WO2008036351A2 - Systems and methods for aggregating search results - Google Patents

Systems and methods for aggregating search results

Info

Publication number
WO2008036351A2
WO2008036351A2 PCT/US2007/020361 US2007020361W WO2008036351A2 WO 2008036351 A2 WO2008036351 A2 WO 2008036351A2 US 2007020361 W US2007020361 W US 2007020361W WO 2008036351 A2 WO2008036351 A2 WO 2008036351A2
Authority
WO
Grant status
Application
Patent type
Prior art keywords
search
results
query
user
plurality
Prior art date
Application number
PCT/US2007/020361
Other languages
French (fr)
Other versions
WO2008036351A3 (en )
Inventor
Tao Yang
Wei Wang
Original Assignee
Iac Search & Media, 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

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/30Information retrieval; Database structures therefor ; File system structures therefor
    • G06F17/30861Retrieval from the Internet, e.g. browsers
    • G06F17/30864Retrieval from the Internet, e.g. browsers by querying, e.g. search engines or meta-search engines, crawling techniques, push systems

Abstract

Systems and methods for aggregating search results are disclosed herein. The systems and methods include receiving a user search query, analyzing the user search query to identify a plurality of properties of the user search query, identifying a plurality of search results that match the user search query, each search result being based on a different scheme, and aggregating the search results to produce a search results list. The search results list may be a combined and selected results list. Feedback-based optimization is also disclosed.

Description

SYSTEMS AND METHODS FOR AGGREGATING SEARCH

RESULTS

BACKGROUND OF THE INVENTION

1). Field of the Invention

[0001] This invention relates to the field of search engines and, in particular, to systems and methods for aggregating search results.

2). Discussion of Related Art

[0002] The Internet is a global network of computer systems and websites.

These computer systems include a variety of documents, files, databases, and the like, which include information covering a variety of topics. It can be difficult for users of the Internet to locate information on the Internet.

Search engines are often used by people to locate information on the

Internet.

[0003] The user of the search engine enters a search query on a website connected to the search engine via the Internet. The search engine returns results matching the search query, which are typically ranked based on relevancy.

[0004] However, multiple users entering the same search query may be searching for different information. For example, first and second users may both enter "CMU" as their search query and receive the same results; however, the first user may be searching for information about Carnegie Mellon University, while the second user may be searching for information about Central Michigan University.

[0005] Ranking of search results is determined using a ranking algorithm and each search engine typically uses a different ranking algorithm. Often, results that include content with rich link topology on the web are ranked higher, while results with weak link support (even though they may be more relevant to the search query) are ranked lower. Combining results from different search engines is used in meta-search web sites, but these meta- search web sites do not have access to inside data/ranking schemes from the search engines for well-informed decisions.

SUMMARY OF THE INVENTION

[0006] A method for aggregating search results is described herein. The method includes receiving a user search query; analyzing the user search query to identify a plurality of properties of the user search query; identifying a plurality of search results that match the user search query, each search result being based on a different scheme utilizing the search query; determining a relevance factor for each scheme; and aggregating the search results from each scheme to produce a search results list. [0007] Analyzing the user search query to identify a plurality of properties may include identifying concepts that differentiate attributes of the query. [0008] Analyzing the user search query to identify a plurality of properties may include matching the user search query to keywords of the databases. [0009] Analyzing the user search query to identify a plurality of properties may include identifying documents in the databases that match the user search query; comparing the number of documents that match the user search query with the total number of results in the database to produce a content ratio hit; and if the content hit ratio exceeds a threshold, then determining the user search query has the property of the database. [0010] The properties may be selected from the group consisting of location/ time, commerce, news, language and homepage. Some of the properties can be derived or optimized through query log analysis. [0011] The user search query may be a natural language query. [0012] Aggregating the search results to produce a search results list may include dividing each of the search results from each database into a plurality of zones; combining each of the search results from each database in each zone; and combining the search results from each zone. [0013] The method may also include providing the search results list to a user. [0014] The method may also include optimizing the aggregation and ranking of search results that match a search query with user feedback.

[0015] Optimizing aggregation of search results may include assessing a user's satisfaction with a ranking scheme.

