US20170039285A1 - Adaptive user interface for real-time search relevance feedback - Google Patents

Adaptive user interface for real-time search relevance feedback Download PDF

Info

Publication number
US20170039285A1
US20170039285A1 US15/207,277 US201615207277A US2017039285A1 US 20170039285 A1 US20170039285 A1 US 20170039285A1 US 201615207277 A US201615207277 A US 201615207277A US 2017039285 A1 US2017039285 A1 US 2017039285A1
Authority
US
United States
Prior art keywords
user
objects
display
search result
search
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.)
Abandoned
Application number
US15/207,277
Inventor
Mark D. Cramer
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Surf Canyon Inc
Original Assignee
Surf Canyon Inc
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
Priority to US11/510,524 priority Critical patent/US20070260597A1/en
Priority to US11/743,076 priority patent/US8442973B2/en
Priority to US13149808P priority
Priority to US12/482,389 priority patent/US8117197B1/en
Priority to US13/355,433 priority patent/US8543570B1/en
Priority to US14/018,791 priority patent/US8924378B2/en
Priority to US14/552,320 priority patent/US9418122B2/en
Application filed by Surf Canyon Inc filed Critical Surf Canyon Inc
Priority to US15/207,277 priority patent/US20170039285A1/en
Publication of US20170039285A1 publication Critical patent/US20170039285A1/en
Abandoned legal-status Critical Current

Links

Images

Classifications

    • G06F17/30867
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2457Query processing with adaptation to user needs
    • G06F16/24578Query processing with adaptation to user needs using ranking
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/248Presentation of query results
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/335Filtering based on additional data, e.g. user or group profiles
    • G06F16/337Profile generation, learning or modification
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/951Indexing; Web crawling techniques
    • G06F17/3053
    • G06F17/30554
    • G06F17/30864
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce, e.g. shopping or e-commerce
    • G06Q30/02Marketing, e.g. market research and analysis, surveying, promotions, advertising, buyer profiling, customer management or rewards; Price estimation or determination
    • G06Q30/0241Advertisement
    • G06Q30/0251Targeted advertisement
    • G06Q30/0255Targeted advertisement based on user history
    • G06Q30/0256User search
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network-specific arrangements or communication protocols supporting networked applications
    • H04L67/22Tracking the activity of the user
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y10TECHNICAL SUBJECTS COVERED BY FORMER USPC
    • Y10STECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y10S707/00Data processing: database and file management or data structures
    • Y10S707/99931Database or file accessing

Abstract

A method and apparatus for dynamically adjusting the user interface of a search engine in order to effectively communicate the improved relevancy achieved through real-time implicit re-ranking of search results is described. Real-time implicit re-ranking occurs without delay after every user action as the search is being conducted, so finding methods of immediately altering the search page without disrupting the user experience is important. Graphical icons next to search results are employed to enable generating and removing re-ranked results, referred to as “recommended” search results. Clusters based on the real-time user model are also displayed to facilitate query reformulations. Sponsored links are selected using the real-time user model along with a combination of RPC and CTR information and are displayed in a manner similar to the organic results, or used to replace the initial sponsored links altogether.

