United States Patent
`Skillen et al.
`
`[191
`
`US006098065A
`Patent Number:
`Date of Patent:
`
`[111
`[451
`
`6,098,065
`Aug. 1, 2000
`
`[54] ASSOCIATIVE SEARCH ENGINE
`
`[75] Inventors: Richard Prescott Skillen, Mississauga;
`Frederick Caldwell Livermore,
`Stittsville, both of Canada
`
`[73] Assignee: Nortel Networks Corporation,
`Montreal, Canada
`
`[21] Appl. No.: 08/798,747
`[22]
`Filed:
`Feb. 13, 1997
`
`[51] Int. Cl.7 .................................................... .. G06F 17/30
`[52] US. Cl. ....................... .. 707/3; 707/5; 707/6; 707/10
`[58] Field of Search ................................ .. 707/3, 5, 6, 10;
`395/20048
`
`[56]
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`5,682,525 10/1997 Bouve et a1. ......................... .. 707/104
`
`5,694,594 12/1997 Chang . . . . . . . . . . .
`
`. . . . . . .. 707/6
`
`5,721,827
`5,721,903
`5,727,129
`5,732,216
`5,740,549
`5,778,362
`
`395/20047
`2/1998 Logan et a1. ..
`......... .. 707/5
`2/1998 Anand 618.1.
`....... .. 706/10
`3/1998 Barrett et a1.
`.. 395/20033
`3/1998 Logan 618.1. ..
`.. 705/14
`4/1998 Riley 618.1.
`7/1998 Deerwester ............................... .. 707/5
`
`FOREIGN PATENT DOCUMENTS
`
`0 749 081 12/1996 European Pat. Off. .
`
`OTHER PUBLICATIONS
`
`Anonymous, “Internet Access: Internet Marketing Revolu
`tion Begins in the US This Sep. Hyper System Promises
`NeW Cash FloW For Providers, Free Access For Users & A
`NeW Marketing Opportunity For Businesses”, EDGE:
`Work—Group Computing Report, vol. 7, Jun. 1996.
`Anonymous, Make Sure Search Engines Find Your Site,
`Options include buying Words, advertising and careful page
`design, PR NeWs, vol. 52, No. 19, May 6, 1996, 515 Words.
`
`Schreiber, EA. et a1., “Dynamic User Pro?les and Flexible
`Queries in Of?ce Document Retrieval Systems”, Decision
`Support Systems, Jan. 1, 1989, vol. 5, No. 1, pp. 13—28.
`
`Database Inspec. No. 5374275, Institute of Electrical Engi
`neers, Stevenage, GB, XP002062388, Kohda, Y. et al., “A
`NeW Advertising Business Framework in the 1:1 Market
`ing”, see abstract and Transactions of the Information Pro
`cessing Society of Japan, Inf. Process. Soc. Japan, Japan,
`Jun. 1996, vol. 37, No. 6, ISSN 0387—5806, pp. 1235—1236.
`
`Database Inspec. No. 5371096, Institute of Electrical Engi
`neers, Stevenage, GB, XP002062389, Yager, R.R., “Intelli
`gent Agents on the World Wide Web”, see abstract and 1996
`Workshop on “Flexible Query—AnsWering Systems”, Rosk
`ilde, Denmark, May 22—24 1996 and Datalogiske Skrifter,
`1996, Univ. Roskilde, Denmark, No. 62, ISSN 0109—9779,
`pp‘ 289_306'
`
`(List continued on neXt page.)
`
`Primary Examiner—Paul R. LintZ
`Attorney, Agent, or Firm—Vernon E. Williams
`
`[571
`
`ABSTRACT
`
`This invention relates to an advertisment machine Which
`provides advertisements to a user searching for desired
`information Within a data netWork. The machine receives,
`from the user, a search request including a search argument
`corresponding to the desired information and searches,
`based upon the received search argument, a ?rst database
`having data netWork related information to generate search
`results. It also correlates the received search argument to a
`particular advertisement in a second database having adver
`tisement related information. The search results together
`With the particular advertisement are provided by the
`machine to the user.
`
`1 Claim, 2 Drawing Sheets
`
`ACCESS PROVIDER EQUIPMENT SITE
`
`32
`\ 3e
`
`TO INTERNET
`
`ROUTER
`
`CONTEXTUAL
`DATABASE
`
`PRODUCT
`DATABASE
`
`39 \_.
`
`END USER
`
`I TERMINATING
`
`EQUIPMENT
`
`SERVER
`
`38
`
`DATABASE =
`SEARCH
`ENGINE
`
`AssocIATIvE
`SEARCH
`ENGINE
`
`,
`
`AOL Ex. 1004
`Page 1 of 7
`
`

