 200 generates a weight corresponding to the attention point according to the source information of the related information, and sorts according to the appearance frequency of the attention point and the weight comprehensive evaluation. In addition, for the points of interest that do not appear in the screening list, based on the frequency and weight of occurrence of the points of interest, the ranking priority of the target object with the corresponding point of interest in the review information is enhanced in the search result.
  • the generating module 200 is further configured to acquire a history record of the user, sort the plurality of filtering conditions according to the historical record of the user, and provide at least a part of the plurality of filtering conditions after the sorting to the user.
  • the screening condition is multiple
  • the generating module 200 may acquire the history record of the user, sort the multiple filtering conditions according to the user's historical record (such as the used filtering condition, the browsed target object, etc.), and after sorting Multiple screening criteria are provided to the user.
  • a part of the screening condition may be selected and provided to the user, and the unprovided part may be directly deleted or hidden in the interactive element in the screening condition providing page, and may be needed by the user (eg, through interaction). Displayed when the operation sends a display command, etc.).
  • the word segmentation sub-module 410 is configured to perform word segmentation on each of the association items in the association information, and perform part-of-speech tagging on the plurality of word segments in the association item after the segmentation.
  • the related item is a comment information such as a word, a short sentence, and the like related to the target object attribute, and the related information may include a plurality of related items. More specifically, the word segmentation sub-module 410 classifies each of the association items in the association information, decomposes the language in the association item into independent words, and then labels the part of the word that is decomposed, for example, "noun", " The adjective "is part of the word.
  • the screening sub-module 420 is configured to filter the related items according to the part of speech of the plurality of participles in the related item after the word segmentation.
  • the generating sub-module 430 is configured to establish a target point of the target object according to the frequency of occurrence of the associated item after the screening.
  • the screening sub-module 420 filters the related items according to the specific part-of-speech matching mode according to the part of speech of the plurality of participles in the related item after the word segmentation, for example, according to the matching mode of “noun + adjective”, the participle is filtered, and the word is removed.
  • the related items with the combination of words that meet the requirements are obtained as the related items of the phrase in the form of "noun + adjective", and then the synonyms of the selected items in the selected related items are classified, such as "large screen” "With the large screen size” can be classified into the same type of description and so on.
  • frequency statistics are performed on the classified similar words, and a sub-module 430 is generated. According to the statistical result, the high-frequency focus of the target object is obtained. For example, the frequency of appearance of the similar words can be sorted from high to low, and the top 10 items are preset as the focus of the target object.
  • the searching device of the embodiment of the present invention acquires the attention point of the target object by acquiring the related information corresponding to the target object, sorts the attention point according to the frequency of occurrence, the source, and the like, and generates the screening condition corresponding to the target object by the sorting result, and combines the user
  • the history records a plurality of filter conditions and provides them to the user for searching, thereby providing users with screening conditions that are more in line with actual needs, the search is more intelligent, the search results are more personalized, and the user experience is improved.

Landscapes

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

Abstract

本发明提出一种筛选条件提供方法、装置和搜索方法、装置，其中筛选条件提供方法包括以下步骤：获取目标对象对应的关联信息；根据关联信息建立目标对象的关注点；以及根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。本发明实施例的筛选条件提供方法，通过获取目标对象对应的关联信息建立目标对象的关注点，根据关注点生成目标对象对应的筛选条件提供给用户以进行搜索，从而允许用户从需求视角，基于个性化设计筛选条件，增加了筛选条件与用户真实需求的相关性，使筛选更加精确快捷，筛选结果更加个性化，提升了用户体验。

