(12) United States Patent
`Wachtel
`
`(10) Patent N0.:
`(45) Date of Patent:
`
`US 6,195,654 B1
`*Feb. 27, 2001
`
`US006195654B1
`
`(54) SYSTEM AND METHOD FOR OBTAINING
`IMPROVED SEARCH RESULTS AND FOR
`DECREASING NETWORK LOADING
`
`5,655,088 * 8/1997 Midorikawa et al. ............. .. 395/237
`5,659,732 * 8/1997 Kirsch .................................... .. 707/5
`5,682,478 * 10/1997 Watson et a1.
`395/200.12
`5,710,884 * 1/1998 Dedrick ...... ..
`395/200.47
`
`
`
`3rd Ave' Inventor: Edward I Wachtel, New York, NY (Us) 10010
`
`
`
`5,813,009
`
`:
`
`
`
`Nielsen . . . . . . . . 9/1998 Johnson et a1. .................... .. 707/100 . . . ..
`
`
`
`(*) Notice:
`
`This' patent issued on a continued pros
`
`eclltlon appllcatlon ?led under 37 CFR
`
`
`
`* 5,875,437 * 2/1999 Atkins . . . . . . . . . .
`
`. . . . . .. 705/40
`
`5,819,243 * 10/1998 Rich et a1. ........................... .. 706/11
`
`*
`
`.
`
`153(9), andis subjectto the twenty year
`patent term provisions of 35 U.S.C.
`154(a)(2).
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U50 154(1)) by 0 days-
`
`(21) Appl. No.: 08/751,613
`(22) Filed:
`Nov. 18, 1996
`
`Related US Application Data
`(60) Provisional application No. 60/006,863, ?led on Nov. 16,
`1995
`
`7
`Int. Cl. .................................................... ..
`(52) US. Cl. ...................... .. 707/3; 707/2; 707/5; 707/10;
`709/217; 709/218; 709/219
`Of Search ................................ ..
`10, 3,
`39500033, ZOO-03> ZOO-09> 20057; 364/284,
`282-4; 346/2821; 709/217, 218, 219
`
`(56)
`
`References Cited
`Us PATENT DOCUMENTS
`
`11/1991 Winters -
`5,063,522
`5,301,314 * 4/1994 Gifford et al. ......................... .. 707/5
`5’313’55 9
`5/1994 0g?“ et a1‘ '
`5,404,505 * 4/1995 Levinson ................................ .. 707/7
`5,483,652 * 1/1996 Sudama et al.
`707/5
`574937728 * 2/1996 solton et a1‘ “
`395/250
`
`705/14
`5,933,811 * 8/1999 Angles et al.
`. 709/224
`5,954,798 * 9/1999 Shelton et al.
`5,960,422 * 9/1999 Prasad ...................... .. 707/2
`6,035,339 * 3/2000 Agraharam et a1. ............... .. 709/246
`* Cited b examiner
`y
`Primary Examiner—Hosain T. Alam
`Assistant Examiner—Jean M. Corrielus
`(74) Attorney, Agent, or Firm—Craig E. Shinners
`(57)
`ABSTRACT
`
`A networked information 'sharing model is described. The
`network descnbed compnses 2} Chent'server model of a
`client only model. There exists a shared information
`database, a shared category database, a shared interest
`pro?le database and a Shared Client enhancement database,
`each of Which is Continually and dynamically updated The
`shared category database contains categories of interests,
`are
`and marked information units_
`Weights are arrived at by empirical use. Marks are main
`tained to distinguish Where the information came from and
`to access information according to client source preference.
`The shared interest pro?le contains a set of pro?les Which
`clients are associated With. Useful client categories Within
`pro?les are offered When requested. Ashared client enhance
`ment list is maintained to identify and Weight useful sources
`of information Aclient Speci?c database is maintained With
`client categories, preferred information sources, Weights and
`- hted information access histor 'This database is used in
`Wehg
`.
`.
`y
`.
`.
`.
`con]unct1on Wllh' the shared databases to provide intelligent
`
`5,499,221 * 3/1996 Ito etal. . . . . . . . .
`
`. . . .. 369/32
`
`lnformatlon sharlng
`
`707/2
`5,600,831 * 2/1997 Levy et a1. ...... ..