`

`6,098,065
`Page 2
`
`OTHER PUBLICATIONS
`
`Database Inspec. No. 5306185, Institute of Electrical Engi
`neers, Stevenage, GB, XP002062390, Kohda, Y. et al.,
`“Ubiquitous Advertising on the WWW: Merging Advertise
`rnent on the Browser”, see abstract and Fifth International
`World Wide Web Conference, Paris, France, May 6—10,
`1996 and Computer Networks and ISDN Systems, Elsevier,
`Netherlands, May 1996, vol. 28, No. 7—11, ISSN
`0169—7552, pp. 1493—1499.
`Database Inspec. No. 3814556, Institute of Electrical Engi
`neers, Stevenage, GB, XP002062391, Myaeng, S.H. et al.,
`
`“Integration of User Pro?les: Models and Experiments in
`Information Retrieval”, see abstract and Information Pro
`cessing & Management, vol. 26, No. 6, ISSN 0306—4573,
`1990, pp. 719—738.
`Database Inspec. No. 2221234, Institute of Electrical Engi
`neers, Stevenage, GB, XP002062392, Chavarria, GarZa H.,
`“Information Retrieval Models Using Queries and User
`Pro?les”, see abstract and Tercera Conferencia Internacional
`en Ciencia de la Cornputacion (3rd International Conference
`on Computer Science) Santigo, Chile, Jun. 20—24, 1983 and
`Ponti?cia Univ. Catolica de Chile, Santiago, Chile, 1983,
`pp. 113—130.
`
`AOL Ex. 1004
`Page 2 of 7
`
`

`

`U.S. Patent
`
`Aug. 1, 2000
`
`Sheet 1 of 2
`
`6,098,065
`
`vcCo
`
`ASVEVLVG
`ASVEVLVd
`
`LONGOdd
`IWOALXSLNOO
`
`O¢
`
`OL
`
`AAILVIDOSSV
`
`ASVEVLVd
`
`HOYVAS
`
`ANIDNA
`
`HOeVsaS
`
`ANIONS
`
`BL91dasnGNA
`
`L‘Sla
`
`AOL Ex. 1004
`Page 3 of 7
`
`AOL Ex. 1004
`Page 3 of 7
`
`
`
`
`

`

`U.S. Patent
`
`Aug. 1,2000
`
`Sheet 2 of2
`
`6,098,065
`
`AOL Ex. 1004
`Page 4 of 7
`
`