Description

筛选条件提供方法、装置和搜索方法、装置
相关申请的交叉引用
本申请要求北京百度网讯科技有限公司于2013年10月15日提交的、发明名称为“筛选条件提供方法、装置和搜索方法、装置”的、中国专利申请号“201310482406.3”的优先权。
技术领域
本发明涉及互联网技术领域，尤其涉及一种筛选条件提供方法、装置和搜索方法、装置。
背景技术
随着互联网技术的发展，人们越来越多地选择通过网络找到自己需要的商品或服务。现有的网络服务，例如搜索或网络销售主要是基于固定的产品型号、价格范围等基本产品描述设置筛选条件，以根据用户的选择对商品或服务进行搜索和筛选。
在实现本发明过程中，发明人发现现有技术至少存在以下问题：由于用户对商品或服务的需求是个性化的，而现在常见的筛选条件大多为商品或服务的固定参数，无法提供与用户真实需求具有充分相关性的筛选条件，因而很难得到符合用户真实需求的筛选结果。另外，对于没有直接使用经验的品类，用户甚至不知道应该关注商品或服务的哪些特征去搜索和筛选，导致基于固定参数设定的筛选条件对用户获取选择对象的过程造成了障碍。因此，现有的固定模式筛选条件无法匹配用户个性化的需求，筛选相对低效，用户体验差。
发明内容
本发明旨在至少解决上述技术问题之一。
为此，本发明的第一个目的在于提出一种筛选条件提供方法。该方法能够提供针对各类商品或服务的个性化筛选条件，从而可以允许用户从需求视角，进行更加精确快捷的筛选，使筛选结果更加个性化，提升了用户体验。
本发明的第二个目的在于提出一种筛选条件提供装置。
本发明的第三个目的在于提出一种搜索方法。
本发明的第四个目的在于提出一种搜索装置。
为了实现上述目的，本发明第一方面实施例的筛选条件提供方法包括以下步骤：获取 目标对象对应的关联信息；根据关联信息建立目标对象的关注点；根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
根据本发明实施例的筛选条件提供方法，通过获取目标对象对应的关联信息建立目标对象的关注点，并生成筛选条件提供给用户以进行选择，从而允许用户从需求视角，基于个性化设计筛选条件，增加了筛选条件与用户真实需求的相关性，使筛选更加精确快捷，筛选结果更加个性化，提升了用户体验。
为了实现上述目的，本发明第二方面实施例的筛选条件提供装置包括：关联信息获取模块，用于获取目标对象对应的关联信息；关注点建立模块，用于根据关联信息建立目标对象的关注点；筛选条件生成模块，用于根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
根据本发明实施例的筛选条件提供装置，通过获取目标对象对应的关联信息建立目标对象的关注点，并生成筛选条件提供给用户以进行选择，从而使筛选条件更贴近用户的真实需求，使筛选更加精确快捷，筛选结果更加个性化，提升了用户体验。
为了实现上述目的，本发明第三方面实施例的搜索方法包括以下步骤：接收搜索请求；根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件；以及将至少一个筛选条件的至少一部分提供至用户。
本发明实施例的搜索方法，根据用户的搜索请求获取目标对象信息以建立相应的关注点，并生成筛选条件提供给用户以进行搜索选择，从而允许用户从需求视角，基于个性化设计搜索条件，进行更加精确快捷的搜索，使搜索更加个性化，提升了用户体验。
为了实现上述目的，本发明第四方面实施例的搜索装置包括：搜索模块，用于接收搜索请求；生成模块，用于根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
本发明实施例的搜索装置，根据用户的搜索请求获取目标对象信息以建立相应的关注点，并生成筛选条件提供给用户以进行搜索选择，从而允许用户从需求视角，基于个性化设计搜索条件，进行更加精确快捷的搜索，使搜索更加个性化，提升了用户体验。
本发明附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本发明的实践了解到。
附图说明
本发明上述的和/或附加的方面和优点从下面结合附图对实施例的描述中将变得明显和容易理解，其中，
图1是根据本发明一个实施例的筛选条件提供方法的流程图；
图2(a)是根据本发明一个实施例的筛选主题和与其对应的筛选条件的示意图；
图2(b)是根据本发明又一个实施例的筛选主题和与其对应的筛选条件的示意图；
图2(c)是根据本发明另一个实施例的筛选主题和与其对应的筛选条件的示意图；
图2(d)是根据本发明再一个实施例的筛选主题和与其对应的筛选条件的示意图；
图3是根据本发明又一个实施例的筛选条件提供方法的流程图；
图4是根据本发明另一个实施例的筛选条件提供方法的流程图；
图5是根据本发明一个实施例的筛选条件提供装置的结构框图；
图6是根据本发明又一个实施例的筛选条件提供装置的结构框图；
图7是根据本发明一个实施例的搜索方法的流程图；
图8是根据本发明又一个实施例的搜索方法的流程图；
图9(a)是根据本发明一个实施例的根据关联信息建立目标对象的关注点的流程图；
图9(b)是根据本发明一个实施例的根据关注点生成目标对象对应的至少一个筛选条件的流程图；
图10是根据本发明一个实施例的搜索装置的结构框图；
图11是根据本发明又一个实施例的搜索装置的结构框图。
具体实施方式
下面详细描述本发明的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，仅用于解释本发明，而不能理解为对本发明的限制。相反，本发明的实施例包括落入所附加权利要求书的精神和内涵范围内的所有变化、修改和等同物。
在本发明的描述中，需要理解的是，术语“第一”、“第二”等仅用于描述目的，而不能理解为指示或暗示相对重要性。在本发明的描述中，需要说明的是，除非另有明确的规定和限定，术语“相连”、“连接”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连。对于本领域的普通技术人员而言，可以具体情况理解上述术语在本发明中的具体含义。此外，在本发明的描述中，除非另有说明，“多个”的含义是两个或两个以上。
流程图中或在此以其他方式描述的任何过程或方法描述可以被理解为，表示包括一个或更多个用于实现特定逻辑功能或过程的步骤的可执行指令的代码的模块、片段或部分，并且本发明的优选实施方式的范围包括另外的实现，其中可以不按所示出或讨论的顺序，包括根据所涉及的功能按基本同时的方式或按相反的顺序，来执行功能，这应被本发明的实施例所属技术领域的技术人员所理解。
下面参考附图描述根据本发明实施例的筛选条件提供方法、装置和搜索方法、装置。
目前的筛选条件提供方法主要是从产品分类角度的筛选模式出发，基于固定的产品型号、价格范围等基本产品描述设置筛选条件，无法匹配用户个性化的需求，相对低效，用户体验差。为此，本发明提出了一种筛选条件提供方法，包括以下步骤：获取目标对象对应的关联信息；根据关联信息建立目标对象的关注点；以及根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
图1是根据本发明一个实施例的筛选条件提供方法的流程图。
如图1所示，筛选条件提供方法包括：
S101，获取目标对象对应的关联信息。
其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。目标对象及其所属品类可通过现有的或未来可能出现的任何能够获取用户指令信息的方法或装置获得。
具体地，在本发明的一个实施例中，可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的途径，收集目标对象对应的关联信息。其中，关联信息可以是用户对某一目标对象的说明、评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
S102，根据关联信息建立目标对象的关注点。
具体地，根据关联信息，可通过自然语言处理等语义挖掘方法或未来可能出现的任何能够实现语义挖掘的方法对关联信息进行语义分析，然后对分析结果进行筛选，得到与目标对象特性相关的关键词，以提取用户在品类内目标对象的关注点。
S103，根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