`5,617,565 * 4/1997 Augenbraun et al. ................. .. 707/4
`
`17 Claims, 14 Drawing Sheets
`
`CLIENT DATABASE &
`
`INTEREST CATEGORIES
`INFORMATION UNITS ACCESSED
`
`J55
`
`PROFILE NUMBERS
`
`PREVIOUSLY REJECTED CATEGORIES J
`
`INTEREST CATEGORY
`
`INDEXES OF INFORMATION ACCESSED
`
`CATEGORY UTILITY WEIGHT
`
`CATEGORY USE COUNT
`
`NOT PREFERRED INFORMATION CLIENT IDS
`
`PREFERRED INFORMATION CLIENT IDS
`
`CATEGORY WEIGHT CHANGE
`
`o
`
`N
`
`5
`
`m
`
`INFORMATION UNITS ACCESSED
`
`INFORMATION ITEMS
`CATEGORIES CHOSEN
`
`SERVER DESTINATIONS
`
`74
`I/
`76
`J
`78
`/
`
`Petitioner Apple Inc. - Exhibit 1005, p. 1
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 1 0f 14
`
`US 6,195,654 B1
`
`18
`
`FIGURE 1
`
`Petitioner Apple Inc. - Exhibit 1005, p. 2
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 2 0f 14
`
`US 6,195,654 B1
`
`SUGGESTED INTEREST CATEGORIES
`
`- NODE6
`
`- NODE?
`
`- NODE8 '
`
`CURRENT INTEREST CATEGORIES
`
`- NODE1
`- NODEZ
`- NODE3
`
`- NODE4
`
`- NlODE5
`
`s4
`/
`
`3e
`j
`
`CHILD EXPANDED NODE
`
`.
`
`as
`
`- NODE 3A
`
`- NODEGB
`
`' NODE 3C
`
`j
`
`PARENT EXPANDED NODE 5Q
`
`NODE 3E
`
`NODE 3D
`
`F|GURE 2
`
`NODE 3
`
`Petitioner Apple Inc. - Exhibit 1005, p. 3
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 3 0f 14
`
`US 6,195,654 B1
`
`INFORMATION UNIT?
`
`USEFULNESS WEIGHTS
`
`ACCESS COUNTS
`
`\, K, \3
`
`INFORMATION ACCESS AREA
`
`FIGURE 3
`
`Petitioner Apple Inc. - Exhibit 1005, p. 4
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 4 0f 14
`
`US 6,195,654 B1
`
`CLIENT DATABASE ggA
`
`58
`J7
`INTEREST CATEGORIES
`2
`J80
`INFORMATION UNITS ACCESSED
`J82
`PROFILE NUMBERS
`PREVIOUSLY REJECTED CATEGORIES J
`
`INTEREST CATEGORY
`
`6O
`J62
`INDEXES OF INFORMATION ACCESSED
`J64
`CATEGORY UTILITY WEIGHT
`J66
`CATEGORY USE COUNT
`NOT PREFERRED INFORMATION CLIENT IDS »/68
`PREFERRED INFORMATION CLIENT IDS
`J70
`
`CATEGORY WEIGHT CHANGE
`
`/
`
`INFORMATION UNITS ACCESSED
`
`INFORMATION ITEMS
`CATEGORIES CHOSEN
`SERVER DESTINATIONS
`
`74
`M76
`J78
`/
`
`FIGURE 4
`
`Petitioner Apple Inc. - Exhibit 1005, p. 5
`
`

`

`U. C
`
`9
`
`84
`
`INITIALIZE /
`SHARED
`DATABAS Ev
`
`ASSIGN
`INTEREST
`PROFILES PER
`CLIENT
`
`GHQ/$820‘?
