JP5480058B2 - Advertisement matching apparatus, method and program - Google Patents

Advertisement matching apparatus, method and program Download PDF

Info

Publication number
JP5480058B2
JP5480058B2 JP2010174388A JP2010174388A JP5480058B2 JP 5480058 B2 JP5480058 B2 JP 5480058B2 JP 2010174388 A JP2010174388 A JP 2010174388A JP 2010174388 A JP2010174388 A JP 2010174388A JP 5480058 B2 JP5480058 B2 JP 5480058B2
Authority
JP
Japan
Prior art keywords
action
search
advertisement
word
candidate
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
JP2010174388A
Other languages
Japanese (ja)
Other versions
JP2012037920A (en
Inventor
浩司 塚本
Original Assignee
ヤフー株式会社
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 ヤフー株式会社 filed Critical ヤフー株式会社
Priority to JP2010174388A priority Critical patent/JP5480058B2/en
Publication of JP2012037920A publication Critical patent/JP2012037920A/en
Application granted granted Critical
Publication of JP5480058B2 publication Critical patent/JP5480058B2/en
Application status is Active legal-status Critical
Anticipated expiration legal-status Critical

Links

Images

Description

  The present invention relates to a search-linked advertisement.

  2. Description of the Related Art Conventionally, in the field of so-called web search, a search-linked advertisement (for example, see Patent Document 1) in which an advertisement of an advertiser who bids on a keyword is displayed is widely known. In a typical search-linked advertisement, an advertiser submits a string advertisement with a link and bids for a desired word (bid word). When the word is searched, A predetermined number of the character string advertisements are displayed on the search result screen in descending order, and the advertiser is charged for the clicked advertisement based on the bid price.

Japanese Patent No. 3676999

  By the way, when the keyword inputted by the searcher is, for example, one word “snake”, it is difficult to find out what intention the searcher is searching for the word. For this reason, conventionally, there is a possibility that an advertisement that is not appropriate in view of the search intention of the searcher may be issued. In this case, there is a problem that the effect of the advertisement cannot be expected. For example, mismatches such as advertising for snake breeding when you want to get rid of a snake will reduce the click-through rate (CTR) of the ad, which is not beneficial to both advertisers and searchers. .

  In view of the above problems, an object of the present invention is to realize an effective search-linked advertisement that matches a searcher's search intention.

  Based on the above object, one aspect (1) of the present invention is an advertisement matching device that associates an advertisement with an action character string representing an action specified as a display condition of the advertisement in a character string. An attribute storage word, a search query acquisition unit that acquires a search query, and an attribute that generates an attribute candidate word that is a word candidate representing an attribute related to the search query based on the acquired search query Candidate generation means, and pattern processing means for extracting the attribute by performing word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms on the generated attribute candidate word Action candidate generating means for generating action candidates corresponding to the extracted attributes, and an action for outputting the generated action candidates Candidate output means, action accepting means for receiving an action designation among the output action candidates, search means for executing a search according to the specified action, and a word corresponding to the action related to the search A character string obtaining means for obtaining a search character string based on the set of the above, an advertisement character string obtained from the advertisement stored in the advertisement storage means, an action character string corresponding to the advertisement, And an advertisement selection means for selecting an advertisement to be output together with the search result.

  According to another aspect (7) of the present invention, the above aspect is captured by a method category, which is an advertisement matching method in which an advertisement and an action designated as a display condition for the advertisement are represented by character strings. The computer realizes advertisement storage means that stores character strings in association with each other, the computer acquires a search query, and the computer searches for the search query based on the acquired search query. An attribute candidate generation process for generating an attribute candidate word that is a word candidate representing an attribute related to a query, and at least one of removal of a predetermined type of word and unification of synonyms with respect to the generated attribute candidate word Pattern processing for extracting the attribute by performing word processing according to a predetermined pattern including, and a computer corresponding to the extracted attribute. Action candidate generation processing for generating action candidates; action candidate output means for outputting the generated action candidates by the computer; and action for accepting specification of an action among the output action candidates A reception process, a search process in which the computer performs a search according to the specified action, and a character string acquisition process in which the computer acquires a search character string based on a set of words corresponding to the action related to the search And the computer determines the suitability between the obtained search string and the action string corresponding to the advertisement from the advertisements stored in the advertisement storage means. And an advertisement selection process for selecting an advertisement to be output together with the result.

  Another aspect (8) of the present invention is an advertisement matching program that selects the advertisement by controlling the computer by capturing the above aspect in the category of the computer program. An action storage unit that stores an action character string representing an action specified as a display condition in association with a character string is realized, and a computer is made to acquire a search query. Based on the search query, an attribute candidate word that is a word candidate representing an attribute related to the search query is generated, and the computer is configured to remove a predetermined type of word and unify synonyms for the generated attribute candidate word. The attribute is extracted by performing word processing with a predetermined pattern including at least one, and the computer extracts the attribute. And generating a candidate for the action corresponding to the attribute that has been generated, causing the computer to output the candidate for the generated action, causing the computer to accept the specification of the action among the output candidate for the action, and causing the computer to The search is executed in accordance with the specified action, the computer is caused to acquire a search character string based on a set of words corresponding to the action related to the search, and the computer stores the search character string. The advertisement to be output together with the result of the search is selected by judging the suitability of the acquired search character string and the action character string corresponding to the advertisement from the advertisement. .

  In this way, in the action proposal type search that suggests action candidates from the specified search query and specifies what suits the search intention, the search character string based on the action specified in the search and the display for each advertisement By selecting and outputting an advertisement based on the suitability judgment of an action character string in which the desired action is associated in advance, an effective search-linked advertisement that matches the searcher's search intention can be realized and the CTR can be improved. The “action” is a search intention represented by a character string using a plurality of words.