[0016] Optimizing ranking of search results may include assessing a user's satisfaction with a document in the search results list.

[0017] Each scheme may have a database associated therewith.

[0018] A search system is described herein. The search system includes a search engine to receive a user search query; a plurality of databases to store a plurality of search results, each database related to a scheme; and a server to analyze the search query to identify a plurality of properties of the search query, match the user search query with search results in the plurality of databases based on the plurality of properties and aggregate the search results from each of the plurality of databases to produce a. search results list.

[0019] The search engine may be further to provide the search results list to a user.

[0020] The plurality of databases may each have one of the plurality of properties associated therewith.

[0021] The server may be further to divide each of the search results from each database into a plurality of zones, combine each of the search results from each database in each zone, and combine the search results from each zone.

[0022] A method of integrating multiple ranking strategies is described herein. The method includes matching a user search query with a plurality of databases, each database relating to one of a plurality of properties and a ranking scheme; producing a list of search results matching the query ranked according to the ranking scheme; and aggregating the list of search results from each database to produce a final search results list; and presenting the final search results list to a user.

[0023] The method may also include dividing each of the search results from each database into a plurality of zones; combining each of the search results from each database in each zone; and combining the search results from each zone.

[0024] The method may also include optimizing the aggregation of search results that match a search query with user feedback.

[0025] Optimizing the aggregation of search results that match a search query with user feedback may include assessing a user's satisfaction with a ranking scheme or assessing a user's satisfaction with a document in the search results list. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The invention is described by way of example with reference to the accompanying drawings, wherein:

[0027] FIG. 1 is a block diagram illustrating a system for searching in accordance with one embodiment of the invention;

[0028] FIG. 2 is a flow diagram illustrating a method for analyzing queries and aggregating rankings;

[0029] FIG. 3 is a flow diagram illustrating a method for analyzing queries;

[0030] FIG. 4 is a block diagram illustrating a method for analyzing query context;

[0031] FIG.5 is a block diagram illustrating a method for matching queries to databases;

[0032] FIG. 6 is a flow diagram illustrating a method for providing search results in response to a user query;

[0033] FIG. 7 is a flow diagram illustrating a method for optimizing search results using user feedback; and

[0034] FIG. 8 is a block diagram illustrating a method for analyzing queries, aggregating rankings and optimizing results. DETAILED DESCRIPTION OF THE INVENTION

[0035] Figure 1 of the accompanying drawings shows a network system 10 which can be used in accordance with one embodiment of the present invention. The network system 10 includes a search system 12, a search engine 14, a network 16, and a plurality of client systems 18. The search system 12 includes a server 20, a database 22, an indexer 24, and a crawler

26. The plurality of client systems 18 includes a plurality of web search applications 28a-f, located on each of the plurality of client systems 18. The server 20 includes a plurality of databases 30a-d.

[0036] The server 12 is connected to the search engine 14. The search engine 14 is connected to the plurality of client systems 18 via the network

16. The server 20 is in communication with the database 22 which is in communication with the indexer 24. The indexer 24 is in communication with the crawler 26. The crawler 26 is capable of communicating with the plurality of client systems 18 via the network 16 as well.

[0037] The web search server 20 is typically a computer system, and may be an HTTP server. It is envisioned that the search engine 14 may be located at the web search server 20. The web search server 20 typically includes at least processing logic and memory.

[0038] The indexer 24 is typically a software program which is used to create an index, which is then stored in storage media. The index is typically a table of alphanumeric terms with a corresponding list of the related documents or the location of the related documents (e.g., a pointer). An exemplary pointer is a Uniform Resource Locator (URL). The indexer 24 may build a hash table, in which a numerical value is attached to each of the terms. The database 22 is stored in a storage media, which typically includes the documents which are indexed by the indexer 24. The index may be included in the same storage media as the database 22 or in a different storage media. The storage media may be volatile or non-volatile memory that includes, for example, read only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices and zip drives.

[0039] The crawler 26 is a software program or software robot, which is typically used to build lists of the information found on Web sites. Another common term for the crawler 26 is a spider. The crawler 26 typically searches Web sites on the Internet and keeps track of the information located in its search and the location of the information.