`DEST'NAT‘ONS-
`CATEGOR'ES'
`INFORMATION
`
`ANALYZE
`INFORMATION
`FOR
`USEFULNESS
`
`[94
`
`CLIENT
`REQUEST AND
`RECEIVE
`DISTINATIONS
`INFORMATION,
`1
`CATEGORIES
`
`UPDATE CLIENT
`INTEREST
`PROFILES
`
`FIGURE 5
`
`Petitioner Apple Inc. - Exhibit 1005, p. 6
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 6 0f 14
`
`US 6,195,654 B1
`
`96
`
`100
`
`CLIENT
`ACCESSES
`SERVER
`
`l
`
`\- CLlENT SELECTS ____________-__-p
`A CATEGORY
`
`98
`CLIENT SERVER
`ACCESS COUNT /
`INCREMENTED
`
`102
`
`CLIENT /
`
`CATEGORY
`ACCESS
`INCREMENTED
`
`106
`
`110
`
`CLIENT SELECTS ________.____-->
`A SUBCATEGORY
`
`CLIENT
`SUBCATEGORY
`ACCESS
`INCREMENTED
`
`RESOURCE
`ACCESS COUNT
`INCREMENTED
`
`i
`108
`\ CLIENT'SELECTS
`A RESOURCE
`FROM THE
`SUBCATEGORY
`
`l
`112
`\ CLIENT RATES
`CATEGORY
`SUBCATEGORY
`AND RESOURCE
`
`FIGURE 6
`
`Petitioner Apple Inc. - Exhibit 1005, p. 7
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 7 0f 14
`
`US 6,195,654 B1
`
`CLIENT OONNEOTS TO SERVER
`
`114
`
`K
`
`SERVER ACCESSES CLIENT USEFULNESS K
`
`INDEX
`
`116
`
`SERVER REQUESTS INFORMATION
`ABOVE AN ACCESS COUNT AND RATING
`LEVEL
`
`CLIENT SENDS SERVER NEW
`INFORMATION WITH COUNTS AND
`RATINGS
`
`CLIENT MARKS INFORMATION AS SENT
`
`118
`
`K
`
`120
`
`K
`
`122 K
`
`SERVER RECORDS INFORMATION WITHIN
`AN ORDERED WEIGHTED LIST
`
`124 K
`
`FIGURE?
`
`Petitioner Apple Inc. - Exhibit 1005, p. 8
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 8 0f 14
`
`US 6,195,654 B1
`
`126
`
`128
`
`136
`
`INFORMATION WITH
`A USE COUNT
`
`RATING =1‘?
`
`NO
`
`WEIGHT = COUNT +
`COUNT
`
`WEIGHT = COUNT +
`(COUNT * 0.7)
`
`WEIGHT = COUNT +
`(COUNT * 0.5)
`
`140
`\NO RATING. WEIGHT
`= COUNT + (COUNT *
`0.7)
`
`142
`
`FINAL WEIGHT =
`WEIGHT * CLIENT
`INDEX
`
`FIGURE 8
`
`Petitioner Apple Inc. - Exhibit 1005, p. 9
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 9 0f 14
`
`US 6,195,654 B1
`
`144
`
`, TRY '
`
`INFORMATION
`
`146
`
`RATED
`UTILITY 7 OR
`MORE?
`
`152
`
`HALT
`
`148
`\_ ovzw??gm
`COUNT +
`UTILITY 2
`
`150
`STATIC COUNT = J '
`STATIC couéw +
`UTIUTY
`
`154
`
`FIRST OF
`NEW MONTH
`'?
`
`NO
`
`156
`\- DYNAMIC COUNT
`
`= DYNAMIC f’ HALT
`COUNT * 0.7
`
`158
`
`FIGURE 9
`
`Petitioner Apple Inc. - Exhibit 1005, p. 10
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 10 0f 14
`
`US 6,195,654 B1
`
`160
`
`CLIENT REQUESTS
`INFORMATION
`FROM A CATEGORY
`
`CLIENT SENDS
`162
`\_ INDEXES OF
`INFORMATION
`ACCESSED
`
`164
`
`CLIENT SENDS
`PREFERRED AND
`NOT PREFERRED
`IDS
`
`O
`N
`
`YES
`
`166
`
`170
`
`PREFERRED
`IDS?
`
`YES
`
`HIGHER
`WEIGHTED
`IDS?’
`
`NO
`
`168
`
`SEND HIGHEST
`WEIGHTED
`INFORMATION TO
`CLIENT
`
`NO
`
`NEW ID
`INFORMATION
`RECENT?
`
`174
`YES
`
`‘I72
`
`SEND PREFERRED
`ID INFORMATION TO
`CLIENT
`
`FIGURE 10
`
`Petitioner Apple Inc. - Exhibit 1005, p. 11
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 11 0f 14
`
`US 6,195,654 B1
`
`176
`
`POST
`REQUEST 7
`
`184
`
`TIMESLICE =
`TIMESLICE UNIT "‘
`SEARCH MASTER
`MULTIPLIER
`
`ICONITINUE
`ARCHIE SEARCH
`
`186
`
`J
`
`178
`
`180
`
`REQUEST
`‘ NSWERED '7
`
`REQUEST
`UTILITY > 6 ?
