Ex Parte Casati et alDownload PDFBoard of Patent Appeals and InterferencesDec 6, 201010177423 (B.P.A.I. Dec. 6, 2010) Copy Citation UNITED STATES PATENT AND TRADEMARK OFFICE ____________ BEFORE THE BOARD OF PATENT APPEALS AND INTERFERENCES ____________ Ex parte FABIO CASATI MEHMET SAYAL, MARIA GUADALUPE CASTELLANOS and DIMITRIOS GUNOPULOS ____________ Appeal 2009-010990 Application 10/177,423 Technology Center 3600 ____________ Before: MURRIEL E. CRAWFORD, ANTON W. FETTING, and BIBHU R. MOHANTY, Administrative Patent Judges. CRAWFORD, Administrative Patent Judge. DECISION ON APPEAL Appeal 2009-010990 Application 10/177,423 2 STATEMENT OF THE CASE This is an appeal from the final rejection of claims 1-26. We have jurisdiction to review the case under 35 U.S.C. §§ 134 and 6 (2002). The claimed invention is directed to systems and methods for analyzing decision points in business processes (Spec. 1:12-13). Claim 1, reproduced below, is further illustrative of the claimed subject matter. 1. A method, comprising: accessing process execution data characterizing execution of one or more instances of a business process involving a set of one or more activities and a set of one or more decision points; and based upon the accessed process execution data, building a predictive quantitative model comprising a set of one or more correlation rules correlating context data associated with respective stages of the business process with possible outcomes of the business process from respective ones of the decision points in the business process. Claims 1, 13-15, and 24 stand rejected under 35 U.S.C. § 102(b) as anticipated by Alexander (US Pat. 5,182,793, iss. Jan. 26, 1993); claims 2-4 stand rejected under 35 U.S.C. § 103(a) as unpatentable over Alexander in view of Du (US Pat. 6,041,306, iss. Mar. 21, 2000); and claims 5-12, 16-23, 25, and 26 stand rejected under 35 U.S.C. § 103(a) as unpatentable over Alexander in view of Ullman (US Pat. 6,631,362 B1, iss. Oct. 7, 2003)1. We AFFIRM. 1 The rejection of claims 15-23 under 35 U.S.C. § 101 was withdrawn by the Examiner. (See Exam’r’s Ans. 2). Appeal 2009-010990 Application 10/177,423 3 ISSUES Did the Examiner err in asserting that claims 1, 13-15, and 24 are anticipated by Alexander? Did the Examiner err in asserting that claims 2-4 are unpatentable over Alexander in view of Du? FINDINGS OF FACT Alexander Alexander discloses that an example of an application of the apparatus is a computer network for monitoring an airline flight operation, in which the domain of the operation includes all events affecting flight schedules. As will be explained below, the system provides real time support for enabling a user to comprehend the scope of a problem, observe the details of problem side effects, generate multiple possibilities for improving the situation, and evaluate alternatives for improvement. The use of this specific application is exemplary, and the apparatus and method of this invention could be used for any number of other applications in many diverse areas (col. 3, ll. 13-24). Decision processor system 60 includes two subsystems, including an effects processor 62 and a strategy processor 66, each associated with special programming to accomplish the functions described below. The hardware associated with each processor may be any one of a number of well-known devices capable of executing computer instructions, such as a microprocessor (col. 4, ll. 33-39). Effects processor 62 embodies the concept of a “‘symbolic spreadsheet.’” The symbolic spreadsheet is analogous to numeric Appeal 2009-010990 Application 10/177,423 4 spreadsheets in common use today, in which numbers or formulas are assigned to cells. The user can change a value in one cell and immediately see the effect on values of other cells. In contrast to numeric spreadsheets, effects processor 62 uses frames as cells to symbolically represent components of a particular domain. Values, complex entities, or even programs may be assigned to cells. The frames are linked by means of descriptions of their relationships. When a change in the domain occurs or when a hypothetical event is proposed by a user, the programming determines the effect of those changes on other aspects of the operation. A term used to refer to this type of programming is “‘constraint propagation’” (col. 4, ll. 40-55). For the airline flight operations example, cells represent