[0040] The network 16 is a local area network (LAN), wide area network (WAN), a telephone network, such as the Public Switched Telephone Network (PSTN), an intranet, the Internet, or combinations thereof. [0041] The plurality of client systems 18 may be mainframes, minicomputers, personal computers, laptops, personal digital assistants (PDA), cell phones, and the like. The plurality of client systems 18 are capable of being connected to the network 16. Web sites may also be located on the client systems 18. The web search application 28a-f is typically an Internet browser or other software.

[0042] The databases 30a-d are stored in storage media located at the server 20. The storage media may be volatile or non-volatile memory that includes, for example, read only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices and zip drives.

[0043] In use, the crawler 26 crawls websites, such as the websites of the plurality of client systems 18, to locate information on the web. The crawler 26 employs software robots to build lists of the information. The crawler 26 may include one or more crawlers to search the web. The crawler 26 typically extracts the information and stores it in the database 22. The indexer 24 creates an index of the information stored in the database 22. Alternatively, if a database 22 is not used, the indexer 24 creates an index of the located information and the location of the information on the Internet (typically a URL).

[0044] When a user of one of the plurality of client systems 18 enters a search on the web search application 28, the search is communicated to the search engine 14 over the network 16. The search engine 14 communicates the search to the server 20 at the search system 12. The server 20 matches the query to one or more of the databases 30a-d to identify a search result. The server 20 communicates the search result to the user via the search engine 14 and network 16.

[0045] Figure 2 shows a method for analyzing user queries and aggregating search results. The process 40 begins at block 42 where a user search query is received. The user search query may be a natural language search query.

[0046] At block 44, an intention and property analysis is performed on the user search query. Any number of techniques may be used to identify the user intent. The user intent is quantified by identifying properties of the search query. Exemplary properties include location (e.g., local vs. national vs. global), time (e.g., recent vs. historical), commerce (e.g., buying/selling products/services), news, language, homepage, and the like. [0047] At block 46, the query is matched with databases based on the intention and property analysis. The databases may be the databases 30a-d located at the server 20. Each of the databases is related to a different property or contains documents with multiple properties matchable with the selected property. Those databases relating to properties identified as query properties are searched. [0048] At block 48, the results are aggregated. The search results from each of the databases searched may be combined to produce an aggregated, ranked list of search results. At block 50, the ranked list of the search results is provided to the user.

[0049] Figure 3 shows the process for identifying a user's intent and the property analysis of the user search query in more detail. The process 60 begins at block 62 where the user search query is analyzed. The query analysis is performed to classify the query. By classifying the query, the intention of the user and the type of content that should be matched and/or the ranking scheme can be identified.

[0050] At block 64, the query is classified by one or more properties Pj and a confidence level Fj. Exemplary properties include location, time, commerce, news, language, homepage and the like.

[0051] The query can be classified by identifying concepts that differentiate attributes of the query (block 66). This can be done by identifying terminology relating to the property, (block 68). For example, if the user query is "Infiniti Silicon Valley," or "'Who sells Infiniti's in the Silicon Valley?" relevant properties that may characterize the search include location (e.g., Silicon Valley indicates a region in northern California), commerce (e.g., Infiniti is a well-known brand of cars and 'sells' is a common commerce term), etc. Other well-known Natural Language Processing (NLP)-based text matching techniques may also be used to classify the query.

[0052] Alternatively, the query can be classified by matching the query to keywords of databases containing documents with the property Pj (block 70). For example, a database relating to commerce may include the keywords: buy, sell, product, service, price, certain brand names, and the like. A query that includes terms matching (exact matches or similar matches) the keywords is likely to be related to that property. The confidence Fj can be determined based on the degree of matching between the query and the keywords (block 72). An offline web data mining system can scan through all web sites on the Internet and identify home pages of persons or organizations. Keywords associated with each page may be stored with the database.

[0053] Alternatively, a content hit ratio can be identified. A content hit ratio is a relative measurement of hits based on a degree of matching towards each property. That is, the number of results matched to the query in a database containing documents with the property Pj is compared to the total number of results in that database (block 74). The following formula can be used in this analysis:

∑ Hk [0054] A high value exceeding a certain threshold indicates that it is more appropriate to match the query with database content of property Pj. For example, if a query is compared to all of the possible databases, and the hit rate is in a range of, for example, 1% to 60%, the databases with, for example, more than a 30% hit ratio are related to a relevant property for the query. The hit ratio threshold may be any value or range of values between 1% and 100%.