  Another aspect (2) of the present invention is an advertisement reception device, which is a basic word acquisition means for acquiring a basic word as a basis for an advertisement bid, and an attribute relating to the basic word based on the acquired basic word Attribute candidate generating means for generating an attribute candidate word that is a candidate for a word representing a word, and a word having a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms with respect to the generated attribute candidate word Pattern processing means for extracting the attribute by performing processing, action candidate generation means for generating an action candidate corresponding to the extracted attribute, action candidate output means for outputting the generated action candidate , Out of the outputted action candidates, action accepting means for accepting the designation of the action, and the designated action And a bid accepting means for accepting a bid for the advertisement and storing the action character string representing the designated action in a character string and the received advertisement in association with each other in a predetermined advertisement storage means. Features.

  Another aspect (6) of the present invention is an advertisement system, and includes the advertisement matching apparatus according to the aspect (1) and the advertisement receiving apparatus according to any one of the aspects (2) to (5). It is characterized by that.

  As described above, based on the word specified at the time of ad submission, the candidate word is generated as a candidate word that represents the attribute of the search intent, and the predetermined type of word is removed in accordance with the search for outputting the advertisement. The searcher accepts the specification of the action to be displayed from the action candidates according to the search intention attribute extracted by, for example, and stores it in association with the advertisement as an action character string, thereby allowing the searcher to specify the action. This makes it easy to submit advertisements based on search, enabling effective search-linked advertisements that match the search intention of the searcher.

  According to another aspect (3) of the present invention, in any one of the above aspects, the attribute candidate generation unit includes a related word storage unit that stores a plurality of words used simultaneously in the search query in association with each other as related words. Means extracts from the related word storage means the word associated with the input basic word as the related word as the attribute candidate word; and the pattern processing means performs the attribute candidate word as the word processing. A proper noun removing means for removing a proper noun from the word, a noun adjective removing means for removing a nounative adjective from the attribute candidate word as the word processing, and a synonym for the attribute candidate word as the word processing Synonym processing means, and attribute candidates remaining as a result of the removal and unification are defined as the attributes.

  Thus, based on the basic word, the related words that are often combined are attribute candidate words, the proper nouns, noun adjectives, and synonyms are excluded, and the remaining words are the attributes of the search intention. By effectively extracting the attributes that can be used, and making it possible to bid for actions according to those attributes, effective search-linked advertising that matches the searcher's search intention is achieved more effectively. can do.

  According to another aspect (4) of the present invention, in any one of the above aspects, the action candidate output unit performs a search using the search character string corresponding to the action candidate, and is based on the search result. At least one of an additional keyword or an image is output together with an action candidate, the action accepting unit accepts designation of at least one of the additional keyword or an image regarding the action candidate, and the bid accepting unit relates to the action candidate Accepting a bid for an advertisement based on at least one of the additional keyword or image, an action character string representing the specified action as a character string by adding at least one information of the additional keyword or image, and the accepted advertisement To be stored in a predetermined advertisement storage means And features.

  In this way, the search character string is searched in advance for each action candidate, and additional keywords such as more specific options and images such as product photos for each product are output together with the candidates based on the search result. This makes it easy to narrow down the information to be acquired, and in the case of images in particular, the contents of the information provided when candidates are specified can be easily understood from the image, and the search usability is improved. Advertisements that specify images and images will be more appealing to the exact demand, and will have an excellent effect.

  According to another aspect (5) of the present invention, in any one of the above aspects, the action candidate output means performs a search corresponding to the action candidates, and action candidates whose search results are less than a predetermined number are displayed. It is excluded from.

  In this way, when there is no search result for an action, the action is excluded from the candidates to be displayed to the advertiser, so that when the searcher selects, the candidate of the action that has no candidate search result is advertised. It is possible to avoid the waste that is displayed and judged by the main party, and the usability in advertising bidding is improved.

  It should be noted that a category (method for the apparatus, program for the method, etc.) different from each of the above-described modes and more specific modes described below are also included in the present invention. For different categories, “means” shall be appropriately read as “process” or “step”.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to implement | achieve the effective search link advertisement according to the search intention of a searcher.

The figure which shows the whole structure of the form (henceforth "embodiment") for implementing this invention. The block diagram of the advertisement reception apparatus (reception apparatus) in embodiment. The block diagram of the advertisement matching apparatus (allocation apparatus) in embodiment. The figure which illustrates the data (related word) used by embodiment. The figure which illustrates the data (correspondence dictionary) used by embodiment. The flowchart which shows the process sequence regarding reception of the advertisement in embodiment. The conceptual diagram which shows the example which extracts an attribute word from an attribute candidate word in embodiment. The figure which shows an example of the screen display in embodiment. The figure which illustrates the data (advertisement) used by embodiment. The flowchart which shows the process sequence regarding the allocation of the advertisement in embodiment. The figure which shows an example of the screen display in embodiment. The figure which shows an example of the screen display in embodiment. The figure which shows an example of the screen display in embodiment. The figure which shows an example of the screen display in embodiment.

  Next, embodiments of the present invention will be described with reference to the drawings. It should be noted that assumptions common to those already described in the background art and problems are omitted as appropriate.

[1. Overview of action proposal type search)
This embodiment relates to advertisement distribution to action proposal type search. Here, the action proposal type search is based on a given search word (search query), and a searcher is displayed while displaying candidates for possible actions (representing a search intention by a character string using a plurality of words). Is a service that outputs search results based on actions specified by.