Description

    CROSS-REFERENCES TO RELATED APPLICATIONS
  • This application claims the benefit of U.S. Provisional Patent Application No. 61/131,498 filed Jun. 10, 2008, entitled “Search result relevancy using real-time implicit relevance feedback”, which is hereby incorporated by reference in its entirety. Related applications are U.S. patent application Ser. No. 11/510,524 filed Aug. 25, 2006, entitled “Dynamic Search Engine Results Employing User Behavior” and U.S. patent application Ser. No. 11/743,076 filed May 1, 2007, entitled “Real Time Implicit User Modeling for Personalized Search”, the disclosures of which are hereby incorporated herein by reference.
  • BACKGROUND OF THE INVENTION
  • The present invention relates to search engines which monitor real-time user behavior in order to improve relevance by immediately re-ranking results in response to user actions.
  • Information Retrieval (IR) systems in general, and search engines in particular, are designed to assist users in locating and identifying documents of interest based on a variety of input regarding the user's information need. This need is most normally received by asking the user to input a query, typically expressed as a set of keywords and sometimes including various Boolean operators. The understanding of the user's intent can then be modified or refined based on other user inputs, both explicit and implicit. The IR system will then comb the universe of available documents for possible matches which, based on any number of different relevancy algorithms, are presented to the user in a descending order of probability of interest. In other words, those documents with the highest probability of being relevant to the user are typically presented first. The objective is to quickly point the user toward documents with the greatest likelihood of satisfying the information need.
  • Information retrieval on very large data communication networks that contain an enormous amount of information, such as the internet, is particularly difficult given the sizable sets of potentially relevant documents. It is not uncommon for queries on the general internet to return hundreds of thousands, if not millions, of results. Optimally ranking these results based on inherently ambiguous user input, even when taking additional information regarding the user or other searchers into account, is extremely challenging.
  • IR systems and search engines have thus employed many strategies for combating this issue, most of which seek to achieve a better understanding of the user's information need. Some explicitly request additional information concerning intent, such as directly asking the user to respond to questions regarding both short- and long-terms interests as well as proposing clusters (e.g., a group of key words or categories), drill-downs or reformulations based on the user's query. Others rely on implicit signals derived from user behavior in order to better divine intent.
  • While the objective of all of these strategies is to produce an optimally ranked set of results based on the user's query and the corpus of user, documents and other information available prior to the user submitting the query, one technique of particular interest involves using real-time user behavior signals to disambiguate intent after the result set has been produced and then exploiting that information to immediately re-rank the results following each user action. (This is described in U.S. patent application Ser. No. 11/510,524 entitled “Dynamic search engines results employing user behavior” and Ser. No. 11/743,076 entitled “Real time implicit user modeling for personalized search.”) While the other strategies harness various types of information, both implicit and explicit, to produce a static and frequently sub-optimally ranked set of documents, the results with real-time re-ranking are dynamic as each new user input immediately generates an improved ranking and modified result set.
  • As an example, a user query for “dolphins” is inherently ambiguous as it is unclear whether the individual's intent refers to the mammal or the football team. By observing the past search behavior of the individual, as well as those of other searchers, many IR strategies will attempt to divine which context has the greater likelihood of satisfying this particular user's information need, at the particular point in time, but it's virtually impossible to know with certainty and so the result set will typically contain a mix of both. Once the user actually selects a result from one context or the other, however, the degree to which the user's intent can be inferred is increased dramatically and the result set can be optimized on the fly by immediately re-ranking the documents. For example, should the user select a result related to the Miami Dolphins football team, documents related to football are “promoted” while simultaneously those related to the mammal are “suppressed.”
  • While a static result set is easily depicted as a list of documents in something close to descending order of probabilistic relevance, presenting dynamic results poses challenges for the user interface (UI). As the re-ranking of results continuously produces an optimally superior order of documents, it is important for any distraction created by the movement of documents to not impair the advantage of the heightened relevancy.
  • There is therefore a need to develop a UI capable of delivering the benefits of real-time re-ranking while being as unobtrusive as possible to the user's search experience.
  • SUMMARY OF THE INVENTION
  • The present invention provides a system and methods for dynamically altering the UI of an IR system to present the fluid nature of search results as they are re-ranked in real time.
  • In one embodiment, the adaptive UI is run as a software application between a user and a standard third-party search engine or multiple third-party search engines. This software application can run on the user's computer, a central server, or both. In another embodiment the software for the UI is built into the underlying IR system and may run on the same server as the IR system or a separate server.
  • In one embodiment, every time a new ranking is produced the search results are immediately redisplayed in their new order. In another embodiment, rather than moving around results that have already been displayed, only the unseen results are re-ranked with the top unseen documents, after the re-ranking, being inserted on the search results page as recommended search results. In another embodiment, such recommendations are nested to several levels with each successive user action bringing forward from subsequent results pages more re-ranked documents. Each set of recommendations may be preceded by some explanatory text or some other design device, or displayed on a different area of the page, to set them apart from the documents that were previously on the results page.
  • In one embodiment the initial position of each document is indicated by some text or graphical element. For example, a re-ranked result on page 1 that was initially on page 2 might be followed by the text “from page 2.”
  • In one embodiment each result is accompanied by a graphical element or link enabling the immediate re-ranking of documents and display of recommended search results. Clicking a graphical element will behave similarly as if the user had selected the particular document and then returned to the result set. These graphical elements may also accompany each recommended search result and operate in a similar manner. In another embodiment, clicking the graphical element for a document that already has recommendations will cause the recommendations to be removed from the page, along with any associated nested recommendations. Subsequently clicking the same graphical element will alternately display and remove the set of recommendations and any associated nested recommendations.
  • In one embodiment, clusters or query reformulation suggestions are presented next to each set of recommended documents. Unlike traditional clusters, however, these clusters are refined and selected based on the current real-time user model. Rather than presenting a set of clusters or query reformulation suggestions based on the user's query or the entire document set, the present invention can produce more highly targeted clusters to facilitate the process of moving forward.
  • In one embodiment, the next page of results, prior to being displayed to the user, is completely re-ranked at the moment the user clicks “next” or “more results.” Once these re-ranked results are displayed they become seen and so re-ranking and display of recommendations proceeds as it did on the previous page.
  • Sponsored links may also be re-ranked or updated based on the real-time model of user intent. In one embodiment, the real-time user model is used to select more appropriate advertisements from the set of advertisements associated with the user's query. In another embodiment, the real-time user model is used to modify the user's query in order to find additional advertisements more precisely related to the user's intent.
  • In one embodiment, whenever the real-time user model is updated the old advertisements are removed from the page and replaced with new advertisements. In another embodiment, the advertisements operate similar to the organic results with recommended sponsored links being inserted under selected sponsored links. In another embodiment, sponsored links on subsequent pages of results are selected based on the current real-time user model.
  • The outcome is a dynamic UI that adapts to the real-time user model in order to display the most relevant search results and advertisements in a manner as unobtrusive to the user experience as possible.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • FIG. 1 is a diagram of one embodiment of a system incorporating the present invention.
  • FIG. 2A is a diagram of one embodiment of the home page.
  • FIG. 2B is a diagram of one embodiment of the initial search results.
  • FIG. 2C is a diagram of one embodiment of the re-ranked search results.
  • FIG. 2D is a diagram of one embodiment of re-ranked search results presented as indented “recommendations”
  • FIG. 2E is a diagram of one embodiment of re-ranked search results presented as nested “recommendations”
  • FIG. 3 is a diagram of one embodiment of adaptive UI tools added to the search results page.
  • DETAILED DESCRIPTION OF THE INVENTION
  • The preferred embodiment of the present invention operates on the interne, and more specifically the World Wide Web. The invention, however, is not limited to the internet, the World Wide Web or any other particular network architecture, software or hardware which may be described herein. The invention is appropriate for any other network architectures, hardware and software. Furthermore, while the following description relates to an embodiment utilizing the interne and related protocols, other networks and protocols, for example, for use with interactive TVs, cell phones, personal digital assistants and the like, can be used as well.
  • The functions described herein are performed by programs including executable code or instructions running on one or more general-purpose computers. The functions described herein, however, can also be implemented using special purpose computers, state machines and/or hardwired electronic circuits. The example processes described herein do not necessarily have to be performed in the described sequence and not all states have to be reached or performed.
  • As used herein, the term “website” refers to a collection of content. Website content is often transmitted to users via one or more servers that implement basic intern& standards. “Website” is not intended to imply a single geographic or physical location but also includes multiple geographically distributed servers that are interconnected via one or more communications systems.
  • As used herein, the term “document” is defined broadly and includes any type of content, data or information contained in computer files and websites. Content stored by servers and/or transmitted via the communications networks and systems described herein may be stored as a single document, a collection of documents or even a portion of a document. The term “document” is not limited to computer files containing text but also includes files containing graphics, audio, video and other multimedia data. Documents and/or portions of documents may be stored on one or more servers.
  • As used herein, the term “click” or “click-through” is defined broadly and refers to clicking on, or hovering over, a hyperlink included within search result listings to view an underlying document or website. The term “clicking on” a link or button, or pressing a key to provide a command or make a selection, may also refer to using other input techniques such as voice input, pen input, mousing over or hovering over an input area or the like.
  • The dynamic UI of the present invention is designed to efficiently communicate the benefits of enhanced relevancy obtained through immediately re-ranking of search results in response to user actions. In order to appreciate the benefits of real-time re-ranking, the display of search results should be updated each time the real-time user model produces a new ranking of search results. In the preferred embodiment, the display of search results is immediately updated every time a re-ranking occurs.
  • FIG. 1 is a high-level illustration of a preferred embodiment of a system incorporating the present invention. Documents on servers or computers 110-118, on a data communication network 140, such as the interne, are scanned and indexed by search engines on computers and associated databases 120-128. On the interne there are currently billions of documents available, with more being added daily. Google, Yahoo! and Bing are examples of popular intern& search engines, but most any web site that enables users to search for documents on a data communications network and retrieve relevant results can be considered a search engine. Users on computers or terminals 100-108 seeking information can connect with the search engine of their choice, enter keywords related to their searches, and then retrieve lists of matched documents, normally sorted and ranked by an algorithm that places the ones with the highest probability of being relevant to the user's query at the top. The real-time re-ranking engine provides a server 132 and database 130 which plug into the data communications network and then uses the behavior of current users to alter the results provided by one, if not several, of the other search engines. An embodiment of the present invention resides on server 132 with the re-ranking engine.
  • The description of this system will focus on a website that takes results from other search engines that reside on the internet, however, other embodiments of the system would involve incorporating the re-ranking engine and present invention directly into one of the other search engines 120-128. Rather than collecting the initial search results via a data communications network, the system can gather the results directly from a search engine.
  • One embodiment of the invention would involve software, which could be made available for download, which resides on users' computers or terminals 100-108. Rather than going to the website of the invention, users will now go directly to a search engine of choice and the process of using user modeling to immediately alter search results will be performed by the software located on the users' computers. The adaptive UI can then be applied on top of the user's search engine within the user's computer.
  • FIG. 2A is a diagram of a home page of a search engine. While other information and links can be present, the principal purpose of the home page is to enable the user to enter a query consisting of a keyword or set of keywords 202 representing the user's information need before clicking on a button 204 to request that the search engine retrieve the results.
  • FIG. 2B is a diagram of an example of a search results page depicting what the initial search results for the keyword “Washington” might look like. Upon receiving a query 212 from a user, expressed as a keyword or set of keywords, the search engine produces a set of results 220-227. Each result is normally comprised of a title, snippet and the Uniform Resource Locater (URL) of the document, the combination of which can be described as the summary. Search engines may also include additional information next to each result, such as images or expanded snippets. Since the number of matched documents rarely fits on a single page, buttons, such as a next results button 216, are typically available to enable users to navigate to subsequent pages of results and then back to previous pages of results. It should also be noted that a search engine may offer supplemental information on the search results page, as is often the case, such as related popular queries, suggested spellings or links to maps and stock quotes.
  • FIG. 2C is a diagram of an example of how the first modified results might look after the user clicks on the document “George Washington University” 226 from the initial results as depicted in FIG. 2B and then, after viewing the document, returns to the search results page. The set of matched documents remains the same, yet their order has changed based on the intelligence gathered as a consequence of the user's click on the document “George Washington University” 226. The re-ranking engine has deduced that the user is most likely interested in universities containing the word “Washington” as opposed to George Washington the man or Washington the city or state. As a result, the “Access Washington Home Page” document 220, which was initially at the top of the list, has fallen to the sixth position 245 because the system has deemed, based on the behavior of the user, that the probability of this document being relevant to the user has declined. On the other hand, the “University of Washington” document 221, which was initially in the second position, has now moved to the top 240 because the system has deemed, again based on the behavior of the user, that the relevance of this document has increased. This process continues, for all practical purposes, for every document returned in the set of search results.