[0055] After the query has been classified with the property Pj and confidence Fj, the confidence level can, optionally, be adjusted (block 76). The confidence level can be adjusted, for example, by analyzing historical query logs. The historical query logs identify past search queries by any number of users, the search results provided, and may also identify the links/documents selected by the users.

[0056] Figure 4 shows an exemplary method for adjusting the confidence level F) 80. Using a historical query log, at least a first query 82a and second query 82b are examined within a user session.

[0057] A transition probability matrix is computed for property changes for two consecutive queries (e.g., query 82a and query 82b) from the same user. The transition probability matrix includes element Xi,j which represents the probability the second query 82b has property Pj 84 when the first query 82a has property Pi 84 in a query context. This element Xi,j may be represented

by the following formula:

# query pairs with property pair (.PnPj ) Total # queries

Total # query pairs # queries with property Pj

[0058] The confidence level can be determined using the following

equation:

[0059] For example, if the user first searches for Britney Spears and

subsequently searches for Christina Aguilera, the relevant properties may

include, for example, time and news. The probability that the search for

Christina Aguilera includes similar properties is very high, and, therefore,

the confidence level that the same ranking schemes should be used for the

related queries can be increased. Thus, certain ranking schemes can be

associated with certain properties, and the confidence that a particular

ranking scheme should be associated with a particular property can be

adjusted.

[0060] Figure 5 shows a method for matching the query with databases.

The process 90 includes identifying a first ranking scheme 92a, second

ranking scheme 92b, and any number of additional ranking schemes 92c.

The ranking schemes represent different methodologies for ranking the

documents in a database. Exemplary ranking schemes include link popularity, page popularity, frequency and location of words in a document, link analysis and the like.

[0061] Each ranking scheme 92 is associated with a database 94. The databases 94a-c each are related to a different property Pi, P2 and Pn, respectively. The databases 94 include a plurality of documents having the associated property.

[0062] Exemplary properties include location, time, commerce, news, language, homepage and the like. Thus, the documents can be classified by the language of the documents, the geographical location of document publishers, the publishing period of a document, the extent of commercial content, the likelihood the site is a home page, etc.

[0063] A query having a set of attributes Pi of confidence Fi, as determined with the user intent and property analysis described above with reference to Figure 3, is matched against the databases 94 to produce a list 96. Each list 96a-c contains documents relating to each property associated with the database and ranked according to the associated ranking scheme 92. [0064] In particular, for each property Pj and its ranked page list Lj, a specific ranking score Sι,j is provided for each matched document di, representing relevancy of the document with respect to the desired property. The ranking score Si,) is determined by the ranking scheme. [0065] As described above, a confidence ratio Fj is also determined for matching the intention of the query with the property Pj. A confidence ratio Gi.) is also determined for classifying the document di for property Pj. The combined confidence for selecting such a document di for the property Pj is G,j = X(Gi,), Fj). For example, G,j = Gi,j x Fj. The combined confidence can be used to improve ranking of the documents.

[0066] Figure 6 shows an alternative process for identifying user intention and identifying search results 100. The process 100 begins at block 102 where a user enters a search query.

[0067] User intention is identified for property R with confidence Fi (block 104). For each document, a confidence for selecting it for Fi is calculated (block 106). The property Pi and confidence Fi can be identified as described above with reference to Figure 3 (and, optionally, Figure 4). [0068] A list of results, separated into multiple zones, is returned for each property Pi (block 108). The list of results is returned by matching the query to databases having the identified properties, as described above with reference to Figure 5. The multiple zones represent tiers of quality. There may be any number of zones including as few as one zone. For example, the multiple zones may include a "highly relevant" zone, a "relevant" zone, and a "probably relevant" zone. In general, for each list Lj, a criterion can be used to divide the list into t zones. The sublist in zone k for list Lj is Lj,k. Each sublist Lj,k includes all of the documents that match the query in that zone.