  For example, FIG. 11 shows an example of the action proposal type search screen. As a result of the user inputting “snake” as the search query K1 in the input field F and operating the search button B2 with the mouse pointer P, several action candidate CAs are displayed. The state where is displayed. In this figure, the underline accompanying each action candidate CA is a hyperlink of a web page for transitioning to a search result page when the search is further advanced by specifying the action.

[2. Constitution〕
As shown in FIG. 1, in the present embodiment, which handles advertisements displayed by action proposal type search as described above, an advertisement matching apparatus 1 (hereinafter referred to as “allocation”) that allocates advertisements (also referred to as “allocation”) for each search. Device 1 ”or simply“ allocation device ”) and an advertisement receiving device 2 that accepts these advertisements from the advertiser (hereinafter simply referred to as“ accepting device 2 ”or simply“ accepting device ”). It is about. Both the allocation device 1 and the reception device 2 can actually be configured to use the function in conjunction with another search device 3 that processes the above-described action proposal type search. Although the accepting device 2 also includes the search device 3 having the same configuration (FIGS. 2 and 3), the allocation device 1, the accepting device 2, and the advertising system X can also be understood as not including the search device 3.

  In addition, in order to allow the allocation apparatus 1 to access the advertisement data received by the reception apparatus 2, it is stored in a predetermined advertisement storage means 95, but the advertisement storage means 95 is configured as an independent data server. Alternatively, it may be provided in one of the allocation device 1 or the accepting device 2 and accessed from the other, or may be provided in both and copied. Here, for the sake of explanation, like the search device 3, the advertisement storage means 95 is present in both the accepting device 2 and the allocation device 1 (FIGS. 2 and 3), and the contents are automatically copied (synchronized). To do.

  Therefore, hereinafter, the search device 3 and the advertisement storage unit 95 will be described by including them in the reception device 2 (FIG. 2) and the allocation device 1 (FIG. 3). The allocation device 1 and the acceptance device 2 are server devices that accept advertisements and respond to search requests by communicating with terminals T (T1, T2) that are accessed via the communication network N via HTTP or the like. . Note that the communication network N is the Internet, a mobile phone network, a LAN, and the like. Although two are shown in FIG. 1, these may be the same.

  The accepting device 2 (FIG. 2) and the allocation device 1 (FIG. 3) have at least an arithmetic control unit 6 such as a CPU and an external storage device (HDD or the like) or a main memory or the like as a general computer configuration. The apparatus 7 and the communication means 8 (LAN adapter, a communication gateway, etc.) with the communication network N are provided. A terminal T (T1, T2) is a client device used by a user, such as a personal computer (PC), a smart phone, or a mobile phone terminal device.

  In the receiving device 2 (FIG. 2) and the allocation device 1 (FIG. 3), a predetermined computer program (not shown) previously stored (installed) in the storage device 7 controls the arithmetic control unit 6 so that Elements (12, 92, 10, 98, etc.) such as each means shown in 2 and 3 are realized. Among these elements, the information storage means can be realized in the storage device 7 in any format such as various files and databases (also referred to as “DB”), variables such as arrays, various stacks and registers, and system setting values.

  Among such storage means, the related word storage means 25 of the accepting device 2 (FIG. 2) and the allocation device 1 (FIG. 3) are simultaneously input as separated by a space when performing an AND search with a search query. It is means for storing a plurality of words used in association with each other as related words. In the example shown in FIG. 4, for each reference word (for example, “snake”), a group of related words (for example, “bred”, “picture book”, etc.) is associated with “snake”. Is free, and instead of (or in combination with) a group of related words for each reference word, a graph index that shows the distance between multiple words can be used to determine more words. High accuracy can be achieved.

  Further, as illustrated in FIG. 5, the correspondence dictionary storage means 75 (FIG. 2, FIG. 3) correlates each type of attribute word that may be extracted for the basic word or the search query, and performs a corresponding action. This is means for storing candidate patterns and search character string patterns. The index storage means 85 (FIGS. 2 and 3) is means for storing index data for web search according to the prior art.

  Further, the advertisement storage unit 95 (FIGS. 2 and 3) is a unit that stores an advertisement and an action character string representing an action designated as a display condition for the advertisement in a character string in association with each other. . Each means other than the storage means is a processing means for realizing and executing the following information processing functions and operations.

[3. (Actions related to advertisement reception)
As an operation of the present embodiment configured as described above, first, a flowchart regarding advertisement reception is shown in FIG. The provision of the accepted advertisement will be described later along the flowchart of FIG.

[3-1. Overview of ad reception]
In this embodiment, when an advertiser (FIG. 1: Advertiser A123, for example, a pest control company) submits an advertisement and places a bid, the reception device 2 is accessed from the terminal T1 used by the advertiser and includes login and the like. Through a predetermined operation, a word that is the basis of the advertisement bid, that is, a keyword (referred to as “basic word”) is input. Then, the accepting device 2 outputs the action candidate corresponding to the search intention that can be considered from the basic word by the following processing according to the action proposal type search by the search device 3 linked with the allocation device 1. Accepts an action specification from the main, and accepts an advertisement bid and associates it with an action so that the advertisement is displayed when the action is specified in the action proposal type search.

[3-2. (Attribute extraction)
Specifically, first (FIG. 6), when the basic word acquisition unit 12 acquires the basic word (for example, “snake”) as the basis of the advertisement bidding (step S21), the attribute candidate generation unit 20 Based on the acquired basic word, an attribute candidate word that is a word candidate representing an attribute of the search intention related to the basic word is generated (step S22). The attribute candidate word is, for example, a related word that is easily AND-searched with the word of the basic word. Here, the attribute candidate generating means 20 is input from the related word storage means 25 (FIG. 4) as an input search query and related words. The associated words are extracted as attribute candidate words (for example, FIG. 7 (1)).