  • FIG. 2D represents another embodiment of the invention where instead of immediately altering the positions of all of the documents, the documents that have already been seen 260-264 remain “static,” while re-ranked results from subsequent pages are pulled forward 270-272 and indented as “recommendations.” In this example, upon the user selecting the document “George Washington University” 263, the re-ranking engine immediately adjusts the position of all the documents in the result set that have not yet been seen by the user and then the adaptive UI pulls forward three to present to the user as recommendations. Should the user return to the search result page by clicking “back” on the web browser, or select the document without leaving the search result page by doing something such as a Ctrl-click on the result link, the recommended documents will immediately be presented.
  • In this example, three documents are pulled forward as recommendations, however, any number of documents may be selected, including zero. In another embodiment of the invention, the number of documents pulled forward as recommendations is a function of the quality of results that are already on the page, as determined by the re-ranking engine. For example, if three documents are selected as candidates for recommendations yet the re-ranking engine determines that one of those is inferior to a result that is already displayed on the page, the adaptive UI will only bring forward the other two as recommendations.
  • Table I is an illustration of how the number of recommended results for a given user action can be made to vary. Assume that the system is configured to bring forward a maximum number of three recommended results after the selection of a document. Documents A through T are ranked from 1 to 20, with the first ten results (documents A through J) being on the first page of search results. At the moment the user selects Document E, which was initially ranked #5, the re-ranking engine calculates the new ranks. Since documents A through J have already been seen by the user, at the moment the user returns to the search results page these documents will remain present in the same order, however, the adaptive UI will bring forward up to three recommended search results from subsequent pages which, in this example, is only the second page. The new ranking indicates that documents M and Q are relevant to the user based on the real-time user model, so they are brought to the first page of results as recommendations. Document T, however, is not since there are still documents on the first page of results (documents H and I) that are more relevant, based on the new ranking, than document T.
  • TABLE I Example of using recommended result relative quality in the determination of the number of recommended search results to display Initial New Rank Rank Document Notes 1 13 Document A Skipped 2 16 Document B Skipped 3 10 Document C Skipped 4 7 Document D Skipped 5 1 Document E Selected document 6 8 Document F 7 18 Document G 8 4 Document H 9 5 Document I 10 20 Document J 11 12 Document K 12 11 Document L 13 2 Document M 1st recommended result 14 14 Document N 15 15 Document O 16 9 Document P 17 3 Document Q 2nd recommended result 18 19 Document R 19 17 Document S 20 6 Document T Not brought forward
  • In some situations a result may be brought forward as a recommended result even if there is a more highly ranked document already on the page. For example, using Table I, should document H have had a new rank of #3 and document Q have had a new rank of #4, the adaptive UI could have still brought forward document Q because, subsequent to the re-ranking, the three most relevant results, not including the selected document, would have been, respectively, documents M, H (already on the first page) and Q. In another embodiment of the invention, the skipped results are not considered when determining how many results to bring forward as, in spite of the fact they might have a high relevance as determined by the real-time user model, they have already been viewed and passed over by the user.
  • In one embodiment of the invention the top re-ranked documents are pulled forward to the current search results page as recommendations, however, in another embodiment the adaptive UI modifies this slightly by selecting a diverse set of re-ranked documents in order to maximize the likelihood of presenting something relevant to the user. Should the re-ranking engine determine that there are multiple possible sub-contexts for the user's intent, represented as subordinate keywords with significant positive weights or multiple clusters associated with the selected documents, the invention may pull forward the top re-ranked document from each of these sub-contexts (the determination, weighting and use of subordinate keywords is described in detail in the Related Applications described and incorporated by reference above). For example, in FIG. 2D at the moment the user selects the result “George Washington University” 263, the real-time user model could produce three recommendations 270-272 using the highly-weighted subordinate keywords “George” and “University” while attempting to have each recommendation correspond to different sub-context, such as “law school,” “medical center” and “athletics.” Even if the top three re-ranked results were all related to the George Washington University Law School, presented results from other sub-contexts enhances the chances that at least one of the recommended search results will correspond to the user's information need.
  • The benefit of maintaining the results that have already been viewed by the user, in the original order, and then bringing forward re-ranked documents as recommended search results, is that the experience is less jarring to the user. Removing results that have already been seen, even if they haven't been selected, or changing the order of results that have already been displayed on the search results page can be confusing and potentially disconcerting to the user. Before selecting a document, users will often retain a visual image of the search results page, which they expect to remain somewhat consistent upon returning to the search results page. By maintaining the relative positions of the documents that have already been seen while inserting re-ranked documents, users can benefit from the improved relevance without having their search experience excessively disrupted.
  • Subsequent pages of results are calculated using the real-time user model at the moment the user selects “Next” or “More results.” All of the documents that have not yet been displayed to the user, either as the original set of results or recommended search results, will be re-ranked before the adaptive UI presents the top re-ranked documents as the next page of results. Once this new page is displayed, however, those documents are considered “seen” and the re-ranking process, with new documents brought forward as recommendations following each user action, proceeds as described above. In another embodiment, much as recommendations may be selected to present a diversity of sub-contexts to the user, as described above, re-ranked documents for subsequent pages of results may also be chosen to present a diversity of sub-contexts to the user.
  • FIG. 2E represents one embodiment of the invention where re-ranked documents, presented as recommended search results, are nested. In this example, after selecting the document “George Washington University” 281 and being presented with three recommendations 285-287, the user proceeds by selecting the document “George Washington University—Official Athletic Site” 287 and is immediately presented with two additional recommendations 290-291. These two new recommendations are based on the real-time user model of inferred intent built from the knowledge that the user has selected “George Washington University” and “George Washington University—Official Athletic Site” as well as skipped over a number of other documents as well as some recommended search results. In this particular example it may be inferred that, while the user's query was for “Washington,” the intent is more specific to athletic programs at the University of Washington. As a consequence, the two nested recommended search results pertain to such. In one embodiment, the number of recommended result objects in each nest level is enough to provide the user a selection, but not so many as to affect the perception that the user is looking at the same page. In one embodiment, between 3 and 5 objects are provided in each nest level.
  • Should the user then move forward by selecting the document “George Washington University Law School” 285, the real-time user model will be employed to determine a new ranking for all of the documents and the adaptive UI will insert any new recommended results directly under the document the was just selected 285. The real-time user model will take into account all of the user's actions during the current information session and, as such, the recommended results produced will be based on all the previously selected and skipped documents.
  • In one embodiment of the invention, a decay is used to adjust the weights of positive and negative subordinate keywords as the search is progressing. After each user action, just prior to updating the real-time user model, all of the subordinate keywords may be multiplied by some scaling factor, such as 90%. This will have the benefit of providing slightly more weight to the most recent user actions, which, during relatively long information sessions, will have the benefit of producing new rankings that compensate for any slight changes or drifts in user intent. Long information sessions often involve a learning process on the part of the user which could have an impact on the real-time model of inferred intent.
  • In one embodiment of the invention, the adaptive UI offers a control, in the form of a link or a button, to enable the user to clear or reset the real-time user model and recommence with the query. This could be particularly useful should the user experience and recognize a significant shift in information need while the search is taking place. Since the real-time user model incorporates all of the user's previous actions in the information session, a significant shift in intent could cause the re-ranking engine to produce results that are potentially no longer relevant. Even if the embodiment includes a “decay” as described above, a control somewhere on the search results page, such as a link called “reset recommendations” or “clear user model,” could be beneficial to the user in this circumstance.
  • FIG. 3 depicts how the invention may add an element to each document in order to provide greater control to the user. In one embodiment, small links, icons or other graphical elements, depicted here as a small image of a bull's eye, are associated with each search result. Selecting one of these icons, by virtue of a click for example, will immediately invoke the re-ranking engine in order to produce recommendations as if the user had selected the associated result. While the re-ranking engine will not be able to associate a dwell time with the selection of an icon, because the document is not actually selected to be opened, this action will enable the user to express a certain interest in a particular result in order to be immediately presented with appropriate recommendations based on that interest.
  • The invention may be configured so that recommendations are displayed when the user selects a result, when the user selects an icon associated with a result, or in just one instance but not the other. The invention may also be configured so that recommendations are never displayed and, as such, re-ranked documents only appear on subsequent pages of results. In another embodiment of the invention, selecting the icon will always produce at least one recommended result, regardless of the quality of that recommendation relative to the other documents that are already on the page. Having the icon “force” at least one recommendation will be a positive indication to the user that the adaptive UI is currently functional.
  • In one embodiment of the invention, if there are already recommendations associated with a particular result, a click on the icon will remove those recommendations and any nested recommendations. Subsequent clicks on the icon will alternately open and close any recommendations and nested recommendations. This functionality will enable users to easily remove unwanted recommendations from the result page and then bring them back should they so desire.
  • In one embodiment of the invention, with respect to updating the real-time user to model to calculate new ranks for the documents, the re-ranking engine considers the action of selecting an icon as separate from actually selecting the document to be opened. The amount of weight applied to the modifications that are subsequently made to the real-time user model of inferred intent can thus be adjusted. For example, clicking an icon could be considered a weaker indicator of intent when compared to actually selecting the document to be opened and, as such, modifications to the subordinate keyword weights will be scaled down appropriately. Should the user subsequently proceed to click the result in order to view the document, modifications to the subordinate keywords weights may be scaled back up to full strength.
  • Selecting the icon to remove recommendations can be considered an indication of negative intent and such an action is similarly incorporated into the real-time user model. In one embodiment, recommendations that are removed from the page by virtue of selecting an icon are treated as if they had been “skipped” for purposes of the re-ranking engine. In another embodiment, removing recommendations is a stronger indication of intent than simply skipping a result and, as such, any modifications to the subordinate keyword weights are given additional negative influence.
  • In one embodiment of the invention, movement indicators are also present next to each recommendation, giving the user an indication of the page or location from where the recommendations were pulled. In the example, the document “George Washington University Law School” 306 was pulled forward “from page 2.” This is an indication that the document, now presented as a recommended search result, was initially ranked on page 2 at the time the query was submitted. In another embodiment this movement indicator could be the initial rank of the document, such as “#25”. In another embodiment, all of the results on subsequent search results pages also include movement indicators, as they too have had their positions altered from the moment the query was initially submitted.
  • Clustering and query suggestions are a well-understood practice for search engines and many offer both to their users. There are many ways to produce clusters or query suggestions, but typically they revolve around analyzing the various sub-contexts in the result set to offer possible refinements to the user. For example, looking at the result set for the query “Washington,” a search engine might propose the following clusters: University of Washington, Washington State, Washington Capitals, Washington Politics, Washington Government, Washington D.C., etc. It's also possible for clusters to be nested within each other. Some search engines also produce query suggestions by looking at refinements produced by previous users who've entered the same query. For example, should the search engine see that many users who query for “Washington” then proceed to query for “University of Washington,” the latter could be offered as a query suggestion to users searching for the former. The location on the search results page where the clusters and query suggestions are presented can vary, although they're typically either at the top or bottom, or off to the side in a vertical side bar.
  • One embodiment of the present invention exploits the real-time user model of inferred intent to produce a list of more highly targeted clusters or query suggestions. For example, at the moment the user selects “George Washington University” 303 to produce recommendations 306-308, the real-time user model has already inferred that the intent is related to “university.” As such, the adaptive UI can suggest real-time clusters 315, such as “university” and “school,” while simultaneously eliminating clusters that are not related to the user's information need, such as “monument” or “president.” In one embodiment, selecting one of these cluster links will rerun the user's query using the cluster information. For example, clicking “university” will run a new query for “Washington University.” In another embodiment, selecting one of these links will simply take the subset of documents from the original result set that belong to that particular cluster. It is a not a requirement for the real-time clusters to appear above the recommendations. In other embodiments of the invention they may appear anywhere on the search results page. In the case of a query reformulation, where the real-time user model is carried over from a previous query, it would make sense to put the real-time clusters either at the top of bottom of the search results page.