[0069] Aggregated results are calculated for each zone based on user intention (block 110). Results from each list are selected and ranked at each zone. For each zone k, there are n sublists: Li,k, Lzk, ..., Ln,k. The documents may be sorted based on an aggregated ranking score. The aggregated ranking score for di at zone k is:

K,,* = ∑ U WJ * Cu *su where Wj is a weighting factor to be adjusted based on the final ranking need, G,j is the combined confidence for selecting document di for property Pj, and Ski,jis a ranking score in the list Lj for document di. For each zone, the results from each sublist are combined.

[0070] At block 112, the multiple zones are combined into a final result. The combined lists from each zone are then combined together into a final result including all the results from all the lists organized by the zones. This approach allows matched results with different properties and confidence scores to be selected and combined to produce an aggregated list. In one embodiment, the system may set a limit to select the top k results to present to the user, based on application needs. [0071] Figure 7 shows a detailed method for optimizing search results 120. At block 122, search results are provided to a user. At block 124, user's selection of links or documents in the search results are monitored.

[0072] At block 126, the satisfaction level towards a scheme choice is identified. Exemplary scheme choices include link popularity, page popularity, frequency and location of words in a document, link analysis and the like. The satisfaction level towards scheme choice can be identified by selecting results from different result lists Lj and computing the satisfactory degree Bj to the use of the list Lj. The satisfactory degree Bj with respect to the use of list Lj, measured by the percentage of times that the list Lj contributes to the final ranking which has been confirmed by users, can be identified by the following formula: where F(Ts,t, j) = 1 if j eTs,t; otherwise 0, TM is the set of list names which have contributed positively to ranking position t for the s-th query in the log, hs is the number of results selected by users for a query, m is the number of times the same query was asked in the log, and

a = Σ " *=1 B L . β — Max" R

°avg - ^ max ~ maxk-A t5 j where n is the number of possible properties. If the absolute value of (Bj-

Bavg)/Bmax exceeds a threshold, then the feedback is strong enough to

influence a new ranking. The adjusted weighting factor with the feedback

can be:

B1 -B v if — — > δ, where δ is a threshold. max

[0073] At block 128, the satisfaction level towards individual results is

identified. The satisfaction level can be determined by calculating a

satisfactory degree Vi, which is assessed for each individual result for

document ά\. m

where Z (Ts,t, di) = 1 if D(TS/l) = di; otherwise 0.

[0074] At block 130, the combined confidence level Cy can be adjusted

based on the satisfaction level towards the scheme choice and/or the

satisfaction level towards individual results. The combined confidence level

G,j can be adjusted as follows:

U,J = CIJ * β + Vl * {l ~jβ)

The final ranking with feedback for zone k can be computed as follows: [0075] For the results clicked by a user, an algorithmic choice may be kept in the query log history. Thus, personalized confidence scores can be identified to improve ranking.

[0076] Figure 8 shows a detailed method for analyzing a query, aggregating results and optimizing the results with user feedback 140. The process 140 begins at block 142 where a user query is received.

[0077] At block 144, an intention and property analysis is performed. The intention and property analysis is performed as described above with reference to Figure 3 (and, optionally, Figure 4).

[0078] At block 146, the query is matched against the databases having different properties. As described above with reference to Figure 5, a first ranking schema 148a, second ranking schema 148b and any additional number of ranking schemas 148c may be included. Each ranking schema 148 includes a respective database 150a, 150b and 150c that relates to a specific property. Results in the database that match the query are presented in respective lists 152a, 152b and 152c.

[0079] At block 154, the results are aggregated. In one embodiment, the results are aggregated by zones 156a, 156b, 156c, as described above with reference to Figure 6.

[0080] At block 158, the aggregated ranking results are presented to the user, as described above with reference to Figure 6. [0081] At block 160, user feedback may be monitored. At block 162, a satisfaction assessment may be performed. The user feedback and satisfaction assessment are performed as described above with reference to Figure 7. In particular, the schema choice (block 164), URL (document) choice (block 166) and context evaluation (block 168) can be monitored and assessed, as described above with reference to Figure 7. The assessment may be used to modify the result aggregation (block.154). [0082] Systems and methods described herein are advantageous because they produce better search results. For example, the systems and methods described herein can automatically determine whether the search query relates to a local search vs. a global search, commercial products, temporal content, mixed languages, etc. Systems and methods described herein also provide for personalization of search methods.

