UNITED STATES PATENT AND TRADEMARK OFFICE
`
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________
`
`Oracle Corporation,
`
`Petitioner,
`
`v.
`
`Clouding IP, LLC
`
`Patent Owner.
`
`____________
`
`IPR2013- _____
`
`Patent 5,944,839
`
`____________
`
`DECLARATION OF PROFESSOR TODD C. MOWRY, PH.D.
`
`
`
`
`1
`
`
`
`
`
`
`
`ORACLE EX. 1007, p. 1
`
`
`
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________
`
`Oracle Corporation,
`
`Petitioner,
`
`v.
`
`Clouding IP, LLC
`
`Patent Owner.
`
`____________
`
`IPR2013- _____
`
`Patent 5,944,839
`
`____________
`
`DECLARATION OF PROFESSOR TODD C. MOWRY, PH.D.
`
`
`
`
`1
`
`
`
`
`
`
`
`ORACLE EX. 1007, p. 1
`
`
`
`I, Prof. Todd C. Mowry, Ph.D., declare as follows:
`
`
`
`I.
`
`
`Background and Qualifications
`
`(1) My name is Todd Mowry. I am a Professor at Carnegie Mellon
`
`University in the Computer Science Department. I have studied and practiced in
`
`the field of computer science for almost 20 years, and have been a professor of
`
`computer science since 1993.
`
`(2)
`
`I received my Doctor of Philosophy (Ph.D.) degree in the field of
`
`Electrical Engineering from Stanford University in 1994. I received my Masters of
`
`Science (M.S.) degree in Electrical Engineering from Stanford University and my
`
`Bachelor of Science (B.S.) degree in Electrical Engineering from the University of
`
`Virginia.
`
`(3) Upon receiving my Ph.D. degree, I joined the faculty of the University
`
`of Toronto in the Department of Electrical and Computer Engineering and the
`
`Department of Computer Science as an Assistant Professor. I relocated and was
`
`promoted to the rank of Associate Professor (initially without tenure) at Carnegie
`
`Mellon University in 1997, was promoted to tenured Associate Professor in 2002,
`
`and I was promoted to the rank of full Professor in 2008. I was the Associate
`
`Department Head for Faculty of the Computer Science Department from 2009-
`
`2010.
`
`
`
`2
`
`ORACLE EX. 1007, p. 2
`
`
`
`(4)
`
`Since becoming a faculty member in 1993, I supervised the research
`
`of 11 Ph.D. dissertations in the field of computer science, and along with my
`
`graduate students, published over 60 technical publications in scientific journals or
`
`conferences in the field of computer science and 20 technical reports published at
`
`Carnegie Mellon University and the University of Toronto. In addition to Ph.D.
`
`dissertations, I have also supervised several graduate student’s Master’s theses.
`
`(5) As part of my research, I focus on monitoring computer systems as
`
`they run to diagnose bugs and other functionality problems. In fact, I have
`
`published a number of papers on this topic.
`
`(6)
`
`I am a member of several professional organizations including the
`
`Institute of Electrical and Electronics Engineers (IEEE) and the Association of
`
`Computing Machinery (ACM). I am an Associate Editor of ACM Transactions on
`
`Computer Systems (TOCS). I received a Sloan Research Fellowship and the TR35
`
`Award from MIT's Technology Review.
`
`(7)
`
`I have served as a consultant for Intel Corporation, Silicon Graphics,
`
`Inc., SandCraft, Inc., and IBM.
`
`(8) A copy of my latest curriculum vitae, which describes in further detail
`
`my qualifications, responsibilities, employment history, honors, awards,
`
`professional associations, distinguished lectures, and publications is attached to
`
`this declaration.
`
`
`
`3
`
`ORACLE EX. 1007, p. 3
`
`
`
`(9)
`
`I have reviewed U.S. Patent No. 5,944,839 (“the ‘839 patent,” Ex.
`
`1001) to Henri J. Isenberg. I have also reviewed the publications cited in the
`
`footnotes of this declaration and referenced in the inter partes review petition
`
`submitted herewith.