具体地，根据关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户，以使用户根据至少一个筛选条件进行搜索。举例而言，图2(a)、图2(b)、图2(c)分别为根据本发明方法提供筛选条件的三种可能的体现形式。如图2(a)所示，舒适性、耐穿性、清洁性为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(a)中轻便、不断底、好刷洗分别为对应的具体筛选条件。
在本发明的一个实施例中，如图2(b)所示，当一个筛选主题下有多个筛选条件时，可根据筛选条件的关注频率等因素选择一部分筛选条件提供给用户，未提供的部分筛选条件可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示 指令)时显示。如图2(c)所示，本发明的筛选条件提供方法还可用于搜索引擎，以及任何有搜索、筛选需求的方案、方法和装置，例如在用户进行搜索时提供搜索条件的提示。
在本发明的再一个实施例中，如图2(d)所示，可以在向用户展示商品或服务的同时，提供个性化的筛选条件，例如发音质量好、风挡投射、省电、资源占用小等。
本发明实施例的筛选条件提供方法，通过获取目标对象对应的关联信息建立目标对象的关注点，并生成筛选条件提供给用户以进行选择，从而允许用户从需求视角，基于个性化设计筛选条件，增加了筛选条件与用户真实需求的相关性，使筛选更加精确快捷，筛选结果更加个性化，提升了用户体验。
图3是根据本发明又一个实施例的筛选条件提供方法的流程图。
为了进一步提升用户体验，对关联项进行筛选以建立目标对象的关注点，根据关注点的出现频次对关注点进行排序，从而生成筛选条件，并根据用户选择的筛选条件查询搜索数据库以获取查询结果，具体地，如图3所示，筛选条件提供方法包括以下步骤：
S301，根据目标对象建立品牌列表，其中，品牌列表中包括多个品牌。
其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。
具体地，根据目标对象建立对应各品类内的品牌列表和对应的具体目标对象型号列表，例如篮球鞋包括耐克、阿迪达斯等品牌，以及对应的具体篮球鞋型号列表。
S302，获取目标对象对应的关联信息。
具体地，在本发明的一个实施例中，可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的途径，收集目标对象对应的关联信息。其中，关联信息包括评论信息等，可以是用户对某一目标对象的说明、评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
S303，对关联信息中的每个关联项进行分词，并对分词之后的关联项中的多个分词分别进行词性标注。
其中，关联项为与目标对象属性相关的词语、短句等评论信息，关联信息可包括多个关联项。具体地，对关联信息中的每个关联项进行分词，将关联项中的语言分解为独立的词，然后为分解出的独立的词标注词性，例如标注“名词”、“形容词”等词性。
S304，根据分词之后的关联项中多个分词的词性对关联项进行筛选，并根据筛选之后的关联项的出现频率建立目标对象的关注点。
具体地，根据分词之后的关联项中多个分词的词性，按照特定的词性匹配模式对关联项进行筛选，如按照“名词+形容词”的匹配模式对分词进行筛选，去掉不含符合要求的词语组合的关联项，得到的筛选结果即为含有“名词+形容词”形式的词组的关联项，进而对 筛选出的关联项中的词组进行同义近义词归类，如“屏幕大”与“屏幕尺寸大”可以归为同一类描述等。然后对归类后的同类词语进行频率统计，根据统计结果得到目标对象的高频关注点，例如，可按照同类词语的出现频次从高到低排序，预设前10项为目标对象的关注点。
S305，针对多个品牌分别建立搜索数据库。
其中，搜索数据库包括多个品牌及多个品牌对应的关注点。具体地，多个品牌对应的关注点包括多个品牌对应的各个型号目标对象的关注点，以便于通过搜索关注点筛选出对应型号的目标对象。
S306，对关注点的出现频次进行统计。
具体地，统计每个关注点在所有的关联项中的出现频次，其中，关注点的出现频次是评估不同关注点的优先级的主要依据。
S307，根据关注点的出现频次对关注点进行排序。
具体地，根据关注点的出现频次，可按照出现频次从高到低对关注点进行排序。
S308，根据排序前预设位数的关注点生成目标对象对应的至少一个筛选条件。
在本发明的一个实施例中，根据排序前预设位数的关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户，以使用户根据至少一个筛选条件进行搜索。如图2(b)所示为选择笔记本电脑时一种可能的为用户提供筛选条件的方式，其中，显示、功能、体验、外观为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(b)中分辨率高、游戏流畅、噪音小、精细等分别为对应的具体筛选条件。
S309，接收用户选择的筛选条件。
具体地，用户根据筛选主题和与其对应的筛选条件，选取自己需要的一项或几项提交，然后接收端接收用户选择的筛选条件。
S310，根据用户选择的筛选条件查询搜索数据库以获取查询结果，并提供至用户。
具体地，按照用户选择的筛选条件，在搜索数据库中对关注点进行检索，将检索到的符合条件的关注点按照预设的方式和顺序排列，然后将检索到的关注点所对应的目标对象返回给用户。
应当理解，本实施例中的筛选和统计方法需要以较大容量的统计样本为分析对象。
本发明实施例的筛选条件提供方法，通过对关联信息中的关联项分词并分析后进行筛选以建立目标对象的用户关注点，根据关注点的出现频次对关注点进行排序，从而生成筛选条件，并根据用户选择的筛选条件查询搜索数据库以获取对应的目标对象，从而使搜索到的目标对象更符合用户的真实需求，搜索过程更加智能化，提升了用户体验。
图4是根据本发明另一个实施例的筛选条件提供方法的流程图。
为了进一步提升用户体验，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重提升有相应点评的目标对象的排序优先级，并结合用户的历史记录对多个筛选条件进行排序，具体地，如图4所示，筛选条件提供方法包括以下步骤。
S401，根据目标对象建立品牌列表，其中，品牌列表中包括多个品牌。
其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。
具体地，建立对应各品类内的品牌列表和对应的具体目标对象型号列表，例如篮球鞋包括耐克，阿迪达斯等品牌，以及对应的具体篮球鞋型号列表。
S402，获取目标对象对应的关联信息。
具体地，在本发明的一个实施例中，可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的垂直行业用户评论信息等途径，收集商品目标对象对应的关联信息。其中，关联信息包括用户点评信息和/或搜索信息评论信息等，可以是用户对某一商品目标对象的说明、评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
S403，对关联信息中的每个关联项进行分词，并对分词之后的关联项中的多个分词分别进行词性标注。
其中，关联项为与目标对象属性相关的词语、短句等评论信息，关联信息可包括多个关联项。具体地，对关联信息中的每个关联项进行分词，将关联项中的语言分解为独立的词，然后为分解出的独立的词标注词性，例如标注“名词”、“形容词”等词性。
S404，根据分词之后的关联项中多个分词的词性对关联项进行筛选，并根据筛选之后的关联项的出现频率建立目标对象的关注点。
具体地，根据分词之后的关联项中多个分词的词性，按照特定的词性匹配模式对关联项进行筛选，如按照“名词+形容词”的匹配模式对分词进行筛选，去掉不含符合要求的词语组合的关联项，得到的筛选结果即为含有“名词+形容词”形式的词组的关联项，进而对筛选出的关联项中的词组进行同义近义词归类，如“屏幕大”与“屏幕尺寸大”可以归为同一类描述等。然后对归类后的同类词语进行频率统计，根据统计结果得到目标对象的高频关注点，例如，可按照同类词语的出现频次从高到低排序，预设前10项为目标对象的关注点。
S405，针对多个品牌分别建立搜索数据库。