[0083] Multiple ranking schemes can be used with the same database, allowing a combination from different ranking strategies and properties. The results with the different ranking strategies can then be aggregated. [0084] The zoning-based aggregation scheme allows for diversification of the results and ensures that documents from each list have a chance to appear in the top positions if they are highly relevant. Thus, one list does not necessarily dominate the final ranking. [0085] The foregoing description with attached drawings is only illustrative of possible embodiments of the described method and should only be construed as such. Other persons of ordinary skill in the art will realize that many other specific embodiments are possible that fall within the scope and spirit of the present idea. The scope of the invention is indicated by the following claims rather than by the foregoing description. Any and all modifications which come within the meaning and range of equivalency of the following claims are to be considered within their scope.

Claims

CLAIMSWhat is claimed:
1. A method for aggregating search results comprising: receiving a user search query; analyzing the user search query to identify a plurality of properties of the user search query; identifying a plurality of search results that match the user search query, each search result being based on a different scheme utilizing the search query; determining a relevance factor for each scheme; and aggregating the search results from each scheme to produce a search results list.
2. The method of claim 1, wherein analyzing the user search query to identify a plurality of properties comprises: identifying concepts that differentiate attributes of the query.
3. The method of claim 1, wherein analyzing the user search query to identify a plurality of properties comprises: matching the user search query to keywords of the databases.
4. The method of claim 1, wherein analyzing the user search query to identify a plurality of properties comprises: identifying documents in a plurality of databases, each database associated with a scheme, that match the user search query; comparing the number of documents that match the user search query with the total number of results in the database to produce a content ratio hit; and if the content hit ratio exceeds a threshold, then determining the user search query has the property of the database.
5. The method of claim 1, wherein the properties are selected from the group consisting of location, time, commerce, news, language and homepage.
6. The method of claim 5, wherein the properties are determined through query and log analysis.
7. The method of claim 1, wherein the user search query is a natural language query.
8. The method of claim 1, aggregating the search results to produce a search results list comprises: dividing each of the search results from each scheme into a plurality of zones; combining each of the search results from each scheme in each zone; and combining the search results from each zone.
9. The method of claim 8, further comprising selecting a portion of the search results to present to the user, wherein the portion presented corresponds to one of the plurality of zones.
10. The method of claim 9, wherein the portion presented corresponds to the highest ranked zone.
11. The method of claim 1, further comprising providing the search results list to a user.
12. The method of claim 11, further comprising optimizing the aggregation and ranking of search results that match a search query with user feedback.
13. The method of claim 12, wherein optimizing aggregation of search results comprises: assessing a user's satisfaction with each scheme.
14. The method of claim 12, wherein optimizing ranking of search results comprises: assessing a user's satisfaction with a document in the search results list.
15. The method of claim 1, wherein each scheme has a database associated therewith.
16. The method of claim 1, wherein the search results list is a combined and selected results list.
17. A search system comprising: a search engine to receive a user search query; a plurality of databases to store a plurality of search results, each database related to a scheme; and a server to analyze the search query to identify a plurality of properties of the search query, match the user search query with search results in the plurality of databases based on the plurality of properties and aggregate the search results from each of the plurality of databases to produce a search results list.
18. The search system of claim 17, wherein the search engine is further to provide the search results list to a user.
19. The search system of claim 17, wherein the plurality of databases each have one of the plurality of properties associated therewith.
20. The search system of claim 17, wherein the server is further to divide each of the search results from each database into a plurality of zones, combine each of the search results from each database in each zone, and combine the search results from each zone.
21. A method of integrating multiple ranking strategies comprising: matching a user search query with a plurality of databases, each database relating to one of a plurality of properties and a ranking scheme; producing a list of search results matching the query ranked according to the ranking scheme; and aggregating the list of search results from each database to produce a final search results list; and presenting the final search results list to a user.
22. The method of claim 21, further comprising: dividing each of the search results from each database into a plurality of zones; combining each of the search results from each database in each zone; and combining the search results from each zone.
23. The method of claim 21, further comprising optimizing the identification of search results that match a search query with user feedback.
24. The method of claim 23, wherein optimizing the identification of search results that match a search query with user feedback comprises assessing a user's satisfaction with a ranking scheme or assessing a user's satisfaction with a document in the search results list with feedback optimization.
PCT/US2007/020361 2006-09-19 2007-09-19 Systems and methods for aggregating search results WO2008036351A3 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US11524011 US8407229B2 (en) 2006-09-19 2006-09-19 Systems and methods for aggregating search results
US11/524,011 2006-09-19