  Subsequently, the pattern processing means 30 performs word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms on the generated attribute candidate word (for example, FIG. 7 (1)). The attribute of the search intention is extracted by applying (step S23). More specifically, the pattern processing means 30 removes the proper noun from the attribute candidate word as the word processing by the proper noun removal means 31 that is difficult to become an attribute (for example, “Miyakojima” in FIG. 7 (2)), Further, the noun adjectives are removed by the noun adjective removing means 32 which is difficult to become an attribute (for example, “gigantic” in FIG. 7 (3)), and the synonym processing means 33 unifies the synonyms for the remaining attribute candidate words. (For example, in FIG. 7 (4), “how to keep”, “dream fortune-telling”, and “extinguishment” are removed), and attribute candidate words remaining as a result of the removal and unification are set as attribute words representing attributes of the search intention. The synonyms here are not limited to complete synonyms, but include words having a predetermined relationship as similar themes.

[3-3. Action candidate generation and output)
Subsequently, the action candidate generating means 40 generates action candidates corresponding to the attribute words extracted as attributes (referred to as “action candidates”) (step S24), and the generated action candidates are output as action candidates. The means 50 outputs (step S25). Note that the remaining attribute words as shown in FIG. 7 (4) are not left as they are. If the reranking means 38 rearranges (reranks) the action words in the order corresponding to the rearrangement results, By adjusting the ranking criteria, the display order of action candidates according to the purpose and circumstances of system operation can be realized.

  Here, the sorting criteria are arbitrary, but for example, based on the words of the basic word, they are sorted by a statistical index (correlation coefficient, mutual information, etc.) based on the number of appearances and the number of co-occurrence of each attribute word. In other words, the order of action candidates using statistics can be realized. In addition, for example, for each word of the same query, initially, each action candidate corresponding to the attribute word is actually displayed randomly to various users, and as a result, a click rate (CTR: click through rate, etc.) If the action candidates corresponding to each of them are output in the order of priority selected from the attribute words with high priority, the click rate can be improved and the search and advertisement opportunities can be effectively utilized.

For example, the attribute words “breeding, picture book, dream, pet, type, chick, extermination, princess, image ...” remaining in the example of FIG. 7 (4) based on the search query “snake” are arranged as described above. Consider a case where the result of the replacement (reranking) is an order such as “type, extermination, dream, illustration, image, breeding ...”. In this case, according to the type of each attribute word, the basic word and the attribute word are applied to the pattern of action candidates stored in the correspondence dictionary storage means 75 illustrated in FIG.

"I want to know about snake types"
"I want to know about snake extermination"
"I want to know about snake dreams"
"I want a snake illustration"

The action candidates generated as described above are output as web pages as illustrated in FIG. The example of FIG. 8 is a conceptual diagram showing an action candidate designation screen as part of screen transitions in advertisement submission and bidding. As a result of the advertiser inputting “snake” as the basic word X1 in the basic word input field F and operating the search button B1 with the mouse pointer P, several action candidates CA are displayed.

[3-4. (Reception of action and storage of advertisement data)
There is a check column CF corresponding to each action candidate CA, and with the mouse pointer P, one or more action candidate check columns CF to be used as advertisement display conditions are operated and checked. If the “Next” button B1 is operated, the action is designated, and the screen transitions to a subsequent screen such as a bid designation or confirmation screen.

  That is, as described above, when the action accepting unit 60 accepts designation of one or more actions that should be the advertisement display condition among the output action candidates (step S26 in FIG. 6), the bid accepting unit 92 Then, an action character string representing the designated action as a character string is generated (step S27), an advertisement bid is accepted based on the designated action, and the action character string is associated with the accepted advertisement. The data is stored in the storage means 95 (step S28).

  Here, the format of the action character string is arbitrary. If the action candidate character string as described above is used as it is (for example, “I want to know about snake extermination”), it is compatible with the application of natural language processing technology. For example, the processing is simplified based on the pattern of the search character string in the correspondence dictionary storage means 75 illustrated in FIG. As an example of the latter case, consider a case where “I want to know about snake extermination” is designated from among the action candidates CA shown in FIG. The attribute word “disinfecting” corresponding to this action candidate is “service name” in the type of attribute word shown in FIG.

The corresponding search string pattern is

[Web search]: {query word} and ({attribute word} or {synonym}) and “I want to know”

It is.

In addition to the query word “snake”, attribute word “disinfect”, and the synonym “defeat” obtained from the synonym or synonym data removed earlier, the search string is

[Web Search]: Snake and (exterminate or extermination) and want to know

From now on, the action character string may be a word pair such as snake, extermination, extermination, or wanting to know.

  FIG. 9 shows an example in which action candidate character strings as they are (for example, “I want to know about snake removal”) are used as action character strings, and advertisement data such as URLs and hyperlinks thereof are stored in association with each other. Show.

[4. (Advertising provisions)
Next, a processing procedure in which the allocation device 1 allocates (assigns) and displays the advertisement received as described above for the action proposal type search is shown in the flowchart of FIG.
[4-1. Overview from search query acquisition to search)
That is, in the allocation device 1, when the search query acquisition unit 10 acquires a search query transmitted from the searcher's terminal T 2 via the communication network N (step S 1), an action candidate is generated from the basic word in accepting the advertisement. Action candidates are generated in the same processing procedure as the processing procedure (steps S22 to S24 in FIG. 6) (steps S2 to S4 in FIG. 10).