`
`
`II. The Person of Ordinary Skill in the Relevant Field in the Relevant
`Timeframe
`
`
`
`(10) I have been informed that “a person of ordinary skill in the relevant
`
`field” is a hypothetical person to whom an expert in the relevant field could assign
`
`a routine task with reasonable confidence that the task would be successfully
`
`carried out. I have been informed that the level of skill in the art is evidenced by
`
`prior art references. The prior art discussed herein demonstrates that a person of
`
`ordinary skill in the art, at the time the ‘839 patent was filed, was aware of several
`
`different computer maintenance tools that used a set of sensors in combination
`
`with a case base type knowledge database to diagnose and solve computer
`
`problems.
`
`(11) Based on my experience, I have an understanding of the capabilities
`
`of a person of ordinary skill in the relevant field. I have supervised and directed
`
`many such persons over the course of my career.
`
`
`
`4
`
`
`
`
`
`ORACLE EX. 1007, p. 4
`
`
`
`III. State of the Art as of 1997
`
`
`(12) Case-based reasoning (CBR) is an AI problem-solving technique that
`
`originated with Roger Schank and his Ph.D. students in the mid-1980s at Yale
`
`University. According to Schank, “A case-based reasoner solves new problems by
`
`adapting solutions that were used to solve old problems.”1 The key word in this
`
`quote is “adapting.” In contrast with rule-based reasoning, which performs a
`
`scripted action for a rule whenever the specific conditional test for that rule is
`
`satisfied, the motivation behind case-based reasoning is to take a more flexible and
`
`adaptive approach to problem-solving that draws upon analogies to earlier
`
`solutions of related (but somewhat different) problems. The proponents of case-
`
`based reasoning argue that drawing upon analogies to solve problems corresponds
`
`well to how humans solve problems, i.e., by recalling situations that remind them
`
`of their current problem, and by attempting to adapt the previous solution to the
`
`current circumstances.
`
`(13) Rather than storing a set of “rules,” a case-based reasoning system
`
`stores a set of “cases” to capture its previous experiences. Cases are descriptions
`
`of diagnostic situations that typically include a set of symptoms, a description of
`
`the failure and its cause, and also a repair strategy for fixing the problem.
`
`
`1 Riesbeck, C. K., et al. Inside Case-Based Reasoning. Hillsdale, NJ: L. Erlbaum Assoc. Inc.,
`1989
`
`
`
`
`5
`
`ORACLE EX. 1007, p. 5
`
`
`
`(14) In addition to the use of analogies, another key feature of case-based
`
`reasoning is that it adapts its set of cases over time based upon the success or
`
`failure of its attempts to solve problems. For example, if an existing case can be
`
`successfully adapted to solve a new problem, the description of that case may be
`
`further generalized. In contrast, if adapting existing cases fails to solve the
`
`problem, then further specialization may be required, including possibly creating a
`
`new case. This learning process where the set of cases are adapted over time is
`
`fundamental to case-based reasoning.
`
`(15) By 1997, case-based reasoning had become a mature research area
`
`that was being actively explored by dozens of research groups around the world.
`
`In fact, there was so much published work on CBR by the early 1990’s that
`
`multiple survey articles were published on this topic: Kolodner2 and Aamodt.3 The
`
`latter of these two articles cites 75 papers on CBR. The Aamodt survey paper also
`
`describes “The CBR Cycle” in Section 3.3 and Figure 1, which is the basic
`
`structure of nearly all CBR systems, and which has been cited numerous times.
`
`
`2 Kolodner, J. L. “An introduction to case-based reasoning.” Artificial Intelligence Review,
`6(1):3–34, March 1992
`3 Aamodt, A., et al. “Case-based reasoning: Foundational issues, methodological variations, and
`system approaches.” AI Communications, 7(1):39-59, March 1994
`
`
`
`6
`
`ORACLE EX. 1007, p. 6
`
`
`
`
`
`(16) Conferences and workshops were also being created that were
`
`devoted entirely to CBR: the 1st European Workshop on CBR (EWCBR)4 began in
`
`1993, and the 1st International Conference on CBR (ICCBR)5 began in 1995.