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
GB0904629A GB0904629D0 (en) 2006-09-19 2007-09-19 Systems and methods for aggregating search results

Publications (2)

Publication Number Publication Date
WO2008036351A2 true true WO2008036351A2 (en) 2008-03-27
WO2008036351A3 true WO2008036351A3 (en) 2008-09-25

Family

ID=39189874

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2007/020361 WO2008036351A3 (en) 2006-09-19 2007-09-19 Systems and methods for aggregating search results

Country Status (3)

Country Link
US (1) US8407229B2 (en)
GB (1) GB0904629D0 (en)
WO (1) WO2008036351A3 (en)

Families Citing this family (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070157093A1 (en) * 2005-12-30 2007-07-05 Patrick Karcher Systems and methods for adaptive help mechanisms for a user
EP1909197A1 (en) * 2006-10-03 2008-04-09 Pointer S.R.L. Systems and methods for ranking search engine results
US8661029B1 (en) * 2006-11-02 2014-02-25 Google Inc. Modifying search result ranking based on implicit user feedback
US7657513B2 (en) * 2006-12-01 2010-02-02 Microsoft Corporation Adaptive help system and user interface
US7720843B2 (en) * 2007-03-27 2010-05-18 Andrew Thomas Brunner Real-time end-user aware interactive search utilizing layered approach
US7792813B2 (en) * 2007-08-31 2010-09-07 Microsoft Corporation Presenting result items based upon user behavior
US8996376B2 (en) 2008-04-05 2015-03-31 Apple Inc. Intelligent text-to-speech conversion
US8504555B2 (en) * 2008-06-25 2013-08-06 Microsoft Corporation Search techniques for rich internet applications
CN101667179B (en) * 2008-09-03 2012-08-15 华为技术有限公司 Mobile search method and system, and method for synchronizing meta-indexing of search server
US8681144B2 (en) 2008-12-19 2014-03-25 International Business Machines Corporation Prioritized rendering of objects in a virtual universe
US9805492B2 (en) 2008-12-31 2017-10-31 International Business Machines Corporation Pre-fetching virtual content in a virtual universe
US8990306B2 (en) 2009-05-28 2015-03-24 International Business Machines Corporation Pre-fetching items in a virtual universe based on avatar communications
US9256896B2 (en) * 2009-08-27 2016-02-09 International Business Machines Corporation Virtual universe rendering based on prioritized metadata terms
US8498974B1 (en) 2009-08-31 2013-07-30 Google Inc. Refining search results
CN102063432A (en) 2009-11-12 2011-05-18 阿里巴巴集团控股有限公司 Retrieval method and retrieval system
US8639688B2 (en) * 2009-11-12 2014-01-28 Palo Alto Research Center Incorporated Method and apparatus for performing context-based entity association
US20110295678A1 (en) * 2010-05-28 2011-12-01 Google Inc. Expanding Ad Group Themes Using Aggregated Sequential Search Queries
US20120246154A1 (en) * 2011-03-23 2012-09-27 International Business Machines Corporation Aggregating search results based on associating data instances with knowledge base entities
US9959326B2 (en) 2011-03-23 2018-05-01 International Business Machines Corporation Annotating schema elements based on associating data instances with knowledge base entities
US20130024448A1 (en) * 2011-07-21 2013-01-24 Microsoft Corporation Ranking search results using feature score distributions
US8645361B2 (en) * 2012-01-20 2014-02-04 Microsoft Corporation Using popular queries to decide when to federate queries
CN103377240B (en) * 2012-04-26 2017-03-01 阿里巴巴集团控股有限公司 Method for providing information, processing server and server consolidation
WO2014197334A3 (en) 2013-06-07 2015-01-29 Apple Inc. System and method for user-specified pronunciation of words for speech synthesis and recognition
US9965521B1 (en) * 2014-02-05 2018-05-08 Google Llc Determining a transition probability from one or more past activity indications to one or more subsequent activity indications
US9633085B2 (en) * 2014-06-06 2017-04-25 Macy's West Stores, Inc. Method for calculating relevancy scores of search results
US9668121B2 (en) 2014-09-30 2017-05-30 Apple Inc. Social reminders
US20170195365A1 (en) * 2015-12-30 2017-07-06 International Business Machines Corporation Curtailing search engines from obtaining and controlling information
US9785717B1 (en) 2016-09-29 2017-10-10 International Business Machines Corporation Intent based search result interaction