`
`182
`
`CLIENT ANSWER
`COUNT = CLIENT
`ANSWER COUNT
`+ UTlLlTY 2
`
`188
`\- HALT
`
`FIGURE 11
`
`Petitioner Apple Inc. - Exhibit 1005, p. 12
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 12 0f 14
`
`US 6,195,654 B1
`
`190
`
`INFORMATION IS USED
`IN AN INTEREST
`CATEGORY
`
`INTEREST
`CATEGORY COUNT
`INCREMENTED
`
`192
`
`NO
`
`EXITYING
`NETWORK ?
`
`194
`
`196
`
`K HOURS ON
`NETWORK ?
`
`NO
`
`205
`
`198
`____-—-’——-——-—-‘
`\ COUNT REDUCED
`FOR CATEGORIES
`WITH K HOURS
`
`200
`
`CATEGORY
`COUNT > J ?
`
`NO
`
`202
`
`204
`
`' REMOVE
`CATEGORlES WITH
`A LOW COUNT
`
`MARK CATEGORIES
`AS CHANGED
`
`FIGURE 12
`
`Petitioner Apple Inc. - Exhibit 1005, p. 13
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 13 0f 14
`
`US 6,195,654 B1
`
`206
`
`CONNECT TO
`SERVICE PROVIDER
`
`216
`
`SEND PREVIOUSLY
`REJECTED
`CATEGORIES
`
`208 .
`
`\\ ' SEND CURRENT
`INTEREST
`CATEGORIES
`
`210
`
`NO
`
`CATEGORY
`PROFILE CHANGE
`?
`
`SEND LIST OF
`CATEGORIES
`
`CATEGORIES
`REJECTED '?
`
`212 ASSIGN CLIENT TO
`\\ CLOSEST PROFILE
`‘N PROFILE
`DIRECTORY
`
`RECORD
`__-_ REJECTION IN
`CLIENT DATABASE
`
`214
`
`REQUEST FOR
`NEW CATEGORIES
`'?
`
`224
`
`ADD CATEGORY TO
`CLIENT DATABASE
`
`CONTINUE SERVICE
`CONNECTION
`
`FIGURE 13
`
`Petitioner Apple Inc. - Exhibit 1005, p. 14
`
`

`

`U.S. Patent
`
`Feb. 27, 2001
`
`Sheet 14 0f 14
`
`US 6,195,654 B1
`
`gag
`
`21;
`
`28B\ & /30A
`
`20A\ F'E'l /18A
`
`32A/
`
`FIGURE 14
`
`Petitioner Apple Inc. - Exhibit 1005, p. 15
`
`

`

`US 6,195,654 B1
`
`1
`SYSTEM AND METHOD FOR OBTAINING
`IMPROVED SEARCH RESULTS AND FOR
`DECREASING NETWORK LOADING
`
`This application claims the bene?t of prior provisional
`application Ser. No. 60/006,863, ?led on Nov. 16, 1995.
`
`BACKGROUND
`
`1. Field of Invention
`This invention relates to computer systems and in par
`ticular to an intelligent means of acquiring, storing and
`sharing information.
`2. Description of Prior Art
`Servers on the Internet contain vast quantities of infor
`mation and are distributed around the globe. HoWever, the
`vast majority of information is of no use to a particular
`person. Finding information of use requires considerable
`knoWledge as Well as time and money. Mosaic offers a
`graphical user interface to the Internet making access easier.
`Yet there are tens of thousands of servers to choose from and
`a large quantity of information to sift through once on an
`individual server. Furthermore, the server is usually sloW
`due to the number of persons logged onto it and by the
`netWork traf?c to communicate With it.
`There are librarian servers on the Internet Which scan
`thousands of servers and catalogue the ?les on the servers.
`HoWever, these librarian servers are sloW due to the mag
`nitude of the search and the large number of requesters.
`Further, one may ?nd hundreds of potential ?les on a given
`topic; accessing and reading the ?les to ?nd useful ones
`takes and Wastes considerable time. These servers may cover
`some topics to a considerable degree and others sparsely.