  • In another embodiment of the invention, clusters are calculated using the set of produced documents at the moment the user submits the query. There exists many techniques for calculating such clusters, but each document should consequently belong to one or more clusters. At the moment the user selects a document, the real-time clusters may be a function of real-time user model as well as the clusters to which the document originally belonged. Additional document selections will further enable the adaptive UI to determine the clusters that are most appropriate to the user by observing the frequency with which selected documents appear in each of the calculated clusters. In another embodiment of the invention, the set of available clusters is recalculated using a subset of the documents in the result set. This subset is selected based on the real-time user model. For example, after each user action the set of available clusters could be recalculated using the top 50 re-ranked documents. This technique will enable the clusters to follow the evolution of the real-time user model.
  • Real-time clusters can be presented to the user every time there is an update to the real-time user model, which typically occurs when the user selects a document or moves to a subsequent page of results. The real-time clusters take advantage of the cumulative nature of the real-time user model. For example, after selecting “George Washington University” 303, should the user continue by selecting “George Washington University—Official Athletic Site” 308, the adaptive UI can present clusters 316 based on both of these actions, such as “university athletics” and “athletics store”. Real-time cluster and query suggestion links may also be shown immediately upon displaying a subsequent page of results.
  • The practical advantage of real-time clusters or query suggestions is that they are targeted to the user's real-time intent. Traditionally clusters and query suggestions have been unable to exploit the information present in a real-time user model of inferred intent and, as such, rely on providing a wide range of clusters related to the entire document set, or alternatively the reformulations from users who previously submitted a similar query. This inevitably requires the user to wade through a considerable number of options, which requires time and energy. Conversely, real-time clusters, by virtue of being based off the real-time user model, are more targeted to the user's intent and are typically fewer in number.
  • The adaptive UI of the present invention can similarly be used to deliver more highly-targeted advertisements. The selection of advertisements can, however, be somewhat different from the re-ranking of organic results. Search engines will typically display advertisements based on a combination of RPC (revenue per click, which is the amount of money the advertiser is willing to pay for a click on a given sponsored link) and CTR (click-through rate, or the rate at which users selected a given sponsored link). The RPC and CTR together are used to maximize the amount of revenue generated for a given user's query by producing a ranked list of available sponsored links.
  • In one embodiment of the invention, the real-time user model is used to re-rank sponsored links in the same manner that is used to re-rank organic results. The subordinate keywords are matched with the titles, snippets and URLs of the available sponsored links for a given query, and the order of the sponsored links is changed.
  • To illustrate how this might work, Table II is a representation of sponsored links returned for a user's query for “digital camera.” The RPC represents the amount of money that the advertiser is willing to pay for a click while the CTR represents the rate at which users click the advertisement when it is displayed on the search page. The CTR will most certainly be a function of where the advertisement is displayed in relation to the others, but for purposes of this illustration consider the CTR to be normalized.
  • TABLE II Example of sponsored advertisements for the user's query “digital camera” ranked by the search engine as a function of, among other things, RPC and CTR Rank RPC CTR Sponsored Advertisement 1 $0.50 2.0% Digital Camera Find Brand Name Digital Cameras at Prices you'll Love. Shop Kmart Now! www.Kmart.com 2 $0.50 1.9% Best Buy ® Official Site 5%-16% Off Select Digital Cameras This Week At Best Buy ®. Shop Now! www.BestBuy.com 3 $0.55 1.7% Sony ® Digital Cameras Official Site: Sony Cyber-shot ® Digital Cameras starting at $129.99 www.SonyStyle.com/cybershot 4 $0.55 1.6% New Digital Camera-TL100 Slim Design with Smart Auto. 12.2 Megapixels, 2.7″ LCD & More www.samsung.com/us/tl100 5 $0.45 2.0% Digital Camera Browse New Digital Cameras at Dell. Compare Brands and Megapixels Here! www.Dell.com 6 $0.40 2.1% Digital Cameras on Sale Free Shipping & No Tax. Authorized Dealer. 24/7 Customer Support! RitzCamera.com 7 $0.35 1.8% Limited Time Camera Deals Save $100 Instantly on Digital Cameras Only at OfficeMax-Buy Now! www.OfficeMax.com 8 $0.35 1.6% Digital Cameras $9.95 Sale-Bargain-All Brands Quantity Discounts-Free Ship www.EcaMeraFilms.com 9 $0.30 1.8% Digital Camera Compare Price and Features. Large Selection In Stock Today. www.Staples.com 10 $0.25 1.7% Digital Camera Bargains Save on Every SLR, Standard Point & Shoot and Compact Digital Camera. www.NexTag.com/Digital-Cameras
  • As can be seen, the advertisements are produced in a certain order which can be a function of many different things, but typically will rely heavily on RPC and CTR. This is because the objective of the search engine is to put the advertisements in an order that will generate the greatest amount of revenue, which is a function of how often an advertisement is clicked multiplied by the amount of revenue generated by each click. That is why advertisements with higher RPCs will sometimes be ranked beneath those with lower RPCs. For example, in Table II the advertisement for “Sony® Digital Cameras” is ranked below the advertisement for “Best Buy® Official Site” even though it has a higher RPC. This is because the CTR for the “Sony® Digital Cameras” is lower, meaning that in aggregate it'll generate less revenue for that query.
  • How these advertisements are displayed is then up to the search engine. In many cases they'll be displayed in a column, in descending order, to the right-hand side of the organic results. Typically only a maximum number of advertisements will be displayed, depending, among other things, on the user's query. In some instances the search engine will decide to take a number of the top advertisements and display them either above or below the organic results. Additionally, there are search engines that will mix the advertisements in with the organic results. In most cases, however, the search engine will indicate which results are sponsored by either putting them in a colored box or indication as such with text.
  • Should a real-time model of the user's intent be available, which is the case when a user has already taken an action during the current information session, then the adaptive UI can exploit this to produce a superior ranking of advertisements. The real-time model of user intent can come from clicks on either organic or sponsored links during the current query or previous queries in the same information session.
  • Continuing with the current illustration, assume that the real-time user model contains the following subordinate keywords with significant positive weights: sale, save, discount and bargain. Assume as well that the real-time user model also contains the following subordinate keywords with significant negative weights: sony and new. As described above, these subordinate keywords may have been produced by virtue of the user selecting documents or advertisements related to “digital camera sales” while skipping results related to “new digital cameras by Sony.”
  • Table III represents how the order of the ads might be changed based on the real-time user model. The ranks of advertisements related to the positive subordinate keywords are improved while the ranks of those ads related to the negative subordinate keywords are reduced. The objective here is still to maximize revenue for the search engine, however, while RPC and CTR information can be taken into account, advertisements that might otherwise have be ranked low are brought forward because they are more closely aligned with the user's real-time information need as indicated by the real-time user model of inferred intent.
  • TABLE III Example of sponsored advertisements for the user's query “digital camera” re-ranked by the real-time user model - for ease of understanding, positive subordinate keywords are represented in bold while negative subordinate keywords are in italic New Rank Old Rank RPC CTR Sponsored Advertisement 1 6 $0.40 2.1% Digital Cameras on Sale Free Shipping & No Tax. Authorized Dealer. 24/7 Customer Support! RitzCamera.com 2 8 $0.35 1.6% Digital Cameras $9.95 Sale-Bargain - All Brands Quantity Discounts- Free Ship www.EcaMeraFilms.com 3 7 $0.35 1.8% Limited Time Camera Deals Save $100 Instantly on Digital Cameras Only at OfficeMax-Buy Now! www.OfficeMax.com 4 1 $0.50 2.0% Digital Camera Find Brand Name Digital Cameras at Prices you'll Love. Shop Kmart Now! www.Kmart.com 5 10 $0.25 1.7% Digital Camera Bargains Save on Every SLR, Standard Point & Shoot and Compact Digital Camera. www.NexTag.com/Digital-Cameras 6 2 $0.50 1.9% Best Buy ® Official Site 5%-16% Off Select Digital Cameras This Week At Best Buy ®. Shop Now! www.BestBuy.com 7 5 $0.45 2.0% Digital Camera Browse New Digital Cameras at Dell. Compare Brands and Megapixels Here! www.Dell.com 8 3 $0.55 1.7% Sony ® Digital Cameras Official Site: Sony Cyber-shot ® Digital Cameras starting at $129.99 www.SonyStyle.com/cybershot 9 4 $0.55 1.6% New Digital Camera-TL100 Slim Design with Smart Auto. 12.2 Megapixels, 2.7″ LCD & More www.samsung.com/us/tl100 10 9 $0.30 1.8% Digital Camera Compare Price and Features. Large Selection In Stock Today. www.Staples.com
  • In another embodiment of the invention, highly weighted subordinate keywords or real-time clusters from the real-time user model are used to enhance the user's query in order to identify additional potentially relevant sponsored links. This is particularly useful when the set of sponsored advertisements for a given query is small, or when the user's real-time intent, as described by the real-time user model, varies considerably from the original query.
  • To illustrate this point, consider a real-time user model where the term “bargain” has a significantly high positive weight, or where “bargain” represents a real-time cluster of significant weight. Table IV is then an example of sponsored advertisements associated with the user query “digital camera bargain.” These sponsored advertisements are ranked by the search engine in a manner similar to the sponsored advertisements in Table II.
  • TABLE IV Example of sponsored advertisements for the user's query “digital camera bargain” ranked by the search engine as a function of, among other things, RPC and CTR Rank RPC CTR Sponsored Advertisement 1 $0.75 2.2% Digital Camera Bargain We Have 800+ Digital Cameras. Digital Camera Bargain on Sale! www.NexTag.com/Digital-Cameras 2 $0.50 1.8% Affordable Cameras Find A Wide Selection Of Affordable Cameras At Best Buy ®. Shop Now! www.bestbuy.com 3 $0.45 1.7% Digital Cameras Cheap Browse for Point & Shoot Digital Cameras from our Secured Merchants. Become. com 4 $0.30 1.6% Save on Digital Cameras Low Low Prices on Brand Names Same Day Shipping, Buy Now! www.TigerDirect.com 5 $0.25 2.0% Digital Camera Store Canon, Nikon, Sony, Olympus & More! Authorized Dealer-Paypal Accepted www.JR.com/Digital_Cameras
  • The adaptive UI of the present invention may subsequently enhance the list of sponsored advertisements from Table II by methodically inserting those from Table IV. The challenge is that while there is a ranking of advertisements for the user's query “digital camera” and another ranking of advertisements for the user's query “digital camera bargain,” there is no combined ranking of the two sets of advertisements.
  • In one embodiment of the invention, the two lists are combined as a function of RPC and CTR and then re-ranked using the real-time user model. For the example, first the advertisements of Table IV are added to the advertisements of Table II similar to if all of the advertisements had been related to the initial user query. Then the real-time user model is used to re-rank the advertisements in the same manner as Table III. Table V is a representation of the re-ranked ads from Tables II and IV. In can be noted that in this particular example that the top advertisement, “Digital Camera Bargain,” wasn't even present in the set of advertisements for the user's original query.
  • TABLE V Example of sponsored advertisements for the user's query “digital camera” re-ranked and then enhanced by sponsored advertisements from a user's query for “digital camera bargain” New Rank Old Rank RPC CTR Sponsored Advertisement 1 Table $0.75 2.2% Digital Camera Bargain IV-1 We Have 800+ Digital Cameras. Digital Camera Bargain on Sale! www.NexTag.com/Digital-Cameras 2 Table II-6 $0.40 2.1% Digital Cameras on Sale Free Shipping & No Tax. Authorized Dealer. 24/7 Customer Support! RitzCamera.com 3 Table $0.50 1.8% Affordable Cameras IV-2 Find A Wide Selection Of Affordable Cameras At Best Buy ®. Shop Now! www.bestbuy.com 4 Table $0.30 1.6% Save on Digital Cameras IV-4 Low Low Prices on Brand Names Same Day Shipping, Buy Now! www.TigerDirect.com 5 Table II-8 $0.35 1.6% Digital Cameras $9.95 Sale-Bargain-All Brands Quantity Discounts-Free Ship www.EcaMeraFilms.com 6 Table 11-7 $0.35 1.8% Limited Time Camera Deals Save $100 Instantly on Digital Cameras Only at OfficeMax-Buy Now! www.OfficeMax.com 7 Table $0.45 1.7% Digital Cameras Cheap IV-3 Browse for Point & Shoot Digital Cameras from our Secured Merchants. Become.com 8 Table II-1 $0.50 2.0% Digital Camera Find Brand Name Digital Cameras at Prices you'll Love. Shop Kmart Now! www.Kmart.com 9 Table $0.25 1.7% Digital Camera Bargains II-10 Save on Every SLR, Standard Point & Shoot and Compact Digital Camera. www.NexTag.com/Digital-Cameras 10 Table II-2 $0.50 1.9% Best Buy ® Official Site 5%-16% Off Select Digital Cameras This Week At Best Buy ®. Shop Now! www.BestBuy.com
  • In the most commonly used search engines, the sponsored links are separate from the organic results. They are typically displayed in a column to the right of the organic results. However, the most attractive ads can be placed either above or below the organic results. In one embodiment of the invention, as soon as a new ranking of advertisements is available, the old set is removed and simply replaced with the new set. While simple to implement from a UI perspective, the disadvantage of this method is that the old advertisement are no longer available to the user, which could cause frustration.
  • In another embodiment, the adaptive UI makes the advertisements operate in a similar fashion to the organic results. When either an organic or sponsored result is selected, recommended sponsored links are immediately presented on the search results page, however, the old advertisement are not removed. The recommended sponsored links may be presented either above or below the original sponsored links, or inserted into the list of sponsored links in a fashion similar to the insertion of recommended organic results into the list of organic search documents.
  • When a subsequent page of results is produced, however, there is an opportunity to display an entirely new set of advertisements based on the real-time user model. In one embodiment of the invention, at the moment the user selects “next” or “more results,” the real-time user model is updated, the sponsored links are immediately re-ranked and selected as described above, and then the top recommended sponsored links are displayed on the new page of results in a fashion similar to the display of the original sponsored links on the first page of search results.
  • In another embodiment of the invention, during a query reformulation that is part of the same information session, the adaptive UI uses the real-time user model to produce a re-ranked set of sponsored links at the moment the new query is submitted. For example, should the user's first query be “digital camera” and the reformulation be “digital camera slr,” then once it is determined that both of these queries belong to the same information session, the real-time user model is carried over to the new query and can used to select new advertisements. Should “canon” have been a highly weighted subordinate keyword, re-ranked advertisements for “canon digital slr cameras” might be selected to appear on the search results page.
  • In search engines where the sponsored results are mixed in with the organic results, the adaptive UI may simply re-rank the combined result set, as described above, as if all the results were organic. In this scenario, in order to maximize the revenue generated from the query, preference may be given to the sponsored results during the re-ranking process.
  • The outcome is an adaptive UI for a search engine that harnesses the power of a real-time user model to continuously change, update and reorganize search results and sponsored links based upon the “at the moment” intent of the user as deduced from the ongoing behavior of the user during the search process.
  • The user's browser may be installed on other devices than a computer, such as a personal digital assistant (PDA), a mobile phone, TV, or any other device. The display can be modified to fit a smaller form factor.
  • As will be understood by those of skill in the art, the present invention could be embodied in other specific forms without departing from the essential characteristics thereof The present invention can use a separate third party search engine, or could be integrated with a search engine. The search engine could be a general search engine that searches the internet, a specialized search engine that searches a particular web site, a database search engine, a meta-search engine that combines the results of multiple other search engines, or any other type of search engine. Accordingly the foregoing description is intended to be illustrative, but not limiting, of the scope of the invention which is set forth in the following claims.