objects, such as airplanes, crews, and airports, and events that affect them, such as flight delays, airport closings, or maintenance delays. When the system receives an event, effects processor 62 determines its effects and updates knowledge base 50 (col. 4, ll. 56-61). A second subsystem of decision processing system 60 is the strategy processor 66. The programming of strategy processor 66 permits several basic functions: solution of “‘best choice’” problems using a particular method, variation of methods, and variation of strategies (col. 5, ll. 3-8). Figures 3-7 illustrate another aspect of the invention, which is a method for using a computer to assist a user in selecting a best choice among alternative choices when making a decision. Essentially, Figures 3, 4A, and 4B illustrate the method of making a best choice, given a particular choice method. Figures 5 and 6 illustrate the method of the invention using strategy selection and learning features. Figure 7 illustrates a method of the Appeal 2009-010990 Application 10/177,423 5 invention using a scenario feature. These methods can be implemented with the functions and data types discussed in connection with Figure 2 (col. 8, ll. 56-67). As shown in Figure 5, using the invention with the strategy selection features involves a first step, Step 510, of adopting a strategy. When a strategy is adopted, the parameter values and choice methods of that strategy become the current values and choice methods. Step 512 is executing the decision making steps of Figure 3 to choose a best candidate. Step 514 determines if the best candidate is acceptable. If it is, Step 516 is using the best candidate in other programming or as desired by the user (col. 12, ll. 59- 68). Steps 518 and 520 of Figure 5 illustrate another feature of the invention, which may be used during the development stage of the decision making system. This is a learning feature, which is invoked by the user after a decision making operation has failed to produce the desired result. In general, the learning feature is a rule weight adjusting feature in connection with the weighted choose method, so that the candidate who is the user's best choice will receive the highest score, and thereby improve the knowledge base. The learning steps are iterative for each “‘rule instance,’” i.e., each part of the weighted choose method in which a particular rule was evaluated for the entire set of candidates surviving elimination. A rule instance is the rule weight and a list of candidate/rank weight pairs (col. 13, ll. 1-15). Step 518 determines whether this learning feature is to be invoked. If so, Step 520 is performing it. The substeps of Step 520 are illustrated in Figure 6. It is assumed that the programming has already saved each candidate and its state information, including the total score for each Appeal 2009-010990 Application 10/177,423 6 candidate, and each rule instance. Step 610 is getting a rule instance. Step 612 ensures that the learning steps continue for each rule instance (col. 13, ll. 16-23). Step 614 is determining the relationship between the current rule and the current candidate's total, T. For each rule, there are three status categories: (1) every candidate that has a higher current T than the desired candidate is ranked higher by the current rule, or (2) every candidate that has a higher current T than the desired candidate is ranked lower by the current rule, or (3) neither category (1) nor (2) exists. According to the category in which a rule falls, the remaining steps increase or decrease that rule's weight or set the rule aside (col. 13, ll. 24-34). If the rule falls within category 3, Step 618 sets the rule aside. Steps 620-626 ensures that if the rule is set aside, after any rule that falls within category (1) or (2) has had its weight adjusted, the set aside rules can be looked at again to determine if any one of them now fall within category (1) or (2) (col. 13, ll. 35-40). Du Du discloses that organizations are currently engaging in workflow process re-engineering in many domains, including financial services, telecommunications services, healthcare services, customer order fulfillment, manufacturing procedure automation, and electronic commerce (col. 1, ll. 34-39). Figure 3 shows, by way of example, a workflow process 18 which is represented as a directed graph 40 consisting of a set of nodes connected by arcs as displayed on the HP OpenPM user interface (col. 6, ll. 17-20). Appeal 2009-010990 Application 10/177,423 7 Associated with each workflow process 18, there is a process data template defined by a workflow process designer module 22a (shown in Figure 2). The process data template is used to provide initial data for the creation of process instances. At run time, based on the process data template and read/write lists of activities defined in a workflow process 18, HP OpenPM will generate a case packet for each process instance to facilitate data passing between activities and the HP OpenPM engine 20 (col. 7, ll. 25-34). Ullman Ullman discloses in Figure 12 that the box labeled “‘Decision’” 52 takes as values the alternatives for resolving the issue represented by the diagram. The circle labeled S(Cc│A a) 53 represents the satisfaction of criterion Cc given alternative A a and will be called a satisfaction node. While we show only one, there can be one for each alternative/criterion combination. This is nonstandard, as a usual influence diagram would have only one node for each criterion. We use contingent nodes to represent each alternative/criterion pair separately, since we may not have arguments for every pair. The pair of two-node chains 54 hanging from S(Cc│A a) represents opinions posted by decision-makers. Two are shown here, but there can be any number of such chains hanging from each of the S(Cc│A a) satisfaction nodes. The higher of the two circles 55 represents the state of participant knowledge about the ability of the alternative to meet the criterion, and the lower circle 56 is a diagram artifact used to encode probabilistic evidence (col. 5, ll. 20-37). Appeal 2009-010990 Application 10/177,423 8 ANALYSIS Claims 1, 13-15, and 24 We are not persuaded that the Examiner erred in asserting that claims 1, 13-15, and 24 are anticipated by Alexander (App. Br. 11-22). By saying that claims 13-15 and 24 are patentable for the same reason as independent claim 1, Appellants have de facto chosen independent claim 1 as being representative of the subject matter of claims 1, 13-15, and 24 (App. Br. 19- 22; Reply Br. 7-8). We do the same. Appellants assert that Alexander does not disclose a method that operates with respect to a single business process, as recited in independent claim 1 (App. Br. 14-15, 18; Reply Br. 2-3). First of all, Alexander discloses methods applied to flight operations and decision selection (col. 3, ll. 13-16; col. 4, ll. 56-59; col. 5, ll. 3-8; col. 8, ll. 56-59). These are business processes under a broadest reasonable interpretation. See In re Am. Acad. of Sci. Tech. Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004). Moreover, Alexander also discloses that the “specific application is exemplary, and the apparatus and method of this invention could be used for any number of other applications in many diverse areas” (col. 3, ll. 21-24). Furthermore, the fact that different structures in Alexander may perform different portions of the business process does not mean the different portions are not a part of the same “single” business process. And in any case, the requirement of a single business process performed by a single structure is not set forth in the claim. See CollegeNet, Inc. v. ApplyYourself, Inc., 418 F.3d 1225, 1231 (Fed. Cir. 2005). Appellants assert that Alexander does not disclose a predictive quantitative model, as recited in independent claim 1 (App. Br. 15, 17; Appeal 2009-010990 Application 10/177,423 9 Reply Br. 3-4). However, the Examiner has cited numerous portions of Alexander which disclose that inputs are used to make predictions on changes and effects, for example, with respect to airline operations (col. 4, ll. 50-54, 59-68; col. 5, ll. 3-8). Furthermore, choice selection, as disclosed in Alexander, is also a prediction of the best scenario (col. 8, ll. 56-59; col. 10, ll. 27-30; col. 12, ll. 53-56). Appellants assert that Alexander does not disclose a set of one or more correlation rules correlating context data associated with respective stages of the business process with possible outcomes of the business process from respective ones of the decision points in the business process[,] as recited in independent claim 1 (App. Br. 16, 18; Reply Br. 4-5). As an initial matter, we note that altering an input resulting in an altered output has a perfect correlation (i.e., correlation coefficient of one) under a broadest reasonable interpretation. Along those lines, Alexander discloses that “[w]hen a change in the domain occurs or when a hypothetical event is proposed by a user, the programming determines the effect of those changes on other aspects of the operation” (col. 4, ll. 50-54, 59-61). Thus, the “hypothetical event” is “predictive,” and the “changes on other aspects of the operation” in Alexander corresponds to the recited “possible outcomes of the business process from respective ones of the decision points in the business process.” Appellants assert that the weights and links for inputs are pre- programmed, and thus are not capable of being dynamically changed “based upon the accesses process execution data,” as recited in independent claim 1 (App. Br. 16, 18; Reply Br. 5-7). However, Alexander discloses that Appeal 2009-010990 Application 10/177,423 10 “[Figures] 5 and 6 illustrate the [] method of the invention using strategy selection and learning features” (col. 8, ll. 61-63; col. 12, ll. 59-63; col. 13, ll. 1-40). Specifically, Alexander discloses a learning feature that adjusts the weights, which dynamically changes the correlations based on data, as claimed. Claims 2-4 We are not persuaded that the Examiner erred in asserting that claims 2-4 are unpatentable over Alexander in view of Du (App. Br. 24-27; Reply Br. 9-10). By saying that claims 3 and 4 are patentable for the same reason as claim 2, Appellants have de facto chosen claim 2 as being representative of the subject matter of claims 2-4 (App. Br. 27; Reply Br. 10). We do the same. Appellants assert that Du does not disclose “partitioning the business process into the stages,” as recited in dependent claim 2 (App. Br. 25). However, the nodes and process instances disclosed in Du correspond to the claimed stages (col. 6, ll. 17-20; col. 7, ll. 25-34). Appellants assert that there is no reason to combine Alexander and Du (App. Br. 25-27; Reply Br. 9-10). However, Du discloses that “[o]rganizations are currently engaging in workflow process re-engineering in many domains,” of which creating nodes and process instances are stages (col. 1, ll. 35-39). Thus, one of ordinary skill would implement the stages of Du in Alexander to re-engineer Alexander, as the Examiner puts it, “to provide a more accurate description of the procedural knowledge and a more efficient workflow analysis,” as desired by Alexander (col. 1, ll. 33-38). Appeal 2009-010990 Application 10/177,423 11 Claims 5-12, 16-23, 25, and 26 We are not persuaded that the Examiner erred in asserting that claims 5-12, 16-23, 25 and 26 are unpatentable over Alexander in view of Ullman (App. Br. 27-33; Reply Br. 10-13). By saying that claims 6-12, 16-23, 25, and 26 are patentable for the same reason as claim 5, Appellants have de facto chosen claim 5 as being representative of the subject matter of claims 6-12, 16-23, 25, and 26 (App. Br. 31-33; Reply Br. 12-13). We do the same. Appellants assert that Ullman does not disclose that each of the correlation rules specifies respective probabilities of the possible outcomes of the business process from a respective one of the decision points conditioned on completion of all activities of the business process up to a respective one of the stages of the business process and a respective context data value[,] as recited in claim 5 (App. Br. 29-30; Reply Br. 11-12). However, Ullman discloses multiple satisfaction nodes at different points in a process dependent on different alternatives/criterion (col. 5, ll. 20-37). The content of the satisfaction nodes corresponds to the recited possible outcomes under a broadest reasonable interpretation. See In re Am. Acad. of Sci. Tech. Ctr., 367 F.3d at 1364. The decision of the Examiner to reject claims 1, 13-15, and 24 under 35 U.S.C. § 102(b) is AFFIRMED. The decision of the Examiner to reject claims 2-4 under 35 U.S.C. § 103(a) is AFFIRMED. The decision of the Examiner to reject claims 5-12, 16-23, 25, and 26 under 35 U.S.C. § 103(a) is AFFIRMED. Appeal 2009-010990 Application 10/177,423 12 No time period for taking any subsequent action in connection with this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R. § 1.136(a)(1)(iv) (2007). AFFIRMED hh HEWLETT-PACKARD COMPANY Intellectual Property Administration 3404 E. Harmony Road Mail Stop 35 FORT COLLINS, CO 80528 Copy with citationCopy as parenthetical citation