其中，搜索数据库包括多个品牌及多个品牌对应的关注点。具体地，多个品牌对应的关注点包括多个品牌对应的各个型号目标对象的关注点，以便于通过搜索关注点筛选出对应型号的目标对象。
S406，对关注点的出现频次进行统计。
具体地，统计每个关注点在所有的关联项中的出现频次，其中，关注点的出现频次是评估不同关注点的优先级的主要依据。
S407，获取关联信息的来源信息。
具体地，获取关联信息的来源信息，并建立用户在垂直行业的关联信息的来源列表，以买篮球鞋为例，百度知道、百度贴吧、电商网站等媒体均为积累了大量篮球鞋关联信息的来源。
S408，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重进行排序。
具体地，由于将关注点排序需要考虑关联信息来源的权威性等因素，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重综合评估进行排序。另外，对于未出现在筛选列表中的关注点，也可以基于这些关注点的出现频次和权重，在搜索结果中提升点评信息中含有相应关注点的目标对象的排序优先级。
S409，根据排序前预设位数的关注点生成目标对象对应的至少一个筛选条件。
在本发明的一个实施例中，根据排序前预设位数的关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户，以使用户根据至少一个筛选条件进行搜索。如图2(b)所示为选择笔记本电脑时一种可能的为用户提供筛选条件的方式，其中，显示、功能、体验、外观为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(b)中分辨率高、游戏流畅、噪音小、精细等分别为对应的具体筛选条件。
S410，获取用户的历史记录，根据用户的历史记录对多个筛选条件进行排序，并将排序之后的多个筛选条件的至少一部分提供至用户。
具体地，当筛选条件为多个时，可获取用户的历史记录，根据用户的历史记录(如使用过的筛选条件、浏览过的目标对象等)对多个筛选条件进行排序，并将排序之后的多个筛选条件提供至用户。在本发明的一个实施例中，排序后可选择一部分筛选条件提供给用户，未提供的部分可以直接删除，也可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示指令等)时显示。
S411，接收用户选择的筛选条件。
具体地，用户根据筛选主题和与其对应的筛选条件，选取自己需要的一项或几项提交，然后接收端接收用户选择的筛选条件。
S412，根据用户选择的筛选条件查询搜索数据库以获取查询结果，并提供至用户。
具体地，按照用户选择的筛选条件，在搜索数据库中对关注点进行检索，将检索到的 符合条件的关注点按照预设的方式和顺序排列，然后将检索到的关注点所对应的目标对象返回给用户。
应当理解，本实施例中的筛选和统计方法需要以较大容量的统计样本为分析对象。
本发明实施例的筛选条件提供方法，通过获取目标对象对应的关联信息建立目标对象的关注点和搜索数据库，将关注点根据出现频次、来源等信息进行排序，由排序结果生成目标对象对应的筛选条件，并结合用户的历史记录对多个筛选条件进行排序，提供给用户以进行搜索，从而为用户提供更符合实际需求的筛选条件，筛选结果更加个性化，提升了用户体验。
为了实现上述实施例，本发明还提出一种筛选条件提供装置。
一种筛选条件提供装置，包括：关联信息获取模块，用于获取目标对象对应的关联信息；关注点建立模块，用于根据关联信息建立目标对象的关注点；筛选条件生成模块，用于根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
图5是根据本发明一个实施例的筛选条件提供装置的结构框图。
如图5所示，筛选条件提供装置包括：关联信息获取模块100、关注点建立模块200和筛选条件生成模块300。
具体地，关联信息获取模块100用于获取目标对象对应的关联信息。其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。目标对象及其所属品类可通过现有的或未来可能出现的任何能够获取用户指令信息的方法或装置获得。在本发明的一个实施例中，关联信息获取模块100可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的途径，收集目标对象对应的关联信息。其中，关联信息包括评论信息等，可以是用户对某一目标对象的评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
关注点建立模块200用于根据关联信息建立目标对象的关注点。更具体地，关注点建立模块200根据关联信息，可通过自然语言处理等语义挖掘方法或未来可能出现的语义挖掘方法对关联信息进行语义分析，然后对分析结果进行筛选，得到与目标对象属性相关的关键词，以提取用户在品类内目标对象的关注点。
筛选条件生成模块300用于根据关注点生成目标对象对应的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。在本发明的一个实施例中，筛选条件生成模块300还用于对关注点的出现频次进行统计，并根据关注点的出现频次对关注点进行排序，以及根据排序前预设位数的关注点生成目标对象对应的至少一个筛选条件。筛选条件生成 模块300根据关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户，以使用户根据至少一个筛选条件进行搜索。举例而言，图2(a)、图2(b)、图2(c)分别为根据本发明方法提供筛选条件的三种可能的体现形式。如图2(a)所示，舒适性、耐穿性、清洁性为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(a)中轻便、不断底、好刷洗分别为对应的具体筛选条件。
本发明实施例的筛选条件提供装置，通过获取目标对象对应的关联信息建立目标对象的关注点，并生成筛选条件提供给用户以进行选择，从而使搜索条件更贴近用户的真实需求，使筛选更加精确快捷，筛选结果更加个性化，提升了用户体验。
图6是根据本发明又一个实施例的筛选条件提供装置的结构框图。
如图6所示，筛选条件提供装置包括：关联信息获取模块100、关注点建立模块200、筛选条件生成模块300、搜索数据库建立模块400、搜索模块500、分词子模块210、筛选子模块220和生成子模块230。其中，关注点建立模块200包括分词子模块210、筛选子模块220和生成子模块230。
具体地，搜索数据库建立模块400用于建立品牌列表，其中，品牌列表中包括多个品牌，并针对多个品牌分别建立搜索数据库，搜索数据库包括多个品牌及多个品牌对应的关注点。例如，搜索数据库建立模块400建立对应各品类内的品牌列表和对应的具体目标对象型号列表，例如篮球鞋包括耐克、阿迪达斯等品牌，以及对应的具体篮球鞋型号列表。
搜索模块500用于接收用户选择的筛选条件，并根据用户选择的筛选条件查询搜索数据库以获取查询结果，以及将查询结果提供至用户。更具体地，用户根据筛选主题和与其对应的筛选条件，选取自己需要的一项或几项提交，然后搜索模块500接收用户选择的筛选条件，并按照用户选择的筛选条件，在搜索数据库中对关注点进行检索，将检索到的符合条件的关注点按照预设的方式和顺序排列，然后将检索到的关注点所对应的目标对象返回给用户。
在本发明的一个实施例中，筛选条件生成模块300还用于获取关联信息的来源信息，并根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重对关注点进行排序。其中，筛选条件生成模块300对关注点的出现频次进行统计，获取关联信息的来源信息，并建立用户在品类内的关联信息的来源列表，以买篮球鞋为例，百度知道、百度贴吧、电商网站等媒体均为积累了大量篮球鞋关联信息的来源。由于将关注点排序需要考虑关联信息来源的权威性等因素，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重综合评估进行排序。另外，对于未出现在筛选列表中的关注点，也可以基于这些关注点的出现频次和权重，在搜索结果中提升点评信息中含有相应 关注点的目标对象的排序优先级。