Claims (10)

1. A computer implemented method for dynamically modifying search results comprising:
receiving user electronic input with one or more keywords for use as search terms from a user;
accessing a database with the search terms;
providing a first set of search result objects to a database;
displaying a first portion of the set of search result objects to the user on a device display;
monitoring user computer inputs to determine which object said user clicks on in interacting with said first set of search result objects;
compiling, with the computer, information to infer user intent based on the content of said object the user clicks on;
before display of objects beyond said first portion of the search result objects, re-ranking a second portion of the set of objects based on said user inferred intent, the second portion being objects not previously displayed to the user on the device display;
wherein the first portion of the set of search result objects previously displayed to the user on the device display is not displayed as re-ranked on the device display;
compiling, with the computer, said re-ranked second portion of the set of search result objects;
displaying to said user on the device display the second portion of re-ranked objects; and
using terms from a title and display text corresponding to a selected object as subordinate keywords for generating said re-ranking of said second portion of said set of search result objects.
2. The method of claim 1 wherein said re-ranking moves forward a number of documents as a function of the quality of the documents being moved forward compared to documents already displayed to said user and wherein the function of quality comprises a determination of which documents are more relevant to said user's inferred intent than documents already displayed.
3-19. (canceled)
20. The method of claim 1 wherein said user requesting a display of objects beyond said first portion of the search result objects comprises the user selecting “next” or “more results.”
21. The method of claim 1 wherein said second portion of the set of search result objects are a subsequent page of search result objects.
22. The method of claim 1 further comprising:
cumulatively using analysis of user intent to re-rank said set of search result objects based on subsequent documents said user clicks on and skips.
23. A computer implemented method for dynamically modifying search results comprising:
receiving user electronic input with one or more keywords for use as search terms from a user;
accessing a database with the search terms;
providing a first set of search result objects to a database;
displaying a first portion of the set of search result objects to the user on a device display;
monitoring user computer inputs to determine which object said user clicks on in interacting with said first set of search result objects;
compiling, with the computer, information to infer user intent based on the content of said object the user clicks on;
before display of objects beyond said first portion of the search result objects, re-ranking a second portion of the set of objects based on said user inferred intent, the second portion being objects not previously displayed to the user on the device display;
wherein the first portion of the set of search result objects previously displayed to the user on the device display is not displayed as re-ranked on the device display;
compiling, with the computer, said re-ranked second portion of the set of search result objects;
displaying to said user on the device display the second portion of re-ranked objects.
24. The method of claim 23 further comprising:
using terms from a title and display text corresponding to a selected object as subordinate keywords for generating said re-ranking of said second portion of said set of search result objects.
25. A non-transitory computer readable storage media having computer readable code thereon for dynamically modifying search results, comprising code for:
receiving user computer input for use as search terms from a user;
accessing a database with the search terms;
providing a set of search result objects to a database;
displaying a first portion of the set of search result objects to the user on a device display;
monitoring user computer inputs to determine whether any of said objects is selected by said user;
compiling, with a computer, information to infer user intent based on analyzing the content of at least one object the user selects or skips;
before display of objects beyond said first portion of the search result objects, re-ranking a second portion of the set of objects based on said user inferred intent, the second portion being objects not previously displayed to the user on the device display;
wherein the first portion of the set of search result objects previously displayed to the user on a device display is not displayed as re-ranked on the device display;
compiling, with the computer, said re-ranked second portion of the set of search result objects; and
displaying to said user on the device display the second portion of re-ranked objects.
26. The non-transitory computer readable storage media of claim 20 wherein said user requesting a display of objects beyond said first portion of the search result objects comprises the user selecting “next” or “more results.”
US15/207,277 2006-05-02 2016-07-11 Adaptive user interface for real-time search relevance feedback Abandoned US20170039285A1 (en)