`Services such as CompuServe and America Online alle
`viate congestion problems to a considerable degree by
`charging money. HoWever, since the on-line service is
`charging per minute, one may not have the time to sift
`through on-line services and bulletin boards to ?nd What one
`is looking for. The on-line service reduces the vast quantities
`of useless information on the Internet by offering a smaller
`set of services and bulletin boards found to be of interest to
`most people. HoWever, the list of services is still very large
`and one is not con?dent Which if any Will be of interest.
`Furthermore, excellent information may be available on the
`Internet or elseWhere Which the particular on-line service
`does not offer.
`Bulletin boards may haphaZardly provide speci?c infor
`mation of interest. HoWever, one must sift through ansWers
`Which may or may not be of interest. Furthermore, one must
`?nd the bulletin board of interest; on the Internet there are
`a vast number Which may or may not suit a person.
`Expert systems are available Which sift through informa
`tion by use of algorithms, driven by rules and stored in a
`knoWledge base. HoWever, expert systems are expensive
`and time consuming to produce and maintain. It Would be
`impossible to cover the vast and evolving information
`located on the Internet. Furthermore, the processing time
`required to run the expert systems Would reduce the
`response time of these already sloW servers considerably.
`Another option is an heuristic database Weighted by the
`usefulness response of clients. Patent number 5301314 to
`Gifford (1991) describes this method. The method consists
`of placing information Which Was determined useful Within
`a category higher on the tree of offered information Within
`the category. This method falls short in several Ways. First,
`one must sift through the categories. Second, once in the
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`category, one must sift through excellent ansWers until one
`?nds the excellent ansWer Which matches the question; the
`ansWer may not even be on the database or may be located
`in a different category. Third, people have differing interests
`even When the subject of a search is narroWly de?ned. For
`this reason the assumption made by prior art search engines
`that everyone is interested in the documents vieWed by the
`majority often results in search results that are of little
`interest or value to the person conducting the search—
`therefore. In a category of movies, as an example, an
`excellent choice for an English Professor may be a poor
`choice to an engineer. Forth, What if one is not sure What
`categories may be of interest. There are millions on the
`Internet and thousands in subscriber services Which may or
`may not suit a given person.
`Another option is to provide trained personnel to search
`for information. This solution is expensive. Further, What is
`a good information to the personnel may not be to the person
`requesting the information. If the question is highly special
`iZed in a given ?eld, the personnel may not have the
`technical knoWledge to ?nd the appropriate information.
`Finally, searching even by a trained person, takes a consid
`erable amount of time.
`An intelligent computer based method to share informa
`tion is needed Which Will reduce traf?c on a netWorked
`system of computers and processing load on server
`machines. This method should offer the best information for
`a particular persons needs, Whether that information is
`located locally on the server, is located on the Internet or is
`chosen from a set of responses on a bulletin board.
`
`OBJECTS AND ADVANTAGES
`
`It is, therefore, an object of the invention to provide a
`method of sharing information betWeen client computers
`Which Will decrease the traffic on a netWork and decrease the
`load on a server machine.
`It is another object of the invention for the server com
`puter or the client computer to choose information Which
`Will optimally best serve the particular client’s need.
`It is another object of the invention to gather information
`from the Internet or other sources Which Will be of future
`value to clients and keep a database of that information or
`pointers to that information.
`It is another object of the invention to provide an effective
`method of offering topics of interest to clients on an indi
`vidual basis.
`Still further objects and advantages Will become apparent
`from a consideration of the ensuing description and accom
`panying draWings.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 shoWs a netWork of computers consisting of clients
`and servers Which Will share information.
`FIG. 2 shoWs a client interface to the interest categories.
`FIG. 3 shoWs the structure of an information unit.
`FIG. 4 shoWs the client database With an interest category
`and an information units accessed area.
`FIG. 5 shoWs the netWorked information sharing model.
`FIG. 6 shoWs the recording of an information access by a
`client
`FIG. 7 shoWs the upload and processing of client infor
`mation to a server.
`FIG. 8 shoWs a method of calculating a client information
`Weight.
`
`Petitioner Apple Inc. - Exhibit 1005, p. 16
`
`

`

`US 6,195,654 B1
`
`3
`FIG. 9 shows a method of Weighting information used by
`clients from a server.
`FIG. 10 is a model showing the intelligent retrieval of
`requested information.
`FIG. 11 is a model for a database Which can be searched
`in or posted to for information.
`FIG. 12 shoWs the intelligent update of a client category
`list on the client database.