  Specifically, also in the allocation device 1, the attribute candidate generation unit 20 uses a search query (for example, K1: “snake” in FIG. 11) to search for a word representing an attribute of a search intention related to the search query. An attribute candidate word that is a candidate is generated (step S2, FIG. 7 (1)), and the pattern processing means 30 performs at least one of removal of a predetermined type of word and unification of synonyms on the generated attribute candidate word. By performing word processing according to a predetermined pattern including (FIG. 7 (2) to FIG. 7 (4)), an attribute of search intention is extracted (step S3).

  Further, the action candidate generating means 40 generates action candidates corresponding to the extracted attribute (FIG. 7 (4)) (step S4), and the action candidate output means 50 generates the generated action candidates in FIG. Output as illustrated (step S5). When the action accepting unit 60 accepts designation of an action to be used as a search condition among the output action candidates (step S6), the retrieving unit 80 retrieves the content according to the designated action. Execute (Steps S7 to S8).

[4-2. Search string example)
Specifically, the search generation means 70 searches for a character string for search corresponding to the specified action (for example, already exemplified)

[Web Search]: Snake and (exterminate or extermination) and want to know

Etc.) (step S7), and the search means 80 performs a search using the generated search character string (step S8).

  If the specified actions are different, the contents of the search to be executed are generally different, but different actions such as “I want to know ○○ ○○” and “I want to check ○○ ○○” Even if is specified, since the word pairs corresponding to each action are common, the search results may be common. In the search string shown above, the search condition corresponding to the action “I want to know” is also included in the search formula, but “snake and (disinfection or extermination)” gives similar results to some extent. In some cases, even if the suggestion (output action candidate) contains a verb that directly expresses the user's intention of “I want to know” in the character string representing the action, the verb is included in the search expression. Is not essential.

[4-3. (Ad selection)
Subsequently, the character string acquisition unit 90 acquires a search character string based on a set of words corresponding to the action related to the search (step S9). As the “character string for search based on a set of words corresponding to an action”, if the search is performed with such a search character string, the search character string may be used as it is, but the search is a special search. If it is performed by natural language processing based on the action string itself (for example, “I want to know about snake removal”) instead of the character string, the word corresponding to the action is based on the action (string) This includes a case where a search character string including the set is not used for search but is acquired separately.

  In any case, based on the search character string acquired in this way, the advertisement selection means 98 selects an advertisement to be output together with the search result (step S10). In the selection of this advertisement, it is determined by determining the suitability between the action character string corresponding to each advertisement and the obtained search character string from the advertisements stored in the advertisement storage means 95. A predetermined number of advertisements are selected with priority given to high-quality advertisements.

  The suitability between the search character string and the action character string is determined by comparing both character strings into pairs of words and taking synonyms into account, and comparing them with a Japanese processing technique such as TF / IDF. For example, it may be determined that “I want to know about snake extermination” and “I want information on snake extermination” are relatively compatible. In addition, in selecting an advertisement, if the bid price for each advertisement that is charged when clicked, the CTR (click-through rate: click rate) of the advertisement, and the like are taken into account through eCPM, the advertising effect Will increase.

  The action character string and search character string are not limited to natural sentences based on word pairs or search expressions in which word pairs are connected by logical operators. It may have a function of switching the type of search engine or search service, such as an API URL. For example, as illustrated in FIG. 5, [Web search] [shopping search] or the like (actually, API URL or the like corresponding to each search type) is switched as a search character string pattern corresponding to the type of attribute word. If the character string for this is set, a search result highly compatible with the search intention can be provided to the user.

[4-4. (Display of advertisement)
The result output means 100 outputs the advertisement selected as described above to the searcher's terminal T2 as a web page together with the executed search result (step S11). For example, when the button B2 is operated by designating “I want to know about snake extermination” among the action candidates CA illustrated in FIG. 11, the search result display screen that is a transition destination based on the button B2 is illustrated in FIG. In addition, an advertisement “pest control is XX company” (advertising ID: 123-01 in FIG. 9) in which the same action “I want to know about snake control” is displayed in the advertisement column AD. Also, in this search result screen example, the display in the input field F is the character string “snake and (disinfection or extermination) and want to know” used for the search, and each link L to the corresponding web page is displayed. Has been.

[5. effect〕
As described above, in the present embodiment, in an action proposal type search that presents action candidates from a specified search query and specifies those that match the search intention, a search character string based on the action specified in the search (For example, FIG. 11) and an action character string (for example, FIG. 9) in which an action desired to be displayed for each advertisement is associated in advance, and an advertisement is selected and output based on the search intention of the searcher. Matched and effective search-linked advertisement (for example, FIG. 12) can be realized and CTR is also improved.

  In particular, in the present embodiment, a word of a predetermined type is generated after generating an attribute candidate word that is a candidate for a word indicating an attribute of a search intention in accordance with a search for outputting an advertisement using a word specified at the time of advertisement submission as a basic word. (For example, FIG. 7) is received (for example, FIG. 7), the designation of the action to be displayed as an advertisement is accepted from the action candidates corresponding to the attribute of the search intention (for example, FIG. 8), and the action character string is associated with the advertisement. By memorizing (for example, FIG. 9), it is possible to smoothly submit an advertisement on the assumption that a searcher designates an action, and an effective search-linked advertisement (for example, FIG. 12) suitable for the searcher's search intention is provided. realizable.