`

`1
`ASSOCIATIVE SEARCH ENGINE
`
`BACKGROUND OF THE INVENTION
`
`The invention generally relates to search engines and, in
`particular, to an associative search methodology, based on a
`contextual search engine, for retrieving related information.
`The Internet provides an excellent vehicle for access to
`information about goods and services on a global basis. In
`theory, anyone can access information about any product. In
`practice, the problem is one of ?nding the correct informa
`tion. Many techniques for solving this problem are knoWn,
`including: indexing systems such as Yahoo, graphical elec
`tronic malls, hall of malls, directories, and text search
`engines, such as OpenText. These techniques and tools for
`searching and retrieving information, in their present form,
`can inundate the user With large amounts of unWanted
`material.
`This later problem can be alleviated someWhat by
`enabling the user to provide search statements as a set of
`criteria Which are combined With logical operators, such as
`‘AND’, ‘OR’ and ‘NOT’ operators. HoWever, many prob
`lems still exist With this. One of them is that the user is left
`to his/her oWn imagination to try to think of all the alter
`native descriptions (i.e. keywords) of a product or service.
`As search engines and techniques become ever more poW
`erful in the number and diversity of databases they can
`access, the amount of information Which it is possible to
`present to a user can quickly become excessively large. The
`problem or opportunity still remains to quickly ?nd the
`relevant information for Which the user is looking.
`
`SUMMARY OF THE INVENTION
`
`It is an object of the present invention to provide a neW
`and improved associative search methodology for retrieving
`related information.
`The invention, therefore, according to a ?rst broad aspect
`provides a method of providing advertisements to a user
`searching for desired information Within a data netWork,
`comprising the steps of: receiving, from the user, a search
`request including a search argument corresponding to the
`desired information; searching, based upon the received
`search argument, a ?rst database having data netWork related
`information to generate search results; correlating the
`received search argument to a particular advertisement in a
`second database having advertisement related information;
`and providing the search results together With the particular
`advertisement to the user.
`According to a second broad aspect, the invention pro
`vides a method of searching for desired information Within
`a data netWork, comprising the steps of: receiving, from a
`user, a search request including a search argument corre
`sponding to the desired information; searching, based upon
`the received search argument and user pro?le data, a data
`base of information to generate a search result; and provid
`ing the search results to the user.
`According to a third broad aspect, the invention provides
`a system for providing advertisements to a user searching for
`desired information Within a data netWork, comprising:
`means for receiving, from the user, a search request includ
`ing a search argument corresponding to the desired infor
`mation; means for searching, based upon the received search
`argument, a ?rst database having data netWork related
`information to generate search results; means for correlating
`the received search argument to a particular advertisement in
`a second database having advertisement related information;
`
`10
`
`15
`
`25
`
`35
`
`45
`
`55
`
`65
`
`6,098,065
`
`2
`and means for providing the search results together With the
`particular advertisement to the user.
`According to a fourth broad aspect, the invention provides
`a system for searching for desired information Within a data
`netWork, comprising: means for receiving, from a user, a
`search request including a search argument corresponding to
`the desired information; means for searching, based upon
`the received search argument and user pro?le data, a data
`base of information to generate a search result; and means
`for providing the search results to the user.
`Conventional search engines, for example OpenText, pro
`vide a basis on Which the methodology according to this
`invention may be implemented. In a particular embodiment,
`the invention is manifested by an advertising system includ
`ing an associative search engine that may be tied into and
`form an integral part of the conventional search engine.
`When an end user accesses the conventional search engine,
`the associative engine of the selling system examines the
`user’s choices and search instructions, that have been input
`by the user, and formulates the necessary strategy and tactics
`to offer products that Would appeal to the end user based on
`his/her inputs and choices up until that point in the search.
`Since the process is dynamic the strategy and tactics can be
`continually re?ned and the results presented to the end user
`in a predictive order and fashion that relates to that end
`user’s past preferences (either for the immediate search
`underWay or including all prior search data stored for that
`user) and a contextual database. This effectively provides
`unobtrusive, related and useful data and options to the end
`user Who is searching for information. The processor used
`for the advertising and selling mechanism can be a part of or
`interconnected With the search engine.
`Some examples Will illustrate hoW this invention can have
`high value to an end user. An end user’s pro?le data may
`contain such information as the make, model and year of
`automobile he/she oWns or leases. When tires are needed,
`the manufacturers recommended tire types and options may
`be considered as sort criteria provided Within a contextual
`database. Another example relates to When airline reserva
`tions are being sought. All of a person’s preferences, Which
`today are normally sorted one by one through a travel agent,
`could be utiliZed by the advertising and selling mechanism
`to provide the informed response.
`In addition to the search criterion, as users interact With
`the advertising system, it Will continue to compile prefer
`ence data (eg a list of keyWords) for each user. All
`preferences, for example, can be left in a type of default
`mode or even presented to the user for him/her to edit and
`re-prioritiZe in order to look for diversity or alternatives. The
`system can detect exceptions and contradictions so that the
`end user can be shoWn immediately that some of the options
`Would violate natural tendencies.
`Some of the search patterns or preferences Will be keyed
`off of natural interests also such as: social, family, political,
`technological, geographical, environmental, educational and
`so on. Once these preferences are knoWn, then an adver
`tisement or a proposed customiZed product brochure can be
`prepared.
`This invention is also enhanced by advances in technol
`ogy occurring in the emerging ubiquitous data World, such
`as neW softWare languages. One example is “Java”. Java is
`an environment that can be exported to an end user’s
`platform to run an application Which may include graphics,
`moving demonstrations, cartoon like explanations or even
`video. The Java like environment could be used to customiZe
`With more than name; it can be used to customiZe With
`detailed examples.
`
`AOL Ex. 1004
`Page 5 of 7
`
`