`FIG. 13 shoWs the intelligent update of the server interest
`pro?le list and the intelligent offer of neW categories.
`FIG. 14 shoWs a netWorked information sharing model
`consisting only of clients.
`
`SUMMARY
`
`A networked information sharing model is described
`comprising a client-server model or a client only model
`Where there exists a shared information database, a shared
`category database, a shared interest pro?le database and a
`shared client enhancement database, each of Which is con
`tinually and dynamically updated, the shared category data
`base containing categories of interests, Within Which are
`Weighted and marked information units, Weights arrived at
`by empirical use and marks maintained to distinguish Where
`the information came from and to access information
`according to client source preference. The shared interest
`pro?le contains a set of pro?les Which clients are associated
`With Whereby useful client categories Within pro?les are
`offered When requested, a shared client enhancement list
`maintained to identify and Weight useful sources of infor
`mation and a client speci?c database maintained With client
`categories, preferred information sources, Weights and
`Weighted information access history Whereby this database
`is used in conjunction With the shared databases to provide
`intelligent information sharing.
`
`PREFERRED EMBODIMENT—DESCRIPTION
`
`FIG. 1 shoWs the preferred netWork structure of the
`invention. This consists of a set of clients Who Will share
`information over a communications link 26. There is a client
`database 28 containing a client index of interest categories
`and associated category choice information. There is one or
`more servers Which contain a complete set of interest
`categories in the form of a complete category database 18.
`The interest category client database 28 is a subset of the
`complete category database 18. Server information database
`20 consists of information or pointers to information. Infor
`mation units pointing to information in server database 20
`are kept for each category stored in complete category
`database 18. There is an interest pro?le database 32 con
`taining a full set of interest pro?les. Each client on the
`information netWork is assigned to one or more interest
`pro?les. There is a client database enhancement list 30. This
`list contains each client and a number representing the
`amount of useful information from this client Which Was
`useful to other clients.
`This invention offers a method of intelligently offering
`and deleting categories from interest client database 28,
`intelligently offering information Within an interest category
`from information database 20 and judiciously gathering
`information for information database 20 for categories in
`complete category database 18. The folloWing examples Will
`clarify these activities.
`As one example, client database 28 for client 22 may
`contain bulletin board interest categories of baseball player
`statistics, baseball game statistics and articles on current
`
`10
`
`15
`
`25
`
`35
`
`45
`
`55
`
`65
`
`4
`baseball teams. Another category may be offered to client 22
`from server 24 of baseball stadium statistics. The decision to
`offer the category of baseball stadium statistics Will be
`determined by arti?cial intelligence.
`As another example, if a request for a prediction of What
`Will occur in the next Yankee game is posted to the baseball
`game statistics interest category by client 22 Which is run by
`server 24, server 24 may return a list of bulletin board
`ansWers prioritiZed based on Which ansWer Will most prob
`ably be best for client 22. The method by Which server 24
`prioritiZes the ansWers for client 22 Will occur Without
`human intervention.
`As another example, server 24 may judiciously gather
`pointers off the Internet of interesting articles on current
`baseball teams and store the pointers in information database
`20 for the current baseball teams category located in com
`plete category database 18. Client 22A may request three
`articles from the database. server 24 Would offer What it
`thinks is the best tWo articles for client 22A. Again, the
`gathering of information and the offering occurs Without
`human intervention.
`FIG. 2 shoWs a user interface for current interest catego
`ries 36 and suggested interest categories 34. Each interest
`category is a node of an interest category tree. Child
`expanded node 38 shoWs node 3 With one level of children.
`If a client decides to get more speci?c than his assigned
`category he can vieW beloW the category. Parent expanded
`node 40 shoWs node 3 With tWo levels of parent nodes. If the
`client Would like to vieW the more general categories above
`his assigned category he can vieW above his assigned
`category. As an example, node 3 may be backgammon. The
`three children of node 3 may be rules of backgammon, great
`backgammon players and backgammon strategy. The par
`ents of backgammon may be board games and the grand
`parent may be leisure activities. The client may be assigned
`the backgammon node and may choose to vieW What cat
`egories are above or beloW him.