  Further, in the present embodiment, related words that are often combined based on basic words are attribute candidate words, and proper nouns, nounative adjectives, and synonyms are excluded (for example, FIG. 7). With this simple process, you can efficiently extract attributes that may become search intents, and make bids for actions according to those attributes, enabling effective searches that match the searcher's search intentions. Linked advertisements can be realized more effectively.

  Note that the action candidate output means 50 performs a search using a search character string corresponding to each action candidate in advance before outputting the action candidate to the advertiser or searcher, and the search result is a predetermined number. Action candidates that are smaller (for example, 0. “Less than a predetermined number” includes “below a predetermined number”) may be excluded from display targets.

  In this way, when there are no search results for an action during ad submission and search, the action is excluded from the display target, thereby avoiding a wasteful operation of 0 search results for the selected candidate. , Usability of ad trafficking and search is improved.

[6. (Additional keywords and image output)
Further, an advertisement may be submitted by designating an additional keyword or image for further narrowing down the search result based on the action candidates. For example, the action candidate output means 50 of the receiving device 2 at the time of accepting an advertisement or the allocation device 1 at the time of searching performs a search using a search character string corresponding to an action candidate, and based on the search result Output additional keywords along with action candidates. For example, the screen example of FIG. 13 is what the searcher sees at the time of the search. The additional keyword AA is associated with the action AC corresponding to the attribute word “present” extracted for the search query “Mother's Day”. Hyperlinks such as “carnation” and “stall” are also displayed.

  As such additional keywords, for example, if you click on “Carnation” searched for “Mother's Day Gift Purchase”, you can immediately switch to the search result page for “Mother's Day Gift Purchase” and “Carnation” It is convenient to do so. For example, these additional keywords are extracted from a web page of a search result or a web page indicated by a URL included in the search result by applying a technique such as a known TF / IDF.

  Further, the action candidate output means 50 of each of the accepting device 2 and the allocation device 1 performs a search using a search character string corresponding to the action candidate, and outputs an image based on the search result together with the action candidate. For example, in the screen example at the time of search shown in FIG. 14, the additional keyword AA according to FIG. 13 is associated with the action A corresponding to the attribute word “present” extracted for the search query “Mother's Day”. In addition to “carnation” and “stall”, an image G of carnation or stall is displayed. Similar to the additional keyword, when such an image G is clicked, for example, a similar image search or the like may be performed in a shopping site.

  As an image based on the search result as described above, for example, a feature word is extracted from a search result web page or a web page indicated by a URL included in the search result by applying a known technique such as TF / IDF. The result of image search using the feature word can be used.

  At the time of advertisement submission, the advertiser designates an additional keyword or image by a click operation or a check column operation on the designation screen according to FIGS. 13 and 14, and the action receiving means 60 Accepting the designation of the image, the bid accepting means 92 is an advertisement storage means in which an action character string in which an action specified by the advertiser is added as a character string with additional keywords and image information added thereto is associated with the accepted advertisement. 95.

  In this manner, a search is performed in advance using a search character string for each action candidate, and based on the search result, additional keywords such as more specific options (for example, FIG. 13) and images such as product photos for each product are displayed. (For example, FIG. 14) By outputting together with candidates, it becomes easy to narrow down the information to be acquired. In particular, in the case of an image, the contents of information provided when a candidate is designated can be easily grasped from the image and searched. Since the usability is improved, advertisements that specify these additional keywords and images will appeal more accurately to the exact demand, and will exhibit excellent effects.

[7. Other embodiments]
In addition, said each embodiment is only an illustration, and this invention includes what is illustrated below and other embodiment other than that. For example, when accepting advertisements, providing a function to display the advertisement display simulation results based on the search character string to the advertiser, it is possible to support effective advertisement submission by confirming the advertisement display in advance. Become.

  For example, in FIG. 8, the underline attached to each action candidate CA is a hyperlink for transitioning to the search result page when the searcher further advances the search by designating the action. The Lord can also try. In order to test, a desired action candidate CA such as “I want to know about snake removal” may be operated as a hyperlink indicated by an underline, separately from the operation in the check column.

  In response to this operation, the search generation means 70 generates a search character string corresponding to the designated action according to the search, and the search means 80 executes a search using the generated search character string. The result output means 100 outputs the result of the search performed as another window display. The result of such a search is based on the search as illustrated in FIG.

  Each element such as the means shown in FIG. 1 to FIG. 3 is not limited to the arithmetic control unit of the computer, and may be realized by other information processing mechanisms such as an electronic circuit based on wired logic or the like. Further, each configuration diagram, data diagram, flowchart diagram, and the like are merely examples, and the presence / absence of each element, its order, specific contents, and the like can be changed as appropriate. For example, the apparatus of the present invention may be realized by using a plurality of apparatuses such as servers, and a configuration in which each storage unit is realized by a separate and independent server apparatus or system is also common. Depending on the function, the configuration can be flexibly changed, for example, by calling an external platform or the like with an API (application program interface) or network computing (so-called cloud or the like).