`

`6,098,065
`
`3
`One other opportunity presented to service providers is
`the potential to create neW and better Ways to improve the
`environment presented to end users of various classes by
`remembering What Was effective for a sale to occur. In this
`Way, key elements of a presentation or search that have
`higher success can be put into a higher priority or category
`for further use and therefore more rapid development
`through even more intense feedback due to subsequent use.
`While this invention can be integrated With traditional
`Internet search engines (such as Yahoo or Lycos), full
`advantage can be obtained by integrating the invention With
`an Internet Access Provider. An Internet Access Provider can
`maintain a more complete user database. Customers are
`localiZed, so individual databases Will contain feWer users.
`This alloWs more database space to be allocated to each user.
`In addition, more detailed information can be maintained,
`such as geographical location, type of home computer
`system, and any additional information the user may choose
`to provide to the Internet Access Provider.
`It is anticipated that the Java environment Will develop so
`that it too Will learn and adapt. As a user continues to stop
`presentations at certain points, scan, re-scan, look for further
`input at certain points, abort a negotiation and so on at
`certain points, then such points or logic routines could be
`avoided or alternatives considered. The present invention
`proposes such a dynamic and relational preferences meth
`odology in order to more rapidly and more effectively couple
`a user’s needs With product or service offerings.
`It is also recogniZed that these techniques may be applied
`to an educational environment Where, for example, a student
`may be able to have an environment that is aurally enhanced,
`another one that is visually enhanced, another one that is
`textually enhanced and yet another that is anecdotally as
`opposed to logically enhanced. There are many other oppor
`tunities due to different personal proclivities, strengths,
`skills or disabilities.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`The invention Will be better understood from the folloW
`ing detailed description of an advertising machine together
`With reference to the accompanying draWings, in Which:
`FIG. 1 is a schematic representation of the advertising
`machine including an associative search engine; and
`FIG. 2 is a schematic representation of the advertising
`machine in combination With an Internet Access Service
`provider.
`
`DETAILED DESCRIPTION
`Referring to FIG. 1, shoWn is an advertising machine,
`generally referenced by 10, together With a data processing
`device 12 and a communications link 14 through Which the
`device 12 interacts With the advertising machine 10. The
`communications link 14 may be provided by a global data
`netWork, typically the Internet, and the data processing
`device 12 may be any conventional hardWare/softWare com
`bination supporting functionality for communications over
`the Internet. Examples of the data processing device 12
`include a personal computer (PC) or Macintosh executing an
`appropriate broWser application, such as, Netscape Naviga
`tor Which functions as an interface to the World Wide Web
`(WWW) of the Internet. Functionality provided by the
`advertising machine 10 may be implemented using an
`appropriately programmed conventional data processing
`server platform.
`The advertising machine 10, in this particular
`embodiment, comprises a database search engine 16 and an
`
`4
`associative search engine 18 Which may access a database
`20 having contextual data 22 and product data 24. The