Priority Applications (8)

Application Number Priority Date Filing Date Title
US11/510,524 US20070260597A1 (en) 2006-05-02 2006-08-25 Dynamic search engine results employing user behavior
US11/743,076 US8442973B2 (en) 2006-05-02 2007-05-01 Real time implicit user modeling for personalized search
US13149808P true 2008-06-10 2008-06-10
US12/482,389 US8117197B1 (en) 2008-06-10 2009-06-10 Adaptive user interface for real-time search relevance feedback
US13/355,433 US8543570B1 (en) 2008-06-10 2012-01-20 Adaptive user interface for real-time search relevance feedback
US14/018,791 US8924378B2 (en) 2006-08-25 2013-09-05 Adaptive user interface for real-time search relevance feedback
US14/552,320 US9418122B2 (en) 2006-08-25 2014-11-24 Adaptive user interface for real-time search relevance feedback
US15/207,277 US20170039285A1 (en) 2006-08-25 2016-07-11 Adaptive user interface for real-time search relevance feedback

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US15/207,277 US20170039285A1 (en) 2006-08-25 2016-07-11 Adaptive user interface for real-time search relevance feedback

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
US14/552,320 Continuation US9418122B2 (en) 2006-05-02 2014-11-24 Adaptive user interface for real-time search relevance feedback

Publications (1)

Publication Number Publication Date
US20170039285A1 true US20170039285A1 (en) 2017-02-09

Family

ID=45561577

Family Applications (5)

Application Number Title Priority Date Filing Date
US12/482,389 Active 2030-04-01 US8117197B1 (en) 2008-06-10 2009-06-10 Adaptive user interface for real-time search relevance feedback
US13/355,433 Active US8543570B1 (en) 2008-06-10 2012-01-20 Adaptive user interface for real-time search relevance feedback
US14/018,791 Active US8924378B2 (en) 2006-05-02 2013-09-05 Adaptive user interface for real-time search relevance feedback
US14/552,320 Active US9418122B2 (en) 2006-05-02 2014-11-24 Adaptive user interface for real-time search relevance feedback
US15/207,277 Abandoned US20170039285A1 (en) 2006-05-02 2016-07-11 Adaptive user interface for real-time search relevance feedback

Family Applications Before (4)

Application Number Title Priority Date Filing Date
US12/482,389 Active 2030-04-01 US8117197B1 (en) 2008-06-10 2009-06-10 Adaptive user interface for real-time search relevance feedback
US13/355,433 Active US8543570B1 (en) 2008-06-10 2012-01-20 Adaptive user interface for real-time search relevance feedback
US14/018,791 Active US8924378B2 (en) 2006-05-02 2013-09-05 Adaptive user interface for real-time search relevance feedback
US14/552,320 Active US9418122B2 (en) 2006-05-02 2014-11-24 Adaptive user interface for real-time search relevance feedback

Country Status (1)

Country Link
US (5) US8117197B1 (en)

Families Citing this family (51)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8117197B1 (en) 2008-06-10 2012-02-14 Surf Canyon, Inc. Adaptive user interface for real-time search relevance feedback
US8335986B2 (en) * 2009-08-26 2012-12-18 Apple Inc. Previewing different types of documents
US8924376B1 (en) * 2010-01-31 2014-12-30 Bryant Christopher Lee Method for human ranking of search results
US20120066359A1 (en) * 2010-09-09 2012-03-15 Freeman Erik S Method and system for evaluating link-hosting webpages
US9069843B2 (en) * 2010-09-30 2015-06-30 International Business Machines Corporation Iterative refinement of search results based on user feedback
US20120110453A1 (en) * 2010-10-29 2012-05-03 Microsoft Corporation Display of Image Search Results
US20120278318A1 (en) * 2011-05-01 2012-11-01 Reznik Alan M Systems and methods for facilitating enhancements to electronic group searches
US8880492B2 (en) * 2011-05-30 2014-11-04 Google Inc. Keyword suggestion for efficient legal E-discovery
US9230014B1 (en) * 2011-09-13 2016-01-05 Sri International Method and apparatus for recommending work artifacts based on collaboration events
US8890827B1 (en) 2011-10-05 2014-11-18 Google Inc. Selected content refinement mechanisms
WO2013052866A2 (en) 2011-10-05 2013-04-11 Google Inc. Semantic selection and purpose facilitation
US9032316B1 (en) 2011-10-05 2015-05-12 Google Inc. Value-based presentation of user-selectable computing actions
US8878785B1 (en) 2011-10-05 2014-11-04 Google Inc. Intent determination using geometric shape input
US9652556B2 (en) 2011-10-05 2017-05-16 Google Inc. Search suggestions based on viewport content
US8825671B1 (en) 2011-10-05 2014-09-02 Google Inc. Referent determination from selected content
US9679568B1 (en) * 2012-06-01 2017-06-13 Google Inc. Training a dialog system using user feedback
US9183310B2 (en) * 2012-06-12 2015-11-10 Microsoft Technology Licensing, Llc Disambiguating intents within search engine result pages
US20140019462A1 (en) * 2012-07-15 2014-01-16 Microsoft Corporation Contextual query adjustments using natural action input
US9563326B2 (en) * 2012-10-18 2017-02-07 Microsoft Technology Licensing, Llc Situation-aware presentation of information
CN103793388B (en) 2012-10-29 2017-08-25 阿里巴巴集团控股有限公司 The sort method and device of search result
US10475134B2 (en) * 2013-01-24 2019-11-12 Facebook, Inc. Sponsored recommendation in a social networking system
US9576022B2 (en) 2013-01-25 2017-02-21 International Business Machines Corporation Identifying missing content using searcher skill ratings
US9613131B2 (en) 2013-01-25 2017-04-04 International Business Machines Corporation Adjusting search results based on user skill and category information
US9830612B2 (en) * 2013-03-11 2017-11-28 Capital One Financial Corporation Systems and methods for providing advertising services
JP6157192B2 (en) * 2013-04-19 2017-07-05 キヤノン株式会社 Information processing apparatus, control method, and computer program
US20140372203A1 (en) * 2013-06-14 2014-12-18 Microsoft Corporation Quality-weighted second-price auctions for advertisements
US9721218B2 (en) 2013-10-21 2017-08-01 Sap Se Determining the user-specific relevance of applications
TW201523481A (en) * 2013-12-05 2015-06-16 Richplay Information Co Ltd Method for recommending books information
US9510057B2 (en) 2014-01-03 2016-11-29 Gracenote, Inc. Interactive programming guide
US10757201B2 (en) * 2014-03-01 2020-08-25 Microsoft Technology Licensing, Llc Document and content feed
US10255563B2 (en) 2014-03-03 2019-04-09 Microsoft Technology Licensing, Llc Aggregating enterprise graph content around user-generated topics
US20150261883A1 (en) * 2014-03-13 2015-09-17 Andrew S. Wigodsky Information Context Descriptions and the Collaborative Research Engine
CN103942319B (en) * 2014-04-25 2017-11-10 北京猎豹网络科技有限公司 A kind of method and device of search
US10303725B2 (en) * 2014-05-01 2019-05-28 Airbnb, Inc. Re-ranking search results for location refining and diversity
US9827714B1 (en) 2014-05-16 2017-11-28 Google Llc Method and system for 3-D printing of 3-D object models in interactive content items
RU2014125412A (en) * 2014-06-24 2015-12-27 Общество С Ограниченной Ответственностью "Яндекс" Method for processing search request (options) and server (options)
RU2580516C2 (en) * 2014-08-19 2016-04-10 Общество С Ограниченной Ответственностью "Яндекс" Method of generating customised ranking model, method of generating ranking model, electronic device and server
US10672030B2 (en) * 2014-09-25 2020-06-02 International Business Machines Corporation Targeted advertisements from intended recipient predictions derived from user information
US10055784B1 (en) * 2014-10-28 2018-08-21 Amazon Technologies, Inc. In-stream pivoting search results
US10146876B2 (en) 2015-05-12 2018-12-04 Iquanti, Inc. Predicting real-time change in organic search ranking of a website
US10007732B2 (en) * 2015-05-19 2018-06-26 Microsoft Technology Licensing, Llc Ranking content items based on preference scores
US10073883B1 (en) * 2015-05-29 2018-09-11 Amazon Technologies, Inc. Returning query results
US20160350669A1 (en) * 2015-06-01 2016-12-01 Linkedin Corporation Blending content pools into content feeds
US20170004548A1 (en) * 2015-06-30 2017-01-05 Linkedin Corporation Generating and ranking service provider recommendations in a social network
RU2637899C2 (en) 2015-07-16 2017-12-07 Общество С Ограниченной Ответственностью "Яндекс" Method and server of determining changes in user interactive interaction with page of search results
US10785310B1 (en) * 2015-09-30 2020-09-22 Open Text Corporation Method and system implementing dynamic and/or adaptive user interfaces
US10229212B2 (en) 2016-04-08 2019-03-12 Microsoft Technology Licensing, Llc Identifying Abandonment Using Gesture Movement
CN107463580A (en) * 2016-06-06 2017-12-12 腾讯科技(深圳)有限公司 Train clicking rate prediction model method and apparatus, clicking rate predictor method and device
US10387559B1 (en) * 2016-11-22 2019-08-20 Google Llc Template-based identification of user interest
US10452413B2 (en) * 2017-03-16 2019-10-22 Ca, Inc. System and method for navigating web-based application programs
US20180329947A1 (en) 2017-05-10 2018-11-15 International Business Machines Corporation Relevance model for session search