`The information database 20 of FIG. 1 is located on the
`server. It consists of information units, as shoWn in FIG. 3,
`containing an information access area 50 to access the
`information. The information may be located locally on the
`server or may be located remotely. Remote location can be
`anyWhere outside of the given server. It contains a global
`information unit index 44 used by the client and the server
`to identify an information unit. It contains usefulness
`Weights 46 Which identify the usefulness of the information
`unit, access counts 48 Which record the number of times the
`information unit has been accessed and an information
`access area 50 Which is a method to get to the information
`or a pointer to a method to get to the information. The client
`?lls these areas as he uses the information unit. The server
`accumulates client values and combines them to produce
`server values Which are then stored in the information unit.
`There can be more than one count or Weight since an
`information unit Will have records of counts and Weights for
`the given piece of information and the categories and server
`leading to that piece of information.
`Referring noW to FIG. 4, the client database 28A is
`located on the client. It contains an ordered list of client
`interest categories 58. Information is kept on each category.
`Indexes of information accessed 60 re?ects information
`units Which have already been accessed from this category.
`The category utility Weight 62 is based on use and satisfac
`tion With the use of the category. Category use count 64 is
`a count of the number of times this category has been used.
`Not preferred information client Ids 66 and preferred infor
`
`Petitioner Apple Inc. - Exhibit 1005, p. 17
`
`

`

`US 6,195,654 B1
`
`5
`mation client Ids 68 re?ect encrypted Ids of other clients
`Whose information this client has used. Information units
`accessed 80 for this category re?ect all information accessed
`Within the past arbitrary period of time (for example the past
`month). These units are given to the server for potential use
`by other clients. Pro?le numbers 81 associate this client With
`the interest pro?le database 32 of FIG. 1. Previously rejected
`categories 82 are maintained to be given to the server before
`requesting neW categories of interest to this client.
`By use of these structures an intelligent information
`sharing system is built across a netWork. The interaction
`betWeen these structures alloWs for the transfer of useful
`information to meet a particular clients interests and needs.
`
`10
`
`PREFERRED EMBODIMENT—OPERATION
`
`15
`
`The netWorked information sharing model is shoWn in
`FIG. 5. Objects described in FIGS. 1, 2, 3 and 4 Will be
`referenced here and in subsequent paragraphs. In step 84 the
`shared information database 20 of FIG. 1 and the shared
`complete category database 18 of FIG. 1 are populated With
`information and cross references to that information. In step
`86 client interest categories 58 of FIG. 4, Which together are
`one pro?le, are assigned to each client; this pro?le of
`categories is matched to one or more generic pro?les in
`shared interest pro?le database 32 of FIG. 1. In step 88
`clients are Watched by the client machine; referring to FIG.
`4, information on server destinations 78, information items
`74 and categories chosen 76 are recorded on the client
`database 28A in the Information Units Accessed 80 table. In
`step 90 information and destinations accessed by the client
`are recorded With usefulness Weights 46 of FIG. 3 and access
`counts 48 of FIG. 3 Within each information unit on the
`server in shared complete category database 18 of FIG. 1. In
`step 92 the clients interest category pro?le assignments in
`shared interest pro?le database 32 of FIG. 1 are updated
`according to the current category pro?le information
`recorded in step 88. The client requests speci?c pieces of
`information, server or database destinations and neW cat
`egories in step 94; With the shared interest pro?le database
`32 of FIG. 1 updated in step 92 and the analysis done in step
`90, data is returned to the client.
`Referring back to FIG. 1, the server information database
`20 and the complete category database 18 is initially popu
`lated by standard information collection. On-line services
`have category directories and information databases. Many
`books can be found With categoriZed directories of on-line
`services. Thus, We begin With a good base of information
`database 20 cross referenced through complete category
`database 18. The complete category database 18 contains
`leaf nodes and non leaf nodes, each representing an interest
`category. Clients Will be assigned to interest category nodes.
`Referring noW to FIG. 4, this information Will be kept in the
`client database 28A as client interest categories 58. Clients
`Will also be assigned a set of pro?le numbers 81 Which Will
`associate them With interest pro?les located in the interest
`pro?le database 32 of FIG. 1. Pro?les and interest categories
`Will be initially assigned by use of an interest survey
`prepared by a marketing and/or psychology group. Sub
`groups Within an interest category can be formed Where
`interest in information Within a category Will differ substan
`tially betWeen one client and another. Subsequent updates
`Will be made dynamically by the intelligence Within the
`client and server computers as described beloW.