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030187839A1 (en) * 2002-03-28 2003-10-02 International Business Machines Corporation Method and structure for federated web service discovery search over multiple registries with result aggregation
US20070038601A1 (en) * 2005-08-10 2007-02-15 Guha Ramanathan V Aggregating context data for programmable search engines

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6408093B1 (en) * 1999-09-08 2002-06-18 Lucent Technologies Inc. Method for comparing object ranking schemes
US20060200455A1 (en) * 2002-12-20 2006-09-07 Redbank Manor Pty Ltd Search engine result reporter

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030187839A1 (en) * 2002-03-28 2003-10-02 International Business Machines Corporation Method and structure for federated web service discovery search over multiple registries with result aggregation
US20070038601A1 (en) * 2005-08-10 2007-02-15 Guha Ramanathan V Aggregating context data for programmable search engines

Also Published As

Publication number Publication date Type
WO2008036351A3 (en) 2008-09-25 application
US20080071742A1 (en) 2008-03-20 application
US8407229B2 (en) 2013-03-26 grant
GB0904629D0 (en) 2009-04-29 grant
GB2455248A (en) 2009-06-10 application

Similar Documents

Publication Publication Date Title
US7953740B1 (en) Detection of behavior-based associations between search strings and items
US7620628B2 (en) Search processing with automatic categorization of queries
US6073130A (en) Method for improving the results of a search in a structured database
Fan et al. Personalization of search engine services for effective retrieval and knowledge management
US6944609B2 (en) Search results using editor feedback
Teevan et al. To personalize or not to personalize: modeling queries with variation in user intent
US8090717B1 (en) Methods and apparatus for ranking documents
Collins-Thompson et al. Query expansion using random walk models
US7340460B1 (en) Vector analysis of histograms for units of a concept network in search query processing
US20050015366A1 (en) Disambiguation of search phrases using interpretation clusters
US7136845B2 (en) System and method for query refinement to enable improved searching based on identifying and utilizing popular concepts related to users' queries
US7568148B1 (en) Methods and apparatus for clustering news content
US20060248059A1 (en) Systems and methods for personalized search
US20060095416A1 (en) Link-based spam detection
US20050165753A1 (en) Building and using subwebs for focused search
US20080065600A1 (en) Method and apparatus for providing search results from content on a computer network
US20070043723A1 (en) Interactive user-controlled relevanace ranking retrieved information in an information search system
US7260573B1 (en) Personalizing anchor text scores in a search engine
Carmel et al. Personalized social search based on the user's social network
US8494897B1 (en) Inferring profiles of network users and the resources they access
US6560600B1 (en) Method and apparatus for ranking Web page search results
US20040049514A1 (en) System and method of searching data utilizing automatic categorization
US20030078928A1 (en) Network wide ad targeting
US20080313144A1 (en) Method for enhancing search results
Leung et al. Personalized web search with location preferences

Legal Events

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

Ref document number: 07838550

Country of ref document: EP

Kind code of ref document: A2

WWE Wipo information: entry into national phase

Ref document number: 0904629.3

Country of ref document: GB

ENP Entry into the national phase in:

Ref document number: 0904629

Country of ref document: GB

Kind code of ref document: A

Free format text: PCT FILING DATE = 20070919

NENP Non-entry into the national phase in:

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 07838550

Country of ref document: EP

Kind code of ref document: A2