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060031214A1 (en) * 2004-07-14 2006-02-09 Microsoft Corporation Method and system for adaptive categorial presentation of search results
US20060064411A1 (en) * 2004-09-22 2006-03-23 William Gross Search engine using user intent
US7174346B1 (en) * 2003-07-31 2007-02-06 Google, Inc. System and method for searching an extended database
US20070156636A1 (en) * 2006-01-03 2007-07-05 Yahoo! Inc. Apparatus and method for controlling content access based on shared annotations for annotated users in a folksonomy scheme
US7254580B1 (en) * 2003-07-31 2007-08-07 Google Inc. System and method for selectively searching partitions of a database
US20070192293A1 (en) * 2006-02-13 2007-08-16 Bing Swen Method for presenting search results
US20070244866A1 (en) * 2006-04-18 2007-10-18 Mainstream Advertising, Inc. System and method for responding to a search request
US20070266025A1 (en) * 2006-05-12 2007-11-15 Microsoft Corporation Implicit tokenized result ranking
US20080244428A1 (en) * 2007-03-30 2008-10-02 Yahoo! Inc. Visually Emphasizing Query Results Based on Relevance Feedback
US20080270395A1 (en) * 2007-04-24 2008-10-30 Gossweiler Iii Richard Carl Relevance Bar for Content Listings
US20090106221A1 (en) * 2007-10-18 2009-04-23 Microsoft Corporation Ranking and Providing Search Results Based In Part On A Number Of Click-Through Features
US20090164929A1 (en) * 2007-12-20 2009-06-25 Microsoft Corporation Customizing Search Results