`database search engine 16 is conventional technology, an
`example of Which is the OpenText engine that provides
`searches based on subject, strings, boolean, text, etc. Such
`input search arguments may be received from an end user
`via the data processing device 12 and data link 14 and,
`accordingly, the database search engine 16 effects a search
`of the contextual data 22 in the database 20 and returns
`results of the search to the end user, as a page displayed on
`the device 12. The contextual database typically contains
`information relating to the Internet, for instance, keyWords
`associated With respective WWW site locations.
`The associative search engine 18, in accordance With the
`present invention, may contain rule based softWare algo
`rithms or non-precise techniques, such as, fuZZy logic that
`can correlate a search argument derived from the user and
`changes in the argument during a single session, to particular
`product data Within the product database 24. The associative
`search engine 18 selects the most logical product from its
`available data 24 and then provides an advertisement insert
`that is added to the end user’s search page, in an attempt to
`present the end user With the product that is closest to the
`need as determined by the associative search engine 18 of
`the selling machine 10.
`In operation, for example, the process effected When an
`end user at the device 12 accesses the advertising machine
`10 is as folloWs.
`1) The end user device 12 sends a search request outlining
`the search argument to the database search engine 16.
`2) The search engine 16, having carried out a traditional
`search of Internet related information in the contextual
`data 22, passes the argument and results of the its search
`to the associative search engine 18 Which then looks for
`a match in the product data 24 of the database 20. The
`associative search engine 18 may determine a logical
`product ?t to the initial search argument, or it may create
`a logical tree analysis of possible product ?ts and selects
`a probable best product for an advertisement WindoW to
`be displayed With the search results.
`3) The associative engine 18 passes the data of the selected
`product to the search engine 16 Which in turn provides the
`results of the search against the given argument together
`With the initial product advertisement to the device 12 for
`displaying to the end user.
`4) The end user re?nes his or her search by either clicking
`on a displayed result or re?ning their search through
`additional arguments or search criteria.
`5) The associative search engine 18 and the database search
`engine 16 again Work together in providing re?ned data as
`in (2) above. In the case of the end user clicking on (i.e.,
`selecting) a speci?c search result, the associative search
`engine 18 further re?nes its logical tree strategy and
`selects the probable best ?t product and generates an
`advertisement.
`6) The end user search results advertisement WindoW is
`continually updated on each selection or re?nement using
`a technology such as Java providing the end user With a
`continually updated product advertisement that is consid
`ered to be most relevant (best ?t) at that point in the
`search. In addition, failure of the end user to click on the
`advertisement is used as a criterion in the logical tree in
`providing the appropriate advertisement.
`7) Once the end user clicks on (selects) the advertisement
`displayed on the device 12, an appropriate message is sent
`to the advertising machine 10 and, in particular, to the
`associative search engine 18 Which may automatically
`
`10
`
`15
`
`25
`
`35
`
`45
`
`55
`
`60
`
`AOL Ex. 1004
`Page 6 of 7
`
`