1 Advertising matching device (allocation device)
2 Advertisement reception device (reception device)
3 Search Device X Advertising System 6 Arithmetic Control Unit 7 Storage Device 8 Communication Unit 10 Search Query Acquisition Unit 20 Attribute Candidate Generation Unit 25 Related Word Storage Unit 30 Pattern Processing Unit 31 Proper Noun Removal Unit 32 Nounative Adjective Removal Unit 33 Synonym Processing Means 38 Reranking means 40 Action candidate generation means 50 Action candidate output means 60 Action acceptance means 70 Search generation means 75 Corresponding dictionary storage means 80 Search means 90 Character string acquisition means 92 Bid acceptance means 95 Advertisement storage means 98 Advertisement selection means 100 Result Output means CA Action candidate AA Additional keyword A123 Advertiser B1, B2 Search button F Input field G Image L Link N Communication network P Mouse pointer T (T1, T2) Terminal K1 Search query X1 Basic word

Claims (8)

  1. An advertisement storage means that stores an advertisement and an action character string that is designated as a display condition of the advertisement and that represents an action indicating a searcher's search intention in a character string;
    A search query acquisition means for acquiring a search query;
    Based on the acquired search query, attribute candidate generation means for generating an attribute candidate word that is a word candidate representing an attribute relating to the search query;
    Pattern processing means for extracting the attribute by performing word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms for the generated attribute candidate word;
    Corresponding to the attribute extracted, it shows a search intent of the searcher, and action candidate generating means for generating a candidate actions involving verbs search intent of the searcher,
    Action candidate output means for outputting the generated action candidates;
    Action accepting means for accepting specification of an action among the output action candidates,
    Search means for executing a search in accordance with the specified action;
    A character string acquisition means for acquiring a search character string based on a set of words corresponding to the action related to the search;
    It should be output together with the result of the search by judging the suitability between the obtained search character string and the action character string corresponding to the advertisement from the advertisements stored in the advertisement storage means. An ad selection means for selecting an ad;
    An advertisement matching device comprising:
  2. A basic word acquisition means for acquiring a basic word as a basis for an advertising bid;
    Based on the acquired basic word, attribute candidate generating means for generating an attribute candidate word that is a word candidate representing an attribute relating to the basic word;
    Pattern processing means for extracting the attribute by performing word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms for the generated attribute candidate word;
    Corresponding to the attribute extracted, it shows a search intent of the searcher, and action candidate generating means for generating a candidate actions involving verbs search intent of the searcher,
    Action candidate output means for outputting the generated action candidates;
    Action accepting means for accepting specification of an action among the output action candidates,
    A bid accepting unit that accepts a bid for an advertisement based on the specified action and stores an action character string representing the specified action in a character string and the received advertisement in association with each other in a predetermined advertisement storage unit When,
    An advertisement receiving apparatus characterized by comprising:
  3. A related word storage means for storing a plurality of words used simultaneously in the search query in association with each other as related words;
    The attribute candidate generation unit extracts, as the attribute candidate word, a word associated with the input basic word and the related word from the related word storage unit,
    The pattern processing means includes
    As the word processing, proper noun removal means for removing proper nouns from the attribute candidate words,
    As the word processing, nounative adjective removal means for removing nounative adjectives from the attribute candidate words,
    Synonym processing means for unifying synonyms for the attribute candidate words as the word processing;
    The advertisement accepting apparatus according to claim 2, wherein the attribute candidate words remaining as a result of the removal and unification are used as the attribute.
  4. The action candidate output means performs a search using the search character string corresponding to the action candidate, and outputs at least one of an additional keyword or an image based on the search result together with the action candidate,
    The action accepting unit accepts designation of at least one of the additional keyword or the image regarding the action candidate,
    The bid accepting unit accepts a bid for an advertisement based on at least one of the additional keyword or image with respect to the action candidate, and adds the information of at least one of the additional keyword or image to the specified action and adds a character string. The advertisement accepting apparatus according to claim 2 or 3, wherein the action character string represented by (2) and the received advertisement are associated with each other and stored in a predetermined advertisement storage unit.
  5.   5. The action candidate output means performs a search corresponding to the action candidates, and excludes action candidates whose search results are less than a predetermined number from display targets. The advertisement reception device described in 1.
  6. An advertisement matching device according to claim 1;
    The advertisement reception device according to any one of claims 2 to 5,
    An advertising system comprising:
  7. The computer realizes an advertisement storage means that stores an advertisement and an action character string that is designated as a display condition for the advertisement and that indicates an action indicating the searcher's search intention in a character string,
    A search query acquisition process in which a computer acquires a search query;
    An attribute candidate generation process in which the computer generates an attribute candidate word that is a word candidate representing an attribute related to the search query based on the acquired search query;
    Pattern processing for extracting the attribute by performing word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms for the generated attribute candidate word;
    Computer, corresponding to the attributes extracted, shows a search intent of the searcher, and action candidate generation processing for generating a candidate actions involving verbs search intent of the searcher,
    Action candidate output means for outputting a candidate for the generated action by the computer;
    An action reception process in which the computer accepts designation of an action among the output action candidates,
    A search process in which the computer executes a search according to the specified action;
    A character string acquisition process in which the computer acquires a search character string based on a set of words corresponding to the action related to the search;
    The computer determines the suitability between the obtained search character string and the action character string corresponding to the advertisement from the advertisements stored in the advertisement storage means, together with the result of the search. An ad selection process for selecting an ad to be output;
    An advertisement matching method comprising:
  8. An advertisement matching program for selecting an advertisement by controlling a computer,
    The computer realizes an advertisement storage means that stores an advertisement and an action character string that is designated as a display condition of the advertisement and that represents an action indicating a searcher's search intention in a character string in association with each other. ,
    Let the computer retrieve the search query,
    Based on the acquired search query, the computer generates an attribute candidate word that is a word candidate representing an attribute related to the search query.
    Let the computer extract the attribute by performing word processing with a predetermined pattern including at least one of removal of a predetermined type of word and unification of synonyms for the generated attribute candidate word,
    The computer, corresponding to the attributes extracted, shows a search intent of the searcher, to produce a candidate actions involving verbs search intent of the searcher,
    Let the computer output the generated action candidates,
    Let the computer accept the designation of the action among the output action candidates,
    Let the computer perform a search according to the specified action,
    Causing a computer to obtain a search character string based on a set of words corresponding to the action related to the search;
    Along with the result of the search, the computer determines the compatibility between the search character string acquired from the advertisement stored in the advertisement storage means and the action character string corresponding to the advertisement. An advertisement matching program that selects advertisements to be output.
JP2010174388A 2010-08-03 2010-08-03 Advertisement matching apparatus, method and program Active JP5480058B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2010174388A JP5480058B2 (en) 2010-08-03 2010-08-03 Advertisement matching apparatus, method and program