`The client machine Watches Where the client goes to
`access information and What information he accesses. Pri
`vacy is of course imperative. Perhaps, the client can turn on
`
`25
`
`35
`
`45
`
`55
`
`65
`
`6
`or off the Watch facility by pressing a button on the display.
`The client Would also have an encrypted and private client
`ID. Perhaps, at the end of an information search the client
`can press a share button to alloW the sharing of the infor
`mation With other clients.
`A script is kept Which When launched leads to the infor
`mation. The computer keeps track of hoW often the infor
`mation site and the nodes leading to the information site is
`accessed; it prompts the user at the end of the information
`for a usefulness Weight from excellent to poor. The computer
`keeps track of hoW often each point leading to the ?nal
`information is accessed. The script leading to the informa
`tion in the information access area 50 of FIG. 3, the access
`counts 48 of FIG. 3 and the usefulness Weights 46 of FIG.
`3 are kept Within the information unit structure Which is
`passed betWeen client database 28A of FIG. 4 and server
`complete category database 18 of FIG. 1. These information
`units point to the information database 20 of FIG. 1 through
`the information access area 50 of FIG. 3.
`FIG. 6 details the process of recording information access
`by a client. Let us call this client Client A. In step 96 client
`A accesses an information server, as an eXample the Hall of
`Malls server; in step 98 the client computer increments the
`count of hoW often client A has accessed the Hall of Malls
`server. In step 100 the client selects a category, as an
`eXample the Florida Mall; the computer increments the
`count for Florida Mall for client A in step 102. Client A goes
`to The Unusual Music Store subcategory in step 104; the
`computer increments the count for The Unusual Music Store
`in step 106. In step 108 the client chooses a resource, in this
`eXample a tape called The Nighf?y; the computer incre
`ments the count for this resource, the Nighf?y tape, in step
`110. In step 112 the client rates each node in this access: the
`Hall of Malls, the Florida Mall, the Unusual Music Store and
`The Nighffly. The client computer records this information
`in the usefulness Weights 46 of FIG. 3 and access counts 48
`of FIG. 3 of the information unit indeX 44 of FIG. 3.
`If client A has a history of giving good information to
`other clients, then the information unit is recorded by the
`server. FIG. 7 details this method. In step 114 the client
`connects to the server. In step 116 the server accesses the
`client usefulness indeX from enhancement database 30. This
`indeX is an historic measurement of hoW useful information
`offered by this client has been to other clients; it is based on
`the number of users accessing the information Who Were
`satis?ed With the information. In step 118 the server requests
`information from the client Which has a use count above a
`particular number and/or a rating above a particular number;
`the count and rating number are loWer for a high client
`usefulness indeX and higher for a loW client usefulness
`indeX. In step 120 the client sends the server the information
`With counts and ratings. In step 122 the client records this
`information as sent; subsequent uploads Will only resend this
`information if counts or ratings have changed substantially
`and Will only send the deltas of the information. In step 124
`the server records the information unit Within an ordered
`Weighted list.
`FIG. 8 gives a method for Weighting information received
`by a client. It uses a rating indeX of 1, 2 or 3. This method
`is for information Which has a count greater than or equal to
`one, as an eXample the Hall of Malls destination information
`unit. In step 126 the information With a use count is received.
`If, in step 128, the rating is 1 (the top rating) then in step 130,
`the Weight is assigned tWo times the count. If, in step 132,
`the rating is 2 then in step 134 the Weight is assigned 1.7
`times the count. If, in step 136, the rating is 3 then in step
`138 the Weight is assigned 1.5 times the count. If, in step
`
`Petitioner Apple Inc. - Exhibit 1005, p. 18
`
`

`

`US 6,195,654 B1
`
`7
`140, there is no rating then the Weight is assigned 1.7 times
`the count. In step 142 the Weight is assigned Whatever the
`current Weight is times the client data usefulness index. This
`method takes into account three factors: the historic useful
`ness of the clients data, the number of times the data Was
`reused and the rated satisfaction When using the information.
`Referring back to FIG. 1, If the information is accessed
`through a category in complete category database 18, the
`information is shared With clients in this category. If the
`information Was not accessed through a category it can be
`offered to clients With a similar interest pro?le accessed from
`the interest pro?le database 32; if the information is not
`categoriZed and turns out to be useful to a number of clients,
`a question can be asked of each client as to Which category
`it belongs in. If one or more categories consistently comes
`up, the information unit can be placed for access Within a
`gi