`

`6,098,065
`
`5
`connect the end user to the seller of the product, for
`instance, by forwarding the browser of the end user device
`12 to the WWW site address of the seller. In the product
`data 24 for this seller, the associative search engine may
`record that this transaction occurred, in order that a toll for
`bringing a buyer and seller together may be extracted.
`Moreover, user pro?le data may be maintained on end
`user device 12 and accessed by the associative search engine
`18. The pro?le data contains, for example, end user prefer
`ences and previous search arguments Which may be used to
`augment the individual search arguments received With the
`search request to select a best ?t product advertisement. The
`associative search engine 18 retrieves and updates the pro?le
`data on the device 12, using appropriate messages
`exchanged over the communications link 14. For example,
`the search arguments from the current search session may be
`added to the user pro?le data.
`Turning to the system illustrated in FIG. 2, the advertising
`machine 30 (similar to the machine 10 in FIG. 1) is
`integrated as part of the functionality embodied at an Inter
`net access provider equipment site 32 Which typically
`includes a telephone netWork terminating equipment 34, a
`router 36 through Which TCP/IP packets are transmitted to
`and received from the Internet, and a server 38 Which in
`general controls operation and couples data calls terminated
`by equipment 34 to router 36. The link 39 represents a data
`call established through the telephone netWork to the access
`provider site 32.
`The associative search engine 40 of machine 30 contains
`softWare algorithms or non-precise techniques, such as,
`fuZZy logic that correlates a search argument derived from
`the user and changes in the argument during a single session
`With the product database 42 Whereby the most logical
`product from its available list may be selected. It then
`provides an advertisement insert that is added to the end
`user’s search page in an attempt to offer the end user With the
`product that is closest to the need as determined by the
`associative search engine 40.
`The access provider site 32, incorporating the advertising
`machine 30, is the sole channel available to the end user for
`accessing the Internet. When an end user carries out a search
`using the search engine 44 and clicks on (selects) a speci?c
`result or chooses an alternate information site address, the
`information is passed to the associative search engine 40 of
`that access provider 32. Thus, not only the page With the
`search results but also the home page of the access provider
`can be updated With the appropriate advertisement Which is
`selected by the associative search engine 40 from the
`product database 42. In addition, since the data processing
`device 12 of the end user is connected through the access
`provider site 32 for all destinations Within the Internet, a
`comprehensive user pro?le database 48 may be maintained
`by the associative search engine 40, about the end users
`preferences and previous search arguments Which may be
`used to augment the individual search argument received
`With the search request to select a best ?t product adver
`tisement.
`In operation, for example, the process effected When an
`end user at the device 12 accesses the Internet through
`access provider site 32 is as folloWs.
`1) The end users access the search engine 44, provided as a
`basic service by the access provider, and request a search
`outlining the search argument.
`2) The search engine 44 passes the argument and results of
`its search in database 46 to the associative search engine
`40 Which looks for a match in the product database 42
`determining a logical product ?t to the initial search
`
`10
`
`15
`
`25
`
`35
`
`45
`
`55
`
`65
`
`6
`argument. The associative search engine 40 creates a
`logical tree analysis of possible product ?ts, and selects or
`creates a probable best product advertisement for the
`advertisement WindoW on the search results. In this case,
`the associative search engine 40 also can utiliZe the
`maintained pro?le on the end user from past search
`sessions and/or historical data gathered on their buying
`habits, in the product selection processing. The associa
`tive search engine correlates the user’s identity to data in
`the user pro?le database 48 Which it maintains and
`updates With data (e.g., search argument received With
`search request) from the current search session.
`3) The search engine 44 provides the end user With the
`results of the search against the given argument including
`the initial product advertisement from the associative
`search engine 40. The results and advertisement are
`displayed by end user device 12.
`4) The end user re?nes his/her search by either clicking on
`a result, or re?ning their search through additional argu
`ments or search criteria.
`5) The associative search engine 40 and the database search
`engine 44 again Work together in providing re?ned data as
`in (2) above. In the case of the end user clicking on
`(selecting) a speci?c search result the associative search
`engine 40 further re?nes its logical tree strategy and
`selects the probable best ?t product and generates an
`advertisement.
`6) The end user device 12 search results advertisement
`WindoW is continually updated on each selection or
`re?nement using a technology such as Java, providing the
`end user With a continually updated product advertise
`ment that is considered to be most relevant (best ?t) at that
`point in the search. In addition, failure of the end user to
`click on the advertisement is used as a criterion in the
`logical tree in providing the appropriate advertisement.
`The access provider’s home page sent to the end user can
`also contain a specialiZed advertisement constructed from
`the end user’s pro?le data (e.g., searches and/or buying
`habits).
`7) Once the end user is led to click on the advertisement, the
`associative search engine 40 of the selling machine 30
`takes control and connects the potential buyer to the
`WWW site of the seller of the product, thereby exacting
`a toll for the transaction of bringing a buyer and seller
`together.
`Those skilled in the art Will recogniZe that various modi
`?cations and changes could be made to the invention With
`out departing from the spirit and scope thereof. It should
`therefore be understood that the claims are not to be con
`sidered as being limited to the precise embodiments of the
`selling machine set forth above, in the absence of speci?c
`limitations directed to each embodiment.
`We claim:
`1. A method of searching for desired information Within
`a data netWork, comprising the steps of:
`receiving, from a user, a search request including a search
`argument corresponding to the desired information;
`searching, based upon the received search argument and
`user pro?le data, a database of information to generate
`a search result; and
`providing the search results to the user
`Wherein searching the database includes correlating, as a
`function of a fuZZy logic algorithm, the received search
`argument and user pro?le data to particular information
`in the database, and providing the particular informa
`tion as the search results.
`
`AOL Ex. 1004
`Page 7 of 7
`
`