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2010174388A JP5480058B2 (en) 2010-08-03 2010-08-03 Advertisement matching apparatus, method and program

Publications (2)

Publication Number Publication Date
JP2012037920A JP2012037920A (en) 2012-02-23
JP5480058B2 true JP5480058B2 (en) 2014-04-23

Family

ID=45849901

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2010174388A Active JP5480058B2 (en) 2010-08-03 2010-08-03 Advertisement matching apparatus, method and program

Country Status (1)

Country Link
JP (1) JP5480058B2 (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101821358B1 (en) 2013-01-22 2018-01-25 네이버 주식회사 Method and system for providing multi-user messenger service
JP6116034B1 (en) * 2016-02-25 2017-04-19 Necプラットフォームズ株式会社 Web site search apparatus, Web site search method, and Web site search program
KR101927962B1 (en) * 2017-12-22 2018-12-12 이왕재 Data transaction method

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4208402B2 (en) * 2000-10-31 2009-01-14 株式会社リコー Document search apparatus, document search method, and recording medium
JP2002230037A (en) * 2001-01-31 2002-08-16 Kddi Corp Retrieval system, retrieval method and program
JP3881638B2 (en) * 2003-07-31 2007-02-14 株式会社東芝 Document search apparatus, document search method, and document search program
US8392249B2 (en) * 2003-12-31 2013-03-05 Google Inc. Suggesting and/or providing targeting criteria for advertisements
JP2006079454A (en) * 2004-09-10 2006-03-23 Fujitsu Ltd Search keyword analysis method, search keyword analysis program and search keyword analysis apparatus
JP4997743B2 (en) * 2005-11-10 2012-08-08 日本電気株式会社 Document search apparatus, document search program, and document search method
JP4962967B2 (en) * 2008-01-11 2012-06-27 ヤフー株式会社 Web page search server and query recommendation method
JP2010049409A (en) * 2008-08-20 2010-03-04 Yahoo Japan Corp Advertisement bidding system, advertisement bidding method, document retrieval system, document retrieval method, program, and control method for client system
KR101042901B1 (en) * 2008-10-09 2011-06-21 엔에이치엔비즈니스플랫폼 주식회사 Method, system and computer-readable recording medium for creating keyword pairs for search advertisement based on keyword purchase history

Also Published As

Publication number Publication date
JP2012037920A (en) 2012-02-23

Similar Documents

Publication Publication Date Title
US7899818B2 (en) Method and system for providing focused search results by excluding categories
US8095582B2 (en) Dynamic search engine results employing user behavior
US7836010B2 (en) Method and system for assessing relevant properties of work contexts for use by information services
US8041601B2 (en) System and method for automatically targeting web-based advertisements
JP5572596B2 (en) Personalize the ordering of place content in search results
US20050028156A1 (en) Automatic method and system for formulating and transforming representations of context used by information services
US20050222989A1 (en) Results based personalization of advertisements in a search engine
US8260664B2 (en) Semantic advertising selection from lateral concepts and topics
CA2732733C (en) Providing posts to discussion threads in response to a search query
JP2008533553A (en) Matching and ranking of sponsored search listings that incorporate web search technology and web content
CN102625936B (en) Query suggestion from document
KR100932999B1 (en) User information and on the basis of the content of a document browsing on the automatically generated links
US8600979B2 (en) Infinite browse
KR20100022467A (en) Cross-language information retrieval
JP5114380B2 (en) Reranking and enhancing the relevance of search results
US7873640B2 (en) Semantic analysis documents to rank terms
US20110225152A1 (en) Constructing a search-result caption
US7092938B2 (en) Universal search management over one or more networks
US20130254189A1 (en) Using Anchor Text to Provide Context
CN103221951B (en) Predictive query suggestion caching
JP2008527509A (en) Systems, methods, software, and interfaces for multilingual information retrieval
CA2635783C (en) Dynamic search box for web browser
KR101335400B1 (en) Identifying comments to show in connection with a document
US8301616B2 (en) Search equalizer
US20130031087A1 (en) Systems and methods for contextual personalized searching

Legal Events

Date Code Title Description
A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20120717

RD02 Notification of acceptance of power of attorney

Free format text: JAPANESE INTERMEDIATE CODE: A7422

Effective date: 20120808

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20120918

A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 20121016

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20130116

A911 Transfer of reconsideration by examiner before appeal (zenchi)

Free format text: JAPANESE INTERMEDIATE CODE: A911

Effective date: 20130123

A912 Removal of reconsideration by examiner before appeal (zenchi)

Free format text: JAPANESE INTERMEDIATE CODE: A912

Effective date: 20130315

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20140213

R150 Certificate of patent or registration of utility model

Ref document number: 5480058

Country of ref document: JP

Free format text: JAPANESE INTERMEDIATE CODE: R150

S531 Written request for registration of change of domicile

Free format text: JAPANESE INTERMEDIATE CODE: R313531

R350 Written notification of registration of transfer

Free format text: JAPANESE INTERMEDIATE CODE: R350

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

S533 Written request for registration of change of name

Free format text: JAPANESE INTERMEDIATE CODE: R313533