`
`
`IV. The ‘839 Patent
`
`
`(17) The ‘839 patent is generally directed to a tool for performing
`
`automatic maintenance, i.e., trouble-shooting, of a computer system. In order to
`
`automatically trouble-shoot the computer system, the maintenance tool collects
`
`information about the behavior of the computer system it is monitoring using
`
`4 Wess, S., et al. Topics in case-based reasoning: First European Workshop, EWCBR-93,
`Kaiserslautern, Germany, November 1-5, 1993, Selected Papers. London, UK: Springer, 1994
`
` Veloso, M. M., et al. Case-Based Reasoning Research and Development: First International
`Conference, ICCBR-95, Sesimbra, Portugal, October 23-26, 1995, Proceedings. London, UK:
`Springer, 1995
`
` 5
`
`
`
`7
`
`ORACLE EX. 1007, p. 7
`
`
`
`software routines called “sensors,” and then diagnoses the problem and
`
`recommends a likely solution through the combination of an artificial intelligence
`
`(“AI”) engine and a “knowledge database.” (Ex. 1001 at 1:57-59).
`
`(18) The features of the trouble-shooting tool described in the ‘839 patent
`
`correspond in a straightforward way to conventional case-based reasoning systems
`
`at that time. In particular, after one of the sensors in the ‘839 patent detects a
`
`problem, it activates the AI engine. (Ex. 1001 at 1:61-64). The AI engine then
`
`attempts to draw analogies between the current problem and the set of cases that it
`
`has stored in its knowledge database to see whether a previous case suggests a
`
`likely solution to the current problem related to trouble-shooting the computer
`
`system. (Id. at 1:65 – 2:1).
`
`(19) In the course of attempting to diagnose the problem, the AI engine
`
`may determine, based upon information stored in matching cases, that additional
`
`information is required to accurately diagnose the problem, in which case it may
`
`activate additional sensors to collect more information about the monitored system.
`
`(Ex. 1001 at 2:5-8). If a likely solution to the problem is determined, then the AI
`
`engine will attempt to perform the repair. (Id. at 2:9-11).
`
`
`
`
`
`
`
`
`
`8
`
`ORACLE EX. 1007, p. 8
`
`
`
`V. Claim Interpretation
`
`(20) In proceedings before the USPTO, I understand that the claims of an
`
`unexpired patent are to be given their broadest reasonable construction in view of
`
`the specification from the perspective of one skilled in the art. I have been
`
`informed that the ‘839 patent has not expired. In comparing the claims of the ‘839
`
`patent to the known prior art, I have carefully considered the ‘839 patent and the
`
`‘839 file history based upon my experience and knowledge in the relevant field. In
`
`my opinion, the claim terms of the ‘839 patent are used in their ordinary and
`
`customary sense as one skilled in the relevant field would understand them except
`
`for those terms specifically addressed in the following paragraphs.
`
`“Sensors”
`
`(21) The ‘839 patent describes the sensors as follows: “[t]he sensors 112
`
`are software programs that gather information from the computer system 300. (See,
`
`e.g., Ex. 1001 at 3:16-17).”
`
`(22) Accordingly, under the broadest reasonable construction, the term
`
`“sensors” should be interpreted as including different aspects of the same software
`
`program or different components of the same application.
`
`“Artificial Intelligence (AI) Engine”
`
`(23) The term “AI engine” should be interpreted as including, under the
`
`broadest reasonable construction, a different aspect of the same software program
`
`
`
`9
`
`ORACLE EX. 1007, p. 9
`
`
`
`as the “sensors” discussed above in paragraph 22. Both the sensors and the AI
`
`engine are different components of the same software program or different
`
`components of the same application.
`
`(24) Although the prior art references discussed below, such as Gurer,
`
`Allen ‘218, and Barnett, may describe the “sensors” as being separate components
`
`from the “AI engine,” both the sensors and the AI engine are pieces of software,
`
`and the real distinction between them is functionality. The sensors interact with
`
`the system that they are monitoring to collect information, and the AI engine
`
`reasons about the likely diagnosis of the problem and the likely solution. Hence,
`
`under the broadest reasonable construction, it would be obvious to a skilled artisan
`
`that the software that performs the sensor functionality and the software that
`
`performs the AI engine functionality might be integrated within the same larger
`
`software application such that the distinction in functionality between the sensors
`
`and the AI engine is no longer existent.
`
`
`VI. Discussion of Relevant Patents and Articles
`
`GURER
`
`(25) An Artificial Intelligence Approach to Network Fault Management
`
`published by Gurer et al. (“Gurer,” Ex. 1003) describes a system that uses case-
`
`based reasoning to automatically diagnose and correct faults in a computer network
`
`that it is monitoring. The raw input to the system is a set of “alarms” which are
`
`
`
`10
`
`ORACLE EX. 1007, p. 10
`
`
`
`produced by either the element manager software on a particular network element
`
`(e.g., an ATM switch) when it notices a hard error (e.g., a link is down), or through
`
`software that performs statistical analysis of the network when it notices a
`
`statistical error (e.g., performance degradation due to congestion). (Id. at 1:30-34).
`
`Since a given network problem might trigger a large number of alarms, these raw
`
`alarms are first passed through either a Neural Network or a Bayesian Belief
`
`Network to filter and correlate the alarms before they are passed into the case-
`
`based reasoning system, thereby reducing the amount of noise in the input before
`
`fault identification begins. (Id. at 2:1-4). Figure 1 of Gurer illustrates the fault
`
`management process described above. (Id. at 2).
`
`
`
`
`
`11
`
`ORACLE EX. 1007, p. 11
`
`
`
`
`
`
`
`(26) As illustrated in Figure 3 below, Gurer further describes how the case-
`
`based reasoning system uses its library of cases, i.e., knowledge database, to
`
`determine a likely solution to the problem, which potentially involves deciding that
`
`the sensors should collect more information regarding the state of the network, and
`
`how the gathered information is stored in the knowledge database in the form of
`
`new cases. (Ex. 1003 at 7:1-10).
`
`
`
`12
`
`ORACLE EX. 1007, p. 12
`
`
`
`
`
`(27) Gurer discloses that “[a]utomation of network management activities
`
`can benefit from the use of artificial intelligence (AI) technologies, including fault
`
`management, performance analysis, and traffic management. Here we focus on
`
`fault management, where the goal is to proactively diagnose the cause of abnormal
`
`network behavior and to propose, and if possible, take corrective actions.” (See Ex.
`
`1003 at 1:11-14).
`
`(28) In the words of Gurer, “Today’s high speed, heterogeneous networks
`
`represent a complex and data intensive environment . . .” (Ex. 1003 at 1:10). The
`
`network elements in those types of sophisticated networks included processors and
`
`memory for handling the complex and data-intensive nature of the network.
`
`Processors were needed to execute software to manage the network elements, and
`
`memory was needed both to execute the software and to store the network packets.
`
`In addition, the case-based reasoning system that is described by Gurer would
`
`
`
`13
`
`ORACLE EX. 1007, p. 13
`
`
`
`necessarily be implemented as a large and sophisticated piece of software. In order
`
`to execute the software, a processor is necessary in order to perform the
`
`instructions in the software. In addition, both the software instructions and any
`
`data values that are accessed by the software must be stored in memory. Hence,
`
`consistent with the preamble of claim 1 of the ‘839 patent reciting “[a] tool for
`
`automatically maintaining a computer system having a processor and a memory,”
`
`it would be clear to a person of ordinary skill in the art that the system described by
`
`Gurer requires a computer system containing both a processor and a memory.
`
`(29) Further, Gurer discloses a knowledge database (“case library”)
`
`holding a plurality of cases describing potential computer problems and
`
`corresponding likely solutions. “Case-based reasoning is based on the premise that
`
`situations recur with regularity. Studies of experts and their problem solving
`
`techniques have found that experts rely quite strongly on applying their previous
`
`experiences to the current problem at hand. CBR can be thought of as such an
`
`expert that applies previous experiences stored as cases in a case library. Thus, the
`
`problem-solving process becomes one of recalling old experiences and interpreting
`
`the new situation in terms of those old experiences.” (Ex. 1003 at 6:35-39).
`
`(30) The structure of case-based reasoning (CBR) systems was well-known
`
`at the time of the Gurer invention. As further described by Gurer, case-based
`
`reasoning is “[a] symbolic AI technology” that operates on a case library. (Ex.
`
`
`
`14
`
`ORACLE EX. 1007, p. 14
`
`
`
`1003 at 6:24). Hence, consistent with claim 1 of the ‘839 patent which recites, in
`
`part, “a knowledge database stored in the memory and holding a plurality of cases
`
`describing potential computer problems and corresponding likely solutions,” it
`
`would be clear to a person of ordinary skill in the art that the software necessary to
`
`implement this symbolic AI technology, i.e., case-based reasoning, must be stored
`
`in the memory of a computer system, and also its knowledge database, i.e., the case
`
`library, must also be stored in the memory of the computer system.
`
`(31) Gurer also discloses a plurality of sensors, or alarms, adapted for
`
`gathering data about the computer system, storing the data in the knowledge
`
`database, and detecting whether a computer problem exists from the data and the
`
`plurality of cases. “The first step in fault management is to collect monitoring and
`
`performance alarms. Typically alarms are produced by either managed network
`
`elements (e.g., ATM switches, customer premise equipment) or by a statistical
`
`analysis of the network that monitors trends and threshold crossings. Alarms can
`
`be classified into two categories, physical and logical, where physical alarms are
`
`hard errors (e.g., a link is down), typically reported through an element manager,
`
`and logical alarms are statistical errors (e.g., performance degradation due to
`
`congestion).” (Ex. 1003 at 1:30-34).
`
`(32) In addition, “[a]larm filtering is a process that analyzes the multitude
`
`of alarms received and eliminates the redundant alarms (e.g., multiple occurrences
`
`
`
`15
`
`ORACLE EX. 1007, p. 15
`
`
`
`of the same alarm). Alarm correlation is the interpretation of multiple alarms such
`
`that new conceptual meanings can be assigned to the alarms, creating derived
`
`alarms. Faults are identified by analyzing the filtered and correlated alarms and by
`
`requesting tests and status updates from the element managers, which provide
`
`additional information for diagnosis.” (Ex. 1003 at 2:1-5).
`
`(33) “The more complex processes of fault management include alarm
`
`filtering and correlation, fault identification, and correction. Many of these
`
`functions involve analysis, correlation, pattern recognition, clustering or
`
`categorization, problem solving, planning, and interpreting data from a knowledge
`
`base that contains descriptions of network elements and topology.” (Ex. 1003 at
`
`3:2-4).
`
`(34) Figure 1 of Gurer illustrated above in paragraph 26 clearly refers to
`
`multiple sensors, including those that collect both “physical alarms” and “logical
`
`alarms.” (Ex. 1003 at 2). These alarms are gathered by pieces of software that are
`
`either (i) the “element manager” for a network element which reports “physical
`
`alarms” or (ii) the statistical analysis software that detects “logical alarms.” (Id. at
`
`1:30-34). Hence, consistent with claim 1 of the ‘839 patent which recites, in part,
`
`“a plurality of sensors stored in the memory and executing on the processor and
`
`adapted for gathering data about the computer system, storing the data in the
`
`knowledge database, and detecting whether a computer problem exists from the
`
`
`
`16
`
`ORACLE EX. 1007, p. 16
`
`
`
`data and the plurality of cases,” it would be clear to a person of ordinary skill in the
`
`art that the instructions for executing these sensors would be stored in a memory
`
`and would execute on a processor, since this is how software is executed. It is also
`
`clear that the collected alarms must be stored somewhere in the memory, so that
`
`they can be passed into the filtering and correlating mechanism and on to the case-
`
`based reasoning system. Moreover, Gurer also talks about storing the results of
`
`tests, i.e., when the CBR system decides that additional information is needed from
`
`a sensor in order to diagnose the problem, in the knowledge database in the form of
`
`a new case: “[t]he value of the tests (i.e., useful, not useful), the steps taken, any
`
`circuitous paths or dead-ends taken, and the success of the analysis should be
`
`stored into a new case. This information, in addition to contextual information,
`
`comprise a new case which is then indexed into the case library.” (Id. at 8:6-8).
`
`(35) Gurer also discloses an AI system executing in response to detection
`
`of a computer problem. Gurer discloses “[a]utomation of network management
`
`activities can benefit from the use of artificial intelligence (AI) technologies,
`
`including fault management, performance analysis, and traffic management. Here
`
`we focus on fault management, where the goal is to proactively diagnose the cause
`
`of abnormal network behavior and to propose, and if possible, take corrective
`
`actions . . . AI technologies may be used to automate the fault management
`
`
`
`17
`
`ORACLE EX. 1007, p. 17
`
`
`
`process, in particular neural networks (NNs) and case-based reasoning (CBR).”
`
`(Ex. 1003 at 1:11-16).
`
`(36) Further, Gurer discloses that “[a]nother area of fault management
`
`where AI technologies can have a positive impact, is fault correction. CBR
`
`systems, ESs [Expert Systems], or intelligent planning systems can develop plans
`
`or courses of action that will correct a fault that has been identified and verified.”
`
`(Ex. 1003 at 3:27-29).
`
`(37) As discussed above in paragraph 31, Gurer discloses that case-based
`
`reasoning (CBR) is “[a] symbolic AI technology.” (Ex. 1003 at 6:24). At the time
`
`the invention was made, it was well known that symbolic artificial intelligence
`
`(AI) technology and case-based reasoning, in particular, was implemented through
`
`software that executed on a computer system, where the computer system would
`
`have at least a memory and a processor. Hence, consistent with claim 1 of the ‘839
`
`patent which recites, in part, “an AI engine stored in the memory and executing on
`
`the processor in response to detection of a computer problem and [the AI engine]
`
`utilizing the plurality of cases to determine a likely solution to the detected
`
`computer problem,” it would be clear to a person of ordinary skill in the art that the
`
`AI engine would necessarily be stored in a memory of a computer system.
`
`(38) Gurer also discloses that “CBR problem solving can be depicted as a
`
`five-step process, as shown in Figure 3[, illustrated above in paragraph 27]:
`
`
`
`18
`
`ORACLE EX. 1007, p. 18
`
`
`
`(1) retrieval, (2) interpretation and adaptation, (3) evaluation and repair,
`
`(4) implementation, and (5) evaluation and learning. The first step is retrieving
`
`cases that best match the current situation or case. Thus it is crucial to use an
`
`appropriate indexing method, such as decision trees or nearest neighbor matching.
`
`Once a case is retrieved, it must be interpreted and then adapted. The
`
`interpretation process is a simple comparison between the retrieved cases and the
`
`current case. Adaptation is a complicated, domain-dependent process that uses
`
`rules to adapt the current case to the problem situation and propose an initial
`
`solution, based on the similarities and differences . . . ” (Ex. 1003 at 7:1-6).
`
`(39) Further, the ‘839 patent discloses that “[q]uestions belong to one or
`
`four categories: Yes/No; Numeric; Text; and List. Yes/No questions are those that
`
`have Yes or No answers. Numeric questions are those having answers that are
`
`integers. Text questions have textual answers. Finally, List questions have
`
`answers selected from a list of legal answers.” (See Ex. 1001 at 3:52-57).
`
`(40) Hence, consistent with the disclosure of the ‘839 patent and claim 2
`
`which recites, in part, “wherein each case comprises: at least one question asking
`
`about a particular aspect of the computer system that can be answered by the data
`
`gathered by the plurality of sensors,” it would be clear to a person of ordinary skill
`
`in the art that the “decision trees” used in the CBR problem solving process of
`
`
`
`19
`
`ORACLE EX. 1007, p. 19
`
`
`
`Gurer would include at least one question about a particular aspect of the computer
`
`system.
`
`(41) Claim 6 of the ‘839 patent is similar to claim 1 discussed above
`
`except that it is broader in scope. Claim 6 is not limited to a knowledge database
`
`holding a plurality of cases. Accordingly, it would be clear to a person of ordinary
`
`skill in the art that the discussion of claim 1 applies correspondingly to claim 6.
`
`(42) Consistent with claim 8 of the ‘839 patent which recites, in part,
`
`“wherein the determining step comprises the substeps of: inferring the likely
`
`solution to the problem from questions, actions, and rules contained in a
`
`knowledge database,” Gurer discloses that “CBR problem solving can be depicted
`
`as a five-step process: (1) retrieval, (2) interpretation and adaptation, (3) evaluation
`
`and repair, (4) implementation, and (5) evaluation and learning. The first step is
`
`retrieving cases that best match the current situation or case. Thus it is crucial to
`
`use an appropriate indexing method, such as decision trees or nearest neighbor
`
`matching. Once a case is retrieved, it must be interpreted and then adapted. The
`
`interpretation process is a simple comparison between the retrieved cases and the
`
`current case. Adaptation is a complicated, domain-dependent process that uses
`
`rules to adapt the current case to the problem situation and propose an initial
`
`solution, based on the similarities and differences. The next step is an evaluation
`
`and repair cycle where the proposed solution is evaluated through comparisons to
`
`
`
`20
`
`ORACLE EX. 1007, p. 20
`
`
`
`cases with similar solutions or through simulation, and the solution is modified
`
`accordingly. After the CBR system has found its best solution, the solution is
`
`implemented and the results are evaluated. (Ex. 1003 at 7:1-10).
`
`(43) Further, claim 8 of the ‘839 patent also recites, in part, “wherein the
`
`AI engine utilizes the selected ones of the plurality of sensors to gather information
`
`when the knowledge database lacks information necessary to answer a question.”
`
`Gurer discloses that “[t]he filtering and correlation of alarms is the first step of
`
`fault diagnosis. The second step involves further analysis and identification of the
`
`exact cause of the alarms, or the fault. This process is an iterative one where alarm
`
`data are analyzed and decisions are made whether more data should be gathered, a
`
`finer grained analysis should be executed, or problem solving should be performed.
`
`Gathering more data can consist of sending tests to network elements or requesting
`
`network performance data.” (Ex. 1003 at 6: 18-22). Gurer also notes earlier that
`
`“[t]ypically alarms are produced by . . . network elements (e.g., ATM switches,
`
`customer premise equipment).” (Id. at 1: 30-31).
`
`(44) Gurer further discloses that “[g]athering more data can consist of
`
`sending tests to network elements or requesting network performance data.” (Ex.
`
`1003 at 6:18-22). Gurer notes earlier that “[t]ypically alarms [or sensors] are
`
`produced by . . . network elements (e.g., ATM switches, customer premise
`
`equipment).” (Id. at 1:30-31). Further, “[f]aults are identified by analyzing the
`
`
`
`21
`
`ORACLE EX. 1007, p. 21
`
`
`
`filtered and correlated alarms and by requesting tests and status updates from the
`
`element managers, which provide additional information for diagnosis.” (Id. at
`
`2:4-5).
`
`(45) The logical alarms described by Gurer involve performing statistical
`
`analysis of the network behavior. Hence, consistent with claim 14 of the ‘839
`
`patent which recites, in part, “wherein the detecting step comprises the steps of:
`
`periodically activating selected ones of the plurality of sensors to gather
`
`information about the computer system,” and similarly claim 17 which recites, in
`
`part, “wherein the sensing step comprises the substep of: periodically activating at
`
`least one sensor to gather information about the computer system,” it would be
`
`clear to a person of ordinary skill in the art that the sensor that performs this
`
`statistical analysis must be periodically activated in order to recognize the logical
`
`alarms. In addition, Gurer describes how the CBR system determines that
`
`additional information is needed from the networks elements in order to diagnose a
`
`problem, which implies that selected sensors would be activated to collect this
`
`information. “After receiving the filtered and correlated alarms, the CBR system
`
`attempts to identify what information would further the diagnosis of the fault. The
`
`CBR system needs to decide on its own what additional tests need to be made on
`
`the network elements and what granularity of NN to use to further analyze the
`
`data.” (Ex. 1003 at 8:1-3).
`
`
`
`22
`
`ORACLE EX. 1007, p. 22
`
`
`
`(46) Claim 15 of the ‘839 patent is similar in scope to claim 1 discussed
`
`above in with the only differences being (1) claim 1 recites a tool whereas claim 15
`
`recites a program storage device for automatically maintaining a computer system,
`
`and (2) the limitation reciting “sensing information about the computer system by
`
`at least one sensor” in claim 15 versus a plurality of sensors in claim 1. It would
`
`be clear to a person of ordinary skill in the art that the method of Gurer may be
`
`readily executed on a computer readable storage medium. Further, the distinction
`
`between the terms “at least one sensor” and “a plurality of sensors” is merely one
`
`of form. Accordingly, the discussion of claim 1 applies correspondingly to claim
`
`15.
`
`ALLEN ‘218
`
`(47) U.S. Patent No. 5,586,218 to Allen (“Allen ‘218,” Ex. 1004) describes
`
`a case-based reasoning tool that includes a sensor, an AI engine, and a knowledge
`
`database containing cases, where one of the three motivating uses of the tool is to
`
`automatically perform preventative maintenance on office equipment, e.g.,
`
`printers, photocopiers, etc. (See FIG. 5 of Allen ‘218, entitled “Knobots In
`
`Diagnosis and Repair”). (Ex. 1004 at 1:51 – 2:3).
`
`
`
`23
`
`ORACLE EX. 1007, p. 23
`
`
`
`
`
`(48) If the case-based reasoning tool described in Allen ‘218 determines
`
`that it needs additional information to diagnose the problem, a “queries message
`
`119” is generated to request this additional information from the monitored system.
`
`(Id. at 4:56-60). Depending on the success or failure of the tool in repairing the
`
`problem, the set of cases in the knowledge database is updated accordingly based
`
`upon this “reinforcement.” (Id. at 4:28-39, 6:28-41).
`
`(49) Further, Allen ‘218 discloses “[a] software agent [101] which
`
`performs autonomous learning in a real-world environment, implemented in a
`
`case-based reasoning system and coupled to a sensor for gathering information
`
`from . . . its environment.” (See Ex.1004 at Abstract).
`
`(50) Although the language in the Allen ‘218 patent refers to a single
`
`“sensor,” the preventative maintenance example illustrated in FIG. 5 above in
`
`paragraph 49 refers to the sensor providing multiple “readings” as input to the AI
`
`engine. Allen ‘218 further discloses “[t]he features message 110 may comprise
`
`
`
`24
`
`ORACLE EX. 1007, p. 24
`
`
`
`sensor readings from the device 501 . . .” (Ex. 1004 at 8:3-4). To monitor a device
`
`as complex as a printer or photocopier, a large number of different features within
`
`these computer systems would need to be monitored. It appears that while the
`
`inventor of the ‘839 patent chose to describe separate sensors collecting different
`
`types of information from the monitored system, the inventor of the Allen ‘218
`
`patent instead chose to refer to a single “sensor” as collecting this wide array of
`
`different information. A skilled artisan would recognize that Allen ‘218 is simply
`
`using “sensor” to refer collectively to the comprehensive set of sensing
`
`functionality, which can be selectively activated using a “queries message 119.”
`
`(Ex. 1004 at 4:56-60).
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