更具体地，筛选条件生成模块300还用于获取用户的历史记录，根据用户的历史记录对多个筛选条件进行排序，并将排序之后的多个筛选条件的至少一部分提供至用户。当筛选条件为多个时，筛选条件生成模块300获取用户的历史记录，根据用户的历史记录(如使用过的筛选条件、浏览过的目标对象等)对多个筛选条件进行排序，并将排序之后的多个筛选条件提供至用户。在本发明的一个实施例中，排序后可选择一部分筛选条件提供给用户，未提供的部分可以直接删除，也可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示指令等)时显示。
分词子模块210用于对关联信息中的每个关联项进行分词，并对分词之后的关联项中的多个分词分别进行词性标注。其中，关联项为与目标对象属性相关的词语、短句等评论信息，关联信息可包括多个关联项。更具体地，分词子模块210对关联信息中的每个关联项进行分词，将关联项中的语言分解为独立的词，然后为分解出的独立的词标注词性，例如标注“名词”、“形容词”等词性。
筛选子模块220用于根据分词之后的关联项中多个分词的词性对关联项进行筛选。更具体地，筛选子模块220根据分词之后的关联项中多个分词的词性，按照特定的词性匹配模式对关联项进行筛选，如按照“名词+形容词”的模式对分词进行筛选，去掉不含所要求的词语组合的关联项，得到的筛选结果即为含有“名词+形容词”形式的词组的关联项。
生成子模块230用于根据筛选之后的关联项的出现频率建立目标对象的关注点。具体地，生成子模块230对筛选出的关联项中的词组进行同义近义词归类，如“屏幕大”与“屏幕尺寸大”可以归为同一类描述等。然后对归类后的同类词语进行频率统计，根据统计结果得到目标对象的高频关注点，例如，可按照同类词语的出现频次从高到低排序，预设前10项为目标对象的关注点。
本发明实施例的筛选条件提供装置，通过获取目标对象对应的关联信息建立目标对象的关注点和搜索数据库，将关注点按照出现频次、来源等信息进行排序以生成筛选条件，并结合用户的历史记录对多个筛选条件进行排序，提供给用户以进行搜索，然后根据用户选择的筛选条件查询搜索数据库以获取对应的目标对象，从而为用户提供更贴近用户真实需求的筛选条件，筛选结果更加个性化，提升了用户体验。
为了实现上述实施例，本发明还提出一种搜索方法。
一种搜索方法，包括以下步骤：接收搜索请求；根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件；以及将至少一个筛选条件的至少一部分提供至用户。
图7是根据本发明一个实施例的搜索方法的流程图。
如图7所示，搜索方法包括：
S701，接收搜索请求。
在本发明的实施例中，可通过搜索引擎或装置接收用户发送的搜索请求。其中，搜索请求中应包含目标对象及其所属品类等信息。
S702，根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件。
其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。
具体地，对搜索请求中的目标对象及其所属品类等信息进行分析处理，建立目标对象的关注点，并根据关注点的相关性、热度、来源等信息对关注点进行聚类，形成与搜索请求所针对目标对象相关联的筛选主题和与其对应的筛选条件。
S703，将至少一个筛选条件的至少一部分提供至用户。
举例而言，图2(a)、图2(b)、图2(c)分别为根据本发明方法提供搜索条件的三种可能的体现形式。如图2(a)所示，舒适性、耐穿性、清洁性为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(a)中轻便、不断底、好刷洗分别为对应的具体筛选条件。当一个筛选主题下有多个筛选条件时，可根据筛选条件的关注频率等因素选择一部分筛选条件提供给用户。如图2(c)所示，本发明的搜索方法可用于搜索引擎，以及任何有搜索、筛选需求的方案、方法和装置，例如在用户进行搜索时提供搜索条件的提示。
在本发明的一个实施例中，如图2(b)所示，当一个筛选主题下有多个筛选条件时，可根据筛选条件的关注频率等因素选择一部分筛选条件提供给用户，未提供的部分筛选条件可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示指令)时显示。
在本发明的再一个实施例中，如图2(d)所示，可以在向用户展示商品或服务的同时，提供个性化的筛选条件，例如发音质量好、风挡投射、省电、资源占用小等。
本发明实施例的搜索方法，根据用户的搜索请求获取目标对象信息以建立相应的关注点，并生成筛选条件提供给用户以进行搜索选择，从而允许用户从需求视角，基于个性化设计搜索条件，进行更加精确快捷的搜索，使搜索更加个性化，提升了用户体验。
图8是根据本发明又一个实施例的搜索方法的流程图。
为了进一步提升用户体验，对目标对象的关联信息进行分析处理以建立目标对象的关注点，根据关注点的出现频次对关注点进行排序，从而生成筛选条件，并根据用户选择的筛选条件查询搜索数据库以获取查询结果，具体地，如图8所示，搜索方法包括以下步骤：
S801，接收搜索请求。
在本发明的实施例中，可通过搜索引擎或装置接收用户发送的搜索请求。其中，搜索请求中应包含目标对象及其所属品类等信息。
S802，获取目标对象对应的关联信息。
具体地，在本发明的一个实施例中，可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的途径，收集目标对象对应的关联信息。其中，关联信息可以是用户对某一目标对象的说明、评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
S803，根据关联信息建立目标对象的关注点。
具体地，根据关联信息，可通过自然语言处理等语义挖掘方法或未来可能出现的任何能够实现语义挖掘的方法对关联信息进行语义分析，然后对分析结果进行筛选，得到与目标对象特性相关的关键词，以提取用户在品类内目标对象的关注点。如图9(a)所示，根据关联信息建立目标对象的关注点进一步包括：
S8031，对关联信息中的每个关联项进行分词，并对分词之后的关联项中的多个分词分别进行词性标注。
其中，关联项为与目标对象属性相关的词语、短句等评论信息，关联信息可包括多个关联项。具体地，对关联信息中的每个关联项进行分词，将关联项中的语言分解为独立的词，然后为分解出的独立的词标注词性，例如标注“名词”、“形容词”等词性。
S8032，根据分词之后的关联项中多个分词的词性对关联项进行筛选，并根据筛选之后的关联项的出现频率建立目标对象的关注点。
具体地，根据分词之后的关联项中多个分词的词性，按照特定的词性匹配模式对关联项进行筛选，如按照“名词+形容词”的匹配模式对分词进行筛选，去掉不含符合要求的词语组合的关联项，得到的筛选结果即为含有“名词+形容词”形式的词组的关联项，进而对筛选出的关联项中的词组进行同义近义词归类，如“屏幕大”与“屏幕尺寸大”可以归为同一类描述等。然后对归类后的同类词语进行频率统计，根据统计结果得到目标对象的高频关注点，例如，可按照同类词语的出现频次从高到低排序，预设前10项为目标对象的关注点。
S804，根据关注点生成目标对象对应的至少一个筛选条件，并提供至于用户。
具体地，搜索装置根据关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件。如图9(b)所示，根据关注点生成目标对象对应的至少一个筛选条件进一步包括：
S8041，对关注点的出现频次进行统计。
具体地，统计每个关注点在所有的关联项中的出现频次，其中，关注点的出现频次是评估不同关注点的优先级的主要依据。例如，可根据关注点的出现频次从高到低对关注点 进行排序。
S8042，获取关联信息的来源信息。
具体地，获取关联信息的来源信息，并建立用户在品类内的关联信息的来源列表，以买篮球鞋为例，百度知道、百度贴吧、电商网站等媒体均为积累了大量篮球鞋关联信息的来源。
S8043，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重对关注点进行排序。
具体地，由于将关注点排序需要考虑关联信息来源的权威性等因素，根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重综合评估进行排序。另外，对于未出现在筛选列表中的关注点，也可以基于这些关注点的出现频次和权重，在搜索结果中提升点评信息中含有相应关注点的目标对象的排序优先级。
S8044，根据排序前预设位数的关注点生成目标对象对应的至少一个筛选条件。
在本发明的一个实施例中，根据排序前预设位数的关注点的相关性、热度、来源等信息对关注点进行聚类，形成筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户，以使用户根据至少一个筛选条件进行搜索。如图2(b)所示为选择笔记本电脑时一种可能的为用户提供筛选条件的方式，其中，显示、功能、体验、外观为筛选主题，每个筛选主题下都至少对应一个具体筛选条件供用户选择，例如图2(b)中分辨率高、游戏流畅、噪音小、精细等分别为对应的具体筛选条件。
S8045，获取用户的历史记录，根据用户的历史记录对多个筛选条件进行排序，并将排序之后的多个筛选条件的至少一部分提供至用户。
具体地，当筛选条件为多个时，可获取用户的历史记录，根据用户的历史记录(如使用过的筛选条件、浏览过的目标对象等)对多个筛选条件进行排序，并将排序之后的多个筛选条件提供至用户。
在本发明的一个实施例中，排序后可选择一部分筛选条件提供给用户，未提供的部分可以直接删除，也可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示指令等)时显示。
应当理解，本实施例中的筛选和统计方法需要以较大容量的统计样本为分析对象。
本发明实施例的搜索方法，通过获取目标对象对应的关联信息建立目标对象的关注点，将关注点根据出现频次、来源等信息进行排序，由排序结果生成目标对象对应的筛选条件，并结合用户的历史记录对多个筛选条件进行排序，提供给用户以进行搜索，从而为用户提供更符合实际需求的筛选条件，使搜索更加智能化，搜索结果更加个性化，提升了用户体验。
为了实现上述实施例，本发明还提出一种搜索装置。
一种搜索装置，包括：搜索模块，用于接收搜索请求；生成模块，用于根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件，并将至少一个筛选条件的至少一部分提供至用户。
图10是根据本发明一个实施例的搜索装置的结构框图。
如图10所示，搜索装置包括：搜索模块100和生成模块200。
搜索模块100用于接收搜索请求。在本发明的实施例中，可通过搜索模块100接收用户发送的搜索请求。其中，搜索请求中应包含目标对象及其所属品类等信息。
生成模块200用于根据搜索请求生成与搜索请求所针对目标对象相关联的至少一个筛选条件，并将至少一个筛选条件提供至用户。其中，目标对象可以是商品，也可以是服务等其他可以筛选的品类对象。更具体地，生成模块200对搜索请求中的目标对象及其所属品类等信息进行分析处理，建立目标对象的关注点，并根据关注点的相关性、热度、来源等信息对关注点进行聚类，形成与搜索请求所针对目标对象相关联的筛选主题和与其对应的筛选条件，并将至少一个筛选条件提供至用户。
本发明实施例的搜索装置，根据用户的搜索请求获取目标对象信息以建立相应的关注点，并生成筛选条件提供给用户以进行搜索选择，从而允许用户从需求视角，基于个性化设计搜索条件，进行更加精确快捷的搜索，使搜索更加个性化，提升了用户体验。
图11是根据本发明又一个实施例的搜索装置的结构框图。
如图11所示，搜索装置包括：搜索模块100、生成模块200、获取模块300、建立模块400、分词子模块410、筛选子模块420和生成子模块430。其中，建立模块400包括分词子模块410、筛选子模块420和生成子模块430。
获取模块300用于获取目标对象对应的关联信息。在本发明的一个实施例中，获取模块300可通过百度知道、百度贴吧等社交媒体或其他能够获取目标对象及其相关信息的途径，收集目标对象对应的关联信息。其中，关联信息可以是用户对某一目标对象的说明、评分、评论、分享心得等形式的评论信息，也可以是评价人数、打分次数、分享次数等统计信息，以及其他类型的相关信息如用户搜索之后的反馈信息等。
建立模块400用于根据关联信息建立目标对象的关注点。更具体地，根据关联信息，建立模块400可通过自然语言处理等语义挖掘方法或未来可能出现的任何能够实现语义挖掘的方法对关联信息进行语义分析，然后对分析结果进行筛选，得到与目标对象特性相关的关键词，以提取用户在品类内目标对象的关注点。
在本发明的一个实施例中，生成模块200还用于根据关注点生成目标对象对应的至少一个筛选条件。更具体地，生成模块200可根据关注点的相关性、热度、来源等信息对关 注点进行聚类，形成筛选主题和与其对应的筛选条件。
生成模块200还用于对关注点的出现频次进行统计，并根据关注点的出现频次对关注点进行排序，以及根据排序前预设位数的关注点生成目标对象对应的至少一个筛选条件。更具体地，生成模块200可统计每个关注点在所有的关联项中的出现频次，其中，关注点的出现频次是评估不同关注点的优先级的主要依据。
生成模块200还用于获取关联信息的来源信息，并根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重进行排序。其中，生成模块200获取关联信息的来源信息，并建立用户在品类内的关联信息的来源列表，以买篮球鞋为例，百度知道、百度贴吧、电商网站等媒体均为积累了大量篮球鞋关联信息的来源。由于将关注点排序需要考虑关联信息来源的权威性等因素，生成模块200根据关联信息的来源信息生成对应关注点的权重，以根据关注点的出现频次和权重综合评估进行排序。另外，对于未出现在筛选列表中的关注点，也可以基于这些关注点的出现频次和权重，在搜索结果中提升点评信息中含有相应关注点的目标对象的排序优先级。
生成模块200还用于获取用户的历史记录，根据用户的历史记录对多个筛选条件进行排序，并将排序之后的多个筛选条件的至少一部分提供至用户。当筛选条件为多个时，生成模块200可获取用户的历史记录，根据用户的历史记录(如使用过的筛选条件、浏览过的目标对象等)对多个筛选条件进行排序，并将排序之后的多个筛选条件提供至用户。在本发明的一个实施例中，排序后可选择一部分筛选条件提供给用户，未提供的部分可以直接删除，也可隐藏在筛选条件提供页面内的交互元素中，并在用户需要(如通过交互操作发送显示指令等)时显示。
分词子模块410用于对关联信息中的每个关联项进行分词，并对分词之后的关联项中的多个分词分别进行词性标注。其中，关联项为与目标对象属性相关的词语、短句等评论信息，关联信息可包括多个关联项。更具体地，分词子模块410对关联信息中的每个关联项进行分词，将关联项中的语言分解为独立的词，然后为分解出的独立的词标注词性，例如标注“名词”、“形容词”等词性。
筛选子模块420用于根据分词之后的关联项中多个分词的词性对关联项进行筛选。
生成子模块430用于根据筛选之后的关联项的出现频率建立目标对象的关注点。
更具体地，筛选子模块420根据分词之后的关联项中多个分词的词性，按照特定的词性匹配模式对关联项进行筛选，如按照“名词+形容词”的匹配模式对分词进行筛选，去掉不含符合要求的词语组合的关联项，得到的筛选结果即为含有“名词+形容词”形式的词组的关联项，进而对筛选出的关联项中的词组进行同义近义词归类，如“屏幕大”与“屏幕尺寸大”可以归为同一类描述等。然后对归类后的同类词语进行频率统计，生成子模块430 根据统计结果得到目标对象的高频关注点，例如，可按照同类词语的出现频次从高到低排序，预设前10项为目标对象的关注点。
应当理解，本实施例中的筛选和统计方法需要以较大容量的统计样本为分析对象。
本发明实施例的搜索装置，通过获取目标对象对应的关联信息建立目标对象的关注点，将关注点根据出现频次、来源等信息进行排序，由排序结果生成目标对象对应的筛选条件，并结合用户的历史记录对多个筛选条件进行排序，提供给用户以进行搜索，从而为用户提供更符合实际需求的筛选条件，搜索更加智能化，搜索结果更加个性化，提升了用户体验。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本发明的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任何的一个或多个实施例或示例中以合适的方式结合。
尽管已经示出和描述了本发明的实施例，本领域的普通技术人员可以理解：在不脱离本发明的原理和宗旨的情况下可以对这些实施例进行多种变化、修改、替换和变型，本发明的范围由权利要求及其等同物限定。

Claims (28)

  1. 一种筛选条件提供方法，其特征在于，包括以下步骤：
    获取目标对象对应的关联信息；
    根据所述关联信息建立所述目标对象的关注点；以及
    根据所述关注点生成所述目标对象对应的至少一个筛选条件，并将所述至少一个筛选条件的至少一部分提供至用户。
  2. 如权利要求1所述的方法，其特征在于，所述关联信息包括评论信息。
  3. 如权利要求1所述的方法，其特征在于，所述关联信息包括多个关联项，所述根据关联信息建立所述目标对象的关注点进一步包括：
    对所述关联信息中的每个关联项进行分词，并对分词之后的所述关联项中的多个分词分别进行词性标注；以及
    根据所述分词之后的所述关联项中多个分词的词性对所述关联项进行筛选，并根据筛选之后的关联项的出现频率建立所述目标对象的关注点。
  4. 如权利要求1-3任一项所述的方法，其特征在于，还包括：
    根据所述目标对象建立品牌列表，其中，所述品牌列表中包括多个品牌；以及
    针对所述多个品牌分别建立搜索数据库，所述搜索数据库包括所述多个品牌及所述多个品牌对应的关注点。
  5. 如权利要求4所述的方法，其特征在于，还包括：
    接收所述用户选择的筛选条件；以及
    根据所述用户选择的筛选条件查询所述搜索数据库以获取查询结果，并提供至所述用户。
  6. 如权利要求1所述的方法，其特征在于，所述根据所述关注点生成所述目标对象对应的至少一个筛选条件进一步包括：
    对所述关注点的出现频次进行统计；
    根据所述关注点的出现频次对所述关注点进行排序；以及
    根据排序前预设位数的关注点生成所述目标对象对应的至少一个筛选条件。
  7. 如权利要求1所述的方法，其特征在于，还包括：
    获取所述关联信息的来源信息；
    根据所述关联信息的来源信息生成对应关注点的权重，以根据所述关注点的出现频次和所述权重对所述关注点进行排序。
  8. 如权利要求1所述的方法，其特征在于，所述筛选条件为多个，还包括：
    获取所述用户的历史记录，根据所述用户的历史记录对所述多个筛选条件进行排序，并将排序之后的所述多个筛选条件的至少一部分提供至所述用户。
  9. 一种筛选条件提供装置，其特征在于，包括：
    关联信息获取模块，用于获取目标对象对应的关联信息；
    关注点建立模块，用于根据所述关联信息建立所述目标对象的关注点；
    筛选条件生成模块，用于根据所述关注点生成所述目标对象对应的至少一个筛选条件，并将所述至少一个筛选条件的至少一部分提供至用户。
  10. 如权利要求9所述的筛选条件提供装置，其特征在于，所述关联信息包括评论信息。
  11. 如权利要求9所述的筛选条件提供装置，其特征在于，所述关联信息包括多个关联项，所述关注点建立模块进一步包括：
    分词子模块，用于对所述关联信息中的每个关联项进行分词，并对分词之后的所述关联项中的多个分词分别进行词性标注；
    筛选子模块，用于根据所述分词之后的所述关联项中多个分词的词性对所述关联项进行筛选；以及
    生成子模块，用于根据筛选之后的关联项的出现频率建立所述目标对象的关注点。
  12. 如权利要求9所述的筛选条件提供装置，其特征在于，还包括：
    搜索数据库建立模块，用于建立品牌列表，其中，所述品牌列表中包括多个品牌，并针对所述多个品牌建立搜索数据库，所述搜索数据库包括所述多个品牌及所述多个品牌对应的关注点。
  13. 如权利要求12所述的筛选条件提供装置，其特征在于，还包括：
    搜索模块，用于接收所述用户选择的筛选条件，并根据所述用户选择的筛选条件查询所述搜索数据库以获取查询结果，以及将查询结果提供至所述用户。
  14. 如权利要求9所述的筛选条件提供装置，其特征在于，所述筛选条件生成模块，具体用于对所述关注点的出现频次进行统计，并根据所述关注点的出现频次对所述关注点进行排序，以及根据排序前预设位数的关注点生成所述目标对象对应的至少一个筛选条件。
  15. 如权利要求14所述的筛选条件提供装置，其特征在于，所述筛选条件生成模块，还用于获取所述关联信息的来源信息，并根据所述关联信息的来源信息生成对应关注点的权重，以根据所述关注点的出现频次和所述权重对所述关注点进行排序。
  16. 如权利要求9所述的筛选条件提供装置，其特征在于，所述筛选条件为多个，所述筛选条件生成模块，还用于获取所述用户的历史记录，根据所述用户的历史记录 对所述多个筛选条件进行排序，并将排序之后的所述多个筛选条件的至少一部分提供至所述用户。
  17. 一种搜索方法，其特征在于，包括以下步骤：
    接收搜索请求；
    根据所述搜索请求生成与所述搜索请求所针对目标对象相关联的至少一个筛选条件；以及
    将所述至少一个筛选条件的至少一部分提供至用户。
  18. 如权利要求17所述的方法，其特征在于，还包括：
    获取目标对象对应的关联信息；
    根据所述关联信息建立所述目标对象的关注点；以及
    根据所述关注点生成所述目标对象对应的所述至少一个筛选条件。
  19. 如权利要求18所述的方法，其特征在于，所述关联信息包括多个关联项，所述根据关联信息建立所述目标对象的关注点进一步包括：
    对所述关联信息中的每个关联项进行分词，并对分词之后的所述关联项中的多个分词分别进行词性标注；以及
    根据所述分词之后的所述关联项中多个分词的词性对所述关联项进行筛选，并根据筛选之后的关联项的出现频率建立所述目标对象的关注点。
  20. 如权利要求18所述的方法，其特征在于，所述根据所述关注点生成所述目标对象对应的至少一个筛选条件进一步包括：
    对所述关注点的出现频次进行统计；
    根据所述关注点的出现频次对所述关注点进行排序；以及
    根据排序前预设位数的关注点生成所述目标对象对应的至少一个筛选条件。
  21. 如权利要求18所述的方法，其特征在于，还包括：
    获取所述关联信息的来源信息；
    根据所述关联信息的来源信息生成对应关注点的权重，以使根据所述关注点的出现频次和所述权重进行排序。
  22. 如权利要求17所述的方法，其特征在于，所述筛选条件为多个，还包括：
    获取所述用户的历史记录，根据所述用户的历史记录对所述多个筛选条件进行排序，并将排序之后的所述多个筛选条件的至少一部分提供至所述用户。
  23. 一种搜索装置，其特征在于，包括：
    搜索模块，用于接收搜索请求；
    生成模块，用于根据所述搜索请求生成与所述搜索请求所针对目标对象相关联的 至少一个筛选条件，并将所述至少一个筛选条件的至少一部分提供至用户。
  24. 如权利要求23所述的装置，其特征在于，还包括：
    获取模块，用于获取目标对象对应的关联信息；
    建立模块，用于根据所述关联信息建立所述目标对象的关注点；以及
    所述生成模块，还用于根据所述关注点生成所述目标对象对应的所述至少一个筛选条件。
  25. 如权利要求24所述的装置，其特征在于，所述关联信息包括多个关联项，所述建立模块具体包括：
    分词子模块，用于对所述关联信息中的每个关联项进行分词，并对分词之后的所述关联项中的多个分词分别进行词性标注；以及
    筛选子模块，用于根据所述分词之后的所述关联项中多个分词的词性对所述关联项进行筛选；以及
    生成子模块，用于根据筛选之后的关联项的出现频率建立所述目标对象的关注点。
  26. 如权利要求24所述的装置，其特征在于，
    所述生成模块，具体用于对所述关注点的出现频次进行统计，并根据所述关注点的出现频次对所述关注点进行排序；以及根据排序前预设位数的关注点生成所述目标对象对应的至少一个筛选条件。
  27. 如权利要求24所述的装置，其特征在于，
    所述生成模块，还用于获取所述关联信息的来源信息，并根据所述关联信息的来源信息生成对应关注点的权重，以根据所述关注点的出现频次和所述权重进行排序。
  28. 如权利要求23所述的装置，其特征在于，所述筛选条件为多个，所述生成模块，还用于获取所述用户的历史记录，根据所述用户的历史记录对所述多个筛选条件进行排序，并将排序之后的所述多个筛选条件的至少一部分提供至所述用户。
PCT/CN2014/088306 2013-10-15 2014-10-10 筛选条件提供方法、装置和搜索方法、装置 WO2015055094A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310482406.3A CN103530376B (zh) 2013-10-15 2013-10-15 筛选条件提供方法、装置和搜索方法、装置
CN201310482406.3 2013-10-15

Publications (1)

Publication Number Publication Date
WO2015055094A1 true WO2015055094A1 (zh) 2015-04-23

Family

ID=49932385

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/088306 WO2015055094A1 (zh) 2013-10-15 2014-10-10 筛选条件提供方法、装置和搜索方法、装置

Country Status (2)

Country Link
CN (1) CN103530376B (zh)
WO (1) WO2015055094A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110109901A (zh) * 2018-02-02 2019-08-09 北京京东尚科信息技术有限公司 筛选目标对象的方法和装置
CN111723117A (zh) * 2020-06-19 2020-09-29 深圳市蓝思网络技术有限公司 大数据产品开发筛选方法、装置及计算机可读存储介质

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103530376B (zh) * 2013-10-15 2016-03-16 北京百度网讯科技有限公司 筛选条件提供方法、装置和搜索方法、装置
CN103886037B (zh) * 2014-03-10 2018-01-19 华为技术有限公司 一种数据筛选的方法及装置
CN104408103B (zh) * 2014-11-19 2018-05-18 北京国双科技有限公司 文本内容的筛选方法及装置
CN105677162B (zh) * 2014-11-19 2019-12-03 深圳市腾讯计算机系统有限公司 匹配条件列表的显示方法和装置
CN106156160B (zh) * 2015-04-15 2020-04-14 腾讯科技（深圳）有限公司 查找企业账号的方法和装置
CN107016586A (zh) * 2016-01-27 2017-08-04 阿里巴巴集团控股有限公司 提供品牌列表信息的方法及装置
CN107092608B (zh) * 2016-04-16 2021-03-23 口碑控股有限公司 一种目标对象的搜索、推荐方法和设备
CN106203292A (zh) * 2016-06-28 2016-12-07 广东欧珀移动通信有限公司 一种图像的增强现实处理的方法、装置及移动终端
CN108733685B (zh) * 2017-04-17 2023-12-26 阿里巴巴集团控股有限公司 对象交互、浏览、搜索和配置方法及装置
CN109213920A (zh) * 2017-06-29 2019-01-15 阿里巴巴集团控股有限公司 搜索方法、客户端、服务器和存储介质
CN110019990B (zh) * 2017-07-14 2023-05-23 阿里巴巴集团控股有限公司 样本筛选的方法和装置、业务对象数据搜索的方法和装置
CN108846055B (zh) * 2018-05-31 2021-08-24 北京五八信息技术有限公司 搜索处理方法、装置、电子设备及可读存储介质
WO2020093289A1 (zh) * 2018-11-07 2020-05-14 深圳市欢太科技有限公司 资源推荐方法、装置、电子设备及存储介质
CN109543188A (zh) * 2018-11-23 2019-03-29 珠海格力电器股份有限公司 一种对象关联方法、装置、服务器及可读存储介质
CN111046222B (zh) * 2019-04-29 2023-06-27 广东小天才科技有限公司 一种听写列表的生成方法及电子设备
CN112558845A (zh) * 2019-09-25 2021-03-26 北京国双科技有限公司 表格操作方法及表格操作装置
CN112579647A (zh) * 2020-12-22 2021-03-30 科华恒盛股份有限公司 一种数据信息获取方法及相关装置
CN112925969A (zh) * 2021-03-23 2021-06-08 北京金山云网络技术有限公司 一种搜索页面的显示方法和装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1388460A (zh) * 2001-05-29 2003-01-01 塑网科技股份有限公司 智能选择产品的方法
CN101055587A (zh) * 2007-05-25 2007-10-17 清华大学 一种基于用户行为信息的搜索引擎检索结果重排序方法
CN101968788A (zh) * 2009-07-27 2011-02-09 富士通株式会社 提取商品属性信息的方法和设备
CN102467505A (zh) * 2010-11-03 2012-05-23 阿里巴巴集团控股有限公司 一种产品信息的搜索方法、系统及装置
CN103530376A (zh) * 2013-10-15 2014-01-22 北京百度网讯科技有限公司 筛选条件提供方法、装置和搜索方法、装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1388460A (zh) * 2001-05-29 2003-01-01 塑网科技股份有限公司 智能选择产品的方法
CN101055587A (zh) * 2007-05-25 2007-10-17 清华大学 一种基于用户行为信息的搜索引擎检索结果重排序方法
CN101968788A (zh) * 2009-07-27 2011-02-09 富士通株式会社 提取商品属性信息的方法和设备
CN102467505A (zh) * 2010-11-03 2012-05-23 阿里巴巴集团控股有限公司 一种产品信息的搜索方法、系统及装置
CN103530376A (zh) * 2013-10-15 2014-01-22 北京百度网讯科技有限公司 筛选条件提供方法、装置和搜索方法、装置

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110109901A (zh) * 2018-02-02 2019-08-09 北京京东尚科信息技术有限公司 筛选目标对象的方法和装置
CN110109901B (zh) * 2018-02-02 2024-05-24 北京京东尚科信息技术有限公司 筛选目标对象的方法和装置
CN111723117A (zh) * 2020-06-19 2020-09-29 深圳市蓝思网络技术有限公司 大数据产品开发筛选方法、装置及计算机可读存储介质

Also Published As

Publication number Publication date
CN103530376B (zh) 2016-03-16
CN103530376A (zh) 2014-01-22

Similar Documents

Publication Publication Date Title
WO2015055094A1 (zh) 筛选条件提供方法、装置和搜索方法、装置
US7788131B2 (en) Advertising keyword cross-selling
CN107729336B (zh) 数据处理方法、设备及系统
KR102075833B1 (ko) 미술 작품 추천 큐레이션 방법 및 시스템
US9298814B2 (en) Systems and methods for classifying electronic documents
US9990368B2 (en) System and method for automatic generation of information-rich content from multiple microblogs, each microblog containing only sparse information
KR101695011B1 (ko) 토픽별 오피니언과 소셜 영향력자를 기반으로 토픽을 탐지하고 추적하는 시스템 및 방법
US20100153210A1 (en) Apparatus and method for selecting online advertisement based on contents sentiment and intention analysis
CN106339502A (zh) 一种基于用户行为数据分片聚类的建模推荐方法
WO2011080899A1 (ja) 情報推薦方法
US8997008B2 (en) System and method for searching through a graphic user interface
KR101355945B1 (ko) 온라인 문맥기반 광고 장치 및 방법
KR20160131477A (ko) 연관 검색어를 이용한 관심사 카테고리 기반 이커머스 시스템
CN115018255A (zh) 一种基于集成学习数据挖掘技术的旅游景区评价信息质量有效性分析方法
CN108205545B (zh) 一种为用户提供推荐信息的方法与设备
TW201234204A (en) Opportunity identification for search engine optimization
KR101850853B1 (ko) 빅데이터를 이용한 검색 방법 및 장치
KR101903717B1 (ko) 상품 정보 자동 추출 방법 및 장치
KR101650888B1 (ko) 컨텐츠 수집, 추천 시스템 및 방법
JP2007249600A (ja) 目的データをカテゴリに分類する方法
KR102170535B1 (ko) 감성 분석을 통한 사용자 선호 기반의 검색 장치 및 방법
WO2008032037A1 (en) Method and system for filtering and searching data using word frequencies
JP2002278983A (ja) 要約抽出プログラム、文書分析支援プログラム、要約抽出方法、文書分析支援方法、文書分析支援システム
WO2015047075A1 (en) A system and method for ranking recommendations
JP2002215642A (ja) フィードバック型インターネット検索方法及びその方法を実施するためのシステムとプログラム記録媒体

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: 14853768

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: 14853768

Country of ref document: EP

Kind code of ref document: A1