Family Cites Families (81)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4972349A (en) * 1986-12-04 1990-11-20 Kleinberger Paul J Information retrieval system and method
US5806077A (en) 1993-01-15 1998-09-08 International Business Machines Corporation Hypertext display system
US5946678A (en) 1995-01-11 1999-08-31 Philips Electronics North America Corporation User interface for document retrieval
US5890152A (en) 1996-09-09 1999-03-30 Seymour Alvin Rapaport Personal feedback browser for obtaining media files
US5867153A (en) * 1996-10-30 1999-02-02 Transaction Technology, Inc. Method and system for automatically harmonizing access to a software application program via different access devices
US6040834A (en) 1996-12-31 2000-03-21 Cisco Technology, Inc. Customizable user interface for network navigation and management
US6182068B1 (en) 1997-08-01 2001-01-30 Ask Jeeves, Inc. Personalized search methods
US6134532A (en) * 1997-11-14 2000-10-17 Aptex Software, Inc. System and method for optimal adaptive matching of users to most relevant entity and information in real-time
US20020198882A1 (en) * 2001-03-29 2002-12-26 Linden Gregory D. Content personalization based on actions performed during a current browsing session
US6490577B1 (en) 1999-04-01 2002-12-03 Polyvista, Inc. Search engine with user activity memory
US6370526B1 (en) 1999-05-18 2002-04-09 International Business Machines Corporation Self-adaptive method and system for providing a user-preferred ranking order of object sets
WO2000075838A1 (en) 1999-06-03 2000-12-14 Cella Charles H Contingency-based options and futures for contingent travel accommodations
US6591261B1 (en) * 1999-06-21 2003-07-08 Zerx, Llc Network search engine and navigation tool and method of determining search results in accordance with search criteria and/or associated sites
US7548874B2 (en) 1999-10-21 2009-06-16 International Business Machines Corporation System and method for group advertisement optimization
US6963867B2 (en) 1999-12-08 2005-11-08 A9.Com, Inc. Search query processing to provide category-ranked presentation of search results
US6473751B1 (en) 1999-12-10 2002-10-29 Koninklijke Philips Electronics N.V. Method and apparatus for defining search queries and user profiles and viewing search results
US6981040B1 (en) * 1999-12-28 2005-12-27 Utopy, Inc. Automatic, personalized online information and product services
US7080064B2 (en) * 2000-01-20 2006-07-18 International Business Machines Corporation System and method for integrating on-line user ratings of businesses with search engines
CA2404319A1 (en) 2000-03-31 2001-10-11 Andrei Mikheev Method and system for gathering, organizing, and displaying information from data searches
US6691106B1 (en) 2000-05-23 2004-02-10 Intel Corporation Profile driven instant web portal
US6766316B2 (en) 2001-01-18 2004-07-20 Science Applications International Corporation Method and system of ranking and clustering for document indexing and retrieval
US20020107776A1 (en) 2001-02-05 2002-08-08 Homestore.Com System and method for anonymous lead generation and management
US7194454B2 (en) 2001-03-12 2007-03-20 Lucent Technologies Method for organizing records of database search activity by topical relevance
US7188106B2 (en) * 2001-05-01 2007-03-06 International Business Machines Corporation System and method for aggregating ranking results from various sources to improve the results of web searching
US6738764B2 (en) 2001-05-08 2004-05-18 Verity, Inc. Apparatus and method for adaptively ranking search results
US6920448B2 (en) 2001-05-09 2005-07-19 Agilent Technologies, Inc. Domain specific knowledge-based metasearch system and methods of using
US7149732B2 (en) 2001-10-12 2006-12-12 Microsoft Corporation Clustering web queries
US7565367B2 (en) 2002-01-15 2009-07-21 Iac Search & Media, Inc. Enhanced popularity ranking
US6991106B2 (en) 2002-04-11 2006-01-31 Ecolab Inc. Modular container for holding labels
JP2004126840A (en) * 2002-10-01 2004-04-22 Hitachi Ltd Document retrieval method, program, and system
US6829599B2 (en) * 2002-10-02 2004-12-07 Xerox Corporation System and method for improving answer relevance in meta-search engines
US7421660B2 (en) 2003-02-04 2008-09-02 Cataphora, Inc. Method and apparatus to visually present discussions for data mining purposes
US7505964B2 (en) 2003-09-12 2009-03-17 Google Inc. Methods and systems for improving a search ranking using related queries
US20050071328A1 (en) 2003-09-30 2005-03-31 Lawrence Stephen R. Personalization of web search
US20050102282A1 (en) 2003-11-07 2005-05-12 Greg Linden Method for personalized search
US20050144158A1 (en) 2003-11-18 2005-06-30 Capper Liesl J. Computer network search engine
US7444327B2 (en) 2004-01-09 2008-10-28 Microsoft Corporation System and method for automated optimization of search result relevance
US7499913B2 (en) 2004-01-26 2009-03-03 International Business Machines Corporation Method for handling anchor text
US20050192863A1 (en) 2004-02-26 2005-09-01 Krishna Mohan Web site vistor incentive program in conjunction with promotion of anonymously identifying a user and/or a group
KR20060132722A (en) * 2004-03-15 2006-12-21 야후! 인크. Search systems and methods with integration of user annotations
US7716223B2 (en) * 2004-03-29 2010-05-11 Google Inc. Variable personalization of search results in a search engine
US7305389B2 (en) 2004-04-15 2007-12-04 Microsoft Corporation Content propagation for enhanced document retrieval
US7685209B1 (en) * 2004-09-28 2010-03-23 Yahoo! Inc. Apparatus and method for normalizing user-selected keywords in a folksonomy
US7490775B2 (en) * 2004-12-30 2009-02-17 Aol Llc, A Deleware Limited Liability Company Intelligent identification of multimedia content for synchronization
US7386544B2 (en) * 2005-01-13 2008-06-10 Bridgewell Inc. Database search system
US8200700B2 (en) * 2005-02-01 2012-06-12 Newsilike Media Group, Inc Systems and methods for use of structured and unstructured distributed data
US7734601B2 (en) * 2005-02-09 2010-06-08 Sap Ag Integration of digital asset management with intellectual property management
US20060218146A1 (en) 2005-03-28 2006-09-28 Elan Bitan Interactive user-controlled relevance ranking of retrieved information in an information search system
US20070022111A1 (en) 2005-07-20 2007-01-25 Salam Aly A Systems, methods, and computer program products for accumulating, storing, sharing, annotating, manipulating, and combining search results
US7743045B2 (en) * 2005-08-10 2010-06-22 Google Inc. Detecting spam related and biased contexts for programmable search engines
US7693830B2 (en) * 2005-08-10 2010-04-06 Google Inc. Programmable search engine
US7716199B2 (en) * 2005-08-10 2010-05-11 Google Inc. Aggregating context data for programmable search engines
US8290810B2 (en) * 2005-09-14 2012-10-16 Jumptap, Inc. Realtime surveying within mobile sponsored content
US7548915B2 (en) * 2005-09-14 2009-06-16 Jorey Ramer Contextual mobile content placement on a mobile communication facility
US7496548B1 (en) 2005-09-26 2009-02-24 Quintura, Inc. Neural network for electronic search applications
US7475072B1 (en) 2005-09-26 2009-01-06 Quintura, Inc. Context-based search visualization and context management using neural networks
US8117187B2 (en) * 2005-10-28 2012-02-14 Mediareif Moestl & Reif Kommunikations-Und Informationstechnologien Oeg Method for controlling a relational database system
US7599918B2 (en) * 2005-12-29 2009-10-06 Microsoft Corporation Dynamic search with implicit user intention mining
US8996406B2 (en) * 2006-02-02 2015-03-31 Microsoft Corporation Search engine segmentation
US20070214158A1 (en) * 2006-03-08 2007-09-13 Yakov Kamen Method and apparatus for conducting a robust search
US20070239531A1 (en) * 2006-03-30 2007-10-11 Francoise Beaufays Controlling the serving of serially rendered ads, such as audio ads
US8078607B2 (en) * 2006-03-30 2011-12-13 Google Inc. Generating website profiles based on queries from webistes and user activities on the search results
US7634474B2 (en) * 2006-03-30 2009-12-15 Microsoft Corporation Using connectivity distance for relevance feedback in search
US20070260597A1 (en) * 2006-05-02 2007-11-08 Mark Cramer Dynamic search engine results employing user behavior
US8442973B2 (en) 2006-05-02 2013-05-14 Surf Canyon, Inc. Real time implicit user modeling for personalized search
US20150032717A1 (en) 2006-05-02 2015-01-29 Surf Canyon Incorporated Real time implicit user modeling for personalized search
WO2008042974A2 (en) * 2006-10-03 2008-04-10 Qps Tech. Limited Liability Company Mechanism for automatic matching of host to guest content via categorization
US8442972B2 (en) * 2006-10-11 2013-05-14 Collarity, Inc. Negative associations for search results ranking and refinement
US20080120289A1 (en) * 2006-11-22 2008-05-22 Alon Golan Method and systems for real-time active refinement of search results
US20090234814A1 (en) * 2006-12-12 2009-09-17 Marco Boerries Configuring a search engine results page with environment-specific information
US7437370B1 (en) 2007-02-19 2008-10-14 Quintura, Inc. Search engine graphical interface using maps and images
US7958104B2 (en) * 2007-03-08 2011-06-07 O'donnell Shawn C Context based data searching
US8015502B2 (en) * 2007-05-22 2011-09-06 Yahoo! Inc. Dynamic layout for a search engine results page on implicit user feedback
US8359319B2 (en) * 2007-08-27 2013-01-22 Sudhir Pendse Tool for personalized search
US9015147B2 (en) * 2007-12-20 2015-04-21 Porto Technology, Llc System and method for generating dynamically filtered content results, including for audio and/or video channels
WO2009134817A1 (en) * 2008-04-28 2009-11-05 Strands, Inc. Method for providing personalized recommendations of financial products based on user data
US8886637B2 (en) * 2008-05-12 2014-11-11 Enpulz, L.L.C. Web browser accessible search engine which adapts based on user interaction
US8117197B1 (en) 2008-06-10 2012-02-14 Surf Canyon, Inc. Adaptive user interface for real-time search relevance feedback
US8239370B2 (en) * 2008-06-27 2012-08-07 Microsoft Corporation Basing search results on metadata of prior results
US20120130858A1 (en) * 2010-11-22 2012-05-24 Matterhorn, Llc System for serving a dynamically ranked list of motor vehicles
US8959073B2 (en) 2012-11-02 2015-02-17 Swiftype, Inc. Automatically modifying a custom search engine for a web site based on user input

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7174346B1 (en) * 2003-07-31 2007-02-06 Google, Inc. System and method for searching an extended database
US7254580B1 (en) * 2003-07-31 2007-08-07 Google Inc. System and method for selectively searching partitions of a database
US20060031214A1 (en) * 2004-07-14 2006-02-09 Microsoft Corporation Method and system for adaptive categorial presentation of search results
US20060064411A1 (en) * 2004-09-22 2006-03-23 William Gross Search engine using user intent
US20070156636A1 (en) * 2006-01-03 2007-07-05 Yahoo! Inc. Apparatus and method for controlling content access based on shared annotations for annotated users in a folksonomy scheme
US20070192293A1 (en) * 2006-02-13 2007-08-16 Bing Swen Method for presenting search results
US20070244866A1 (en) * 2006-04-18 2007-10-18 Mainstream Advertising, Inc. System and method for responding to a search request
US20070266025A1 (en) * 2006-05-12 2007-11-15 Microsoft Corporation Implicit tokenized result ranking
US20080244428A1 (en) * 2007-03-30 2008-10-02 Yahoo! Inc. Visually Emphasizing Query Results Based on Relevance Feedback
US20080270395A1 (en) * 2007-04-24 2008-10-30 Gossweiler Iii Richard Carl Relevance Bar for Content Listings
US20090106221A1 (en) * 2007-10-18 2009-04-23 Microsoft Corporation Ranking and Providing Search Results Based In Part On A Number Of Click-Through Features
US20090164929A1 (en) * 2007-12-20 2009-06-25 Microsoft Corporation Customizing Search Results

Also Published As

Publication number Publication date
US8117197B1 (en) 2012-02-14
US8924378B2 (en) 2014-12-30
US8543570B1 (en) 2013-09-24
US20150081691A1 (en) 2015-03-19
US9418122B2 (en) 2016-08-16
US20140032311A1 (en) 2014-01-30

Similar Documents

Publication Publication Date Title
US10706112B1 (en) Personalization techniques using image clouds
US9953076B2 (en) Predictive query completion and predictive search results
US20160162556A1 (en) Predictive Information Retrieval
US9830390B2 (en) Related entities
US20160259861A1 (en) Systems and methods for organizing and displaying electronic media content
US20170061485A1 (en) Search assistant system and method
US10175860B2 (en) Search intent preview, disambiguation, and refinement
JP5808384B2 (en) Search system and method integrating user annotations
US9098599B1 (en) Query suggestions for a document based on user history
US9223873B2 (en) Method and system for incrementally selecting and providing relevant search engines in response to a user query
US8996560B2 (en) Search engine utilizing user navigated documents
US8782071B1 (en) Fresh related search suggestions
US9177030B2 (en) Systems and methods for providing search results
US20160299981A1 (en) Search engine that applies feedback from users to improve search results
US20140317106A1 (en) Method and apparatus for a ranking engine
RU2611971C2 (en) Displaying actions and providers associated with subjects
US9378247B1 (en) Generating query refinements from user preference data
RU2696230C2 (en) Search based on combination of user relations data
US20170109450A1 (en) System and method for conducting a profile based search
US8924372B2 (en) Dynamic image display area and image display within web search results
US8484343B2 (en) Online ranking metric
US10180979B2 (en) System and method for generating suggestions by a search engine in response to search queries
US7627826B2 (en) System for organization, display, and navigation of digital information
US7801885B1 (en) Search engine system and method with user feedback on search results
KR101963916B1 (en) Presenting related searches on a toolbar

Legal Events

Date Code Title Description
STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION