`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 1 of 50 PageID 1Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 1 of 50 PageID 1
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE NORTHERN DISTRICT OF TEXAS
`FORT WORTH DIVISION
`
`
`
`
`VALTRUS INNOVATIONS LTD.,
`
`
`Plaintiff,
`
`v.
`
`Case No. 4:22-cv-00020
`
`JURY TRIAL DEMANDED
`
`
`
`
`
`
`GOOGLE LLC,
`
`
`
`
`
`
`
`Defendant.
`
`
`VALTRUS INNOVATIONS LTD.’S COMPLAINT FOR PATENT INFRINGEMENT
`
`11034001
`
`
`
`
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 2 of 50 PageID 2Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 2 of 50 PageID 2
`
`Plaintiff Valtrus Innovations Limited (“Valtrus”), by and through its undersigned counsel,
`
`
`
`
`
`pleads the following against Google LLC (“Google”) and alleges as follows:
`
`THE PARTIES
`
`1.
`
`Plaintiff Valtrus is the successor in interest to a substantial patent portfolio created
`
`by Hewlett Packard Enterprise and its predecessor companies, including Compaq, Verity, and
`
`Hewlett-Packard Development Company (collectively, “HPE”). Valtrus is an Irish entity duly
`
`organized and existing under the laws of the Republic of Ireland. The address of the registered
`
`office of Valtrus is: The Glasshouses GH2, 92 Georges Street Lower, Dun Laoghaire, Dublin
`
`A96 VR66, Ireland. HPE’s worldwide corporate headquarters is located in Houston, Texas. One
`
`of HPE’s primary US facilities is located in Plano, Texas.
`
`2.
`
`Valtrus is the assignee and owns all right and title to U.S. Patent Nos. 6,728,704
`
`(“the ’704 Patent”), 6,738,764 (“the ’764 Patent”), 6,816,809 (“the ’809 Patent”), 7,346,604
`
`(“the ’604 Patent”), 7,523,454 (“the ’454 Patent”), and 7,748,005 (“the ’005 Patent”)
`
`(collectively, the “Asserted Patents”).
`
`3.
`
`The Asserted Patents were developed by inventors working for HPE. HPE and
`
`its predecessors have been developing innovative search, computer processing, and server
`
`technology for decades.
`
`4.
`
`On information and belief, Defendant Google is a limited liability company duly
`
`organized and existing under the laws of the State of Delaware, having a regular and established
`
`place of business in the Northern District of Texas, including at 3800 Railport Parkway,
`
`Midlothian, Texas 76065.
`
`5.
`
`Google has become the dominant search engine in the United States and beyond,
`
`with a reported market share of around 90%. Google’s strategy has included taking technology
`
`from other companies without regard to intellectual property rights and combining that
`
`11034001
`
`
`
`- 2 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 3 of 50 PageID 3Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 3 of 50 PageID 3
`
`
`
`technology to create and maintain market dominance. A core part of this strategy involved
`
`taking the innovations in the Asserted Patents. One need look no further than Google’s
`
`employment records, which have listed as some of Google’s most senior technologists a large
`
`number of inventors on the Asserted Patents, including a Senior Vice President at Google
`
`responsible for search, the first Director of Google R&D Bangalore, and the Vice President at
`
`Google responsible for Google Apps and Cloud.
`
`JURISDICTION AND VENUE
`
`6.
`
`This is an action arising under the patent laws of the United States, 35 U.S.C. § 1
`
`et seq. Accordingly, this Court has subject matter jurisdiction pursuant to 28 U.S.C. §§ 1331 and
`
`1338(a).
`
`7.
`
`This Court has personal jurisdiction over Google because Google creates products
`
`and services that are and have been used, offered for sale, sold, and purchased in the Northern
`
`District of Texas, and Google has committed, and continues to commit, acts of infringement in
`
`the Northern District of Texas, has conducted business in the Northern District of Texas, and/or
`
`has engaged in continuous and systematic activities in the Northern District of Texas.
`
`8.
`
`Under 28 U.S.C. §§ 1391(b)-(d) and 1400(b), venue is proper in this judicial
`
`district because Google maintains a regular and established place of business in this district and
`
`has committed and regularly commits acts of infringement within this judicial district giving rise
`
`to this action. For example, Google operates a 260,000 square foot data center in Midlothian,
`
`Texas. Valtrus is informed and believes that this data center includes the infringing systems and
`
`practices the infringing methods described herein. This data center is one of only fourteen in all
`
`of North America. Valtrus is informed and believes that Google is preparing to open a Google
`
`Cloud location in Dallas, which will be one of only eleven Cloud locations in North America,
`
`and which similarly is and/or will be a location of significant infringement. On information and
`
`11034001
`
`
`
`- 3 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 4 of 50 PageID 4Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 4 of 50 PageID 4
`
`
`
`belief, Google also maintains an office in the Dallas-Fort Worth area, and runs a content
`
`distribution network (CDN) node in the Dallas area. Google also provides, sells, and offers for
`
`sale infringing products and services to users in the Northern District of Texas.
`
`9.
`
`Google also has significant operations in nearby cities, including Austin. For
`
`example, on information and belief, Google owns approximately 550,000 square feet of office
`
`space across three locations in downtown Austin, Texas. Google is also continuing to grow its
`
`presence in Texas. For example, Google is preparing to open an additional 750,000 square feet
`
`of offices in Austin at Block 185, a new office tower located at 601 West Second Street that is
`
`set to open in 2023. Valtrus is informed and believes that these offices include employees
`
`responsible for the subject matter of this patent suit. For example, on information and belief,
`
`Google employees in Austin work on Google Cloud, finance, and engineering. On information
`
`and belief, Google also has an office in Houston, and employs more than 1700 people in Texas.
`
`FIRST CLAIM
`
`(Infringement of U.S. Patent No. 6,728,704)
`
`10.
`
`Valtrus re-alleges and incorporates herein by reference Paragraphs 1-9 of its
`
`Complaint.
`
`11.
`
`The ’704 Patent, entitled “Method and apparatus for merging result lists from
`
`multiple search engines,” was duly and lawfully issued on April 27, 2004. A true and correct
`
`copy of the ’704 Patent is attached hereto as Exhibit 1.
`
`12.
`
`The ’704 Patent names Jianchang Mao, Rajat Mukherjee, Prabhakar Raghavan,
`
`and Panayiotis Tsaparas as co-inventors.
`
`13.
`
`The ’704 Patent has been in full force and effect since its issuance. Valtrus owns
`
`by assignment the entire right and title in and to the ’704 Patent, including the right to seek
`
`damages for any infringement thereof.
`
`11034001
`
`
`
`- 4 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 5 of 50 PageID 5Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 5 of 50 PageID 5
`
`
`
`14.
`
`The ’704 Patent “relates generally to search engine technology. More
`
`specifically, this invention relates to reducing the computational overhead associated with
`
`merging results from multiple search engines.” Ex. 1 at 1:7-10.
`
`15.
`
`The ’704 Patent explains that “the merging of multiple result lists into a single list
`
`is usually accomplished by examining and ranking every single entry of every list. . . . Thus, for
`
`large lists or large numbers of lists, the computation time required by the merging process can
`
`nullify any advantage gained by operating multiple search engines at the same time.” Id. at 2:48-
`
`51, 53-56.
`
`16.
`
`The ’704 Patent states that the “invention allows for a reduction in computational
`
`overhead when merging and re-ranking multiple result lists. Ranking of results is accomplished
`
`by evaluating a subset of entries instead of every single one, thus reducing the number of
`
`calculations required.” Id. at 3:20-24.
`
`17.
`
`The method described by the ’704 Patent improves the function of a computer
`
`utilizing said method by reducing the considerable computational overhead associated with
`
`merging many results from numerous search engines. A reduction in computational overhead
`
`enables the computer to operate multiple search engines simultaneously without sacrificing
`
`response time, allowing the gathering, ranking, and presentation of millions of search results
`
`from across the Internet in a very short amount of time. The ’704 Patent provides detailed,
`
`specific steps for merging multiple result lists in this manner. For example, in one embodiment,
`
`each result list is assigned a representative scoring value based on a selected subset of entries
`
`from the list. Id. at 5:44-55; 7:12-14. In another embodiment, each list is instead “assigned a
`
`probability value equal to its average scoring value’s percentage of the total of all average
`
`11034001
`
`
`
`- 5 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 6 of 50 PageID 6Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 6 of 50 PageID 6
`
`
`
`scoring values,” and entries are selected from each list based on its probability value. Id. at 7:35-
`
`39.
`
`18.
`
`Valtrus is informed and believes, and thereon alleges, that Google has infringed
`
`and unless enjoined will continue to infringe one or more claims of the ’704 Patent, in violation
`
`of 35 U.S.C. § 271, by, among other things, using, selling, and offering for sale, without
`
`authority or license, Google products that use the claimed method of merging result lists in an
`
`infringing manner. Google practices every step of at least claim 1 of the ’704 Patent in the
`
`United States, including one or more steps that it practices in the Northern District of Texas.
`
`19.
`
`For example, the ’704 accused product, Google Search, embodies every limitation
`
`of at least claim 1 of the ’704 Patent, literally or under the doctrine of equivalents, as set forth
`
`below. The further descriptions below, which are based on publicly available information, are
`
`preliminary examples and are non-limiting.
`
`20.
`
`Google Search practices a method of merging result lists from multiple search
`
`engines, comprising the elements set forth below.
`
`11034001
`
`
`
`- 6 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 7 of 50 PageID 7Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 7 of 50 PageID 7
`
`
`
`21.
`
`For example, a query transmitted to Google Search returns results from multiple
`
`search engines displayed to the user as a single merged list of results. See, e.g., Google Search
`
`Results for “in-n-out burger” Query 1 (red arrows added):
`
`22.
`
`Google Search practices a method comprising transmitting a query to a set of
`
`search engines.
`
`11034001
`
`
`
`- 7 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 8 of 50 PageID 8Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 8 of 50 PageID 8
`
`
`
`23.
`
`For example, a query transmitted to Google Search returns results from multiple
`
`search engines. This process begins by transmitting that query to said search engines. See, e.g.,
`
`id. (red arrows added):
`
`
`
`
`
`
`
`11034001
`
`
`
`- 8 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 9 of 50 PageID 9Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 9 of 50 PageID 9
`
`
`
`24.
`
`Google Search practices a method comprising receiving in response to said query
`
`a result list from each search engine of said set of search engines, each result list including one or
`
`more entries.
`
`25.
`
`For example, a query transmitted to Google Search returns one or more results
`
`from each of several search engines. Said results are grouped into result lists associated with
`
`each of said search engines. In order to display said result lists, Google Search first receives a
`
`result list from each search engine. See, e.g., id. (red arrows added):
`
`
`
`11034001
`
`
`
`- 9 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 10 of 50 PageID 10Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 10 of 50 PageID 10
`
`
`
`26.
`
`Google Search practices a method comprising selecting a subset of entries from
`
`each result list to form a set of selected entries.
`
`27.
`
`For example, a query transmitted to Google Search may return many millions of
`
`results across a set of multiple search engines. Only a subset of these results, or entries, is
`
`selected for display to the user near the top of the Google Search result list. See, e.g., Google
`
`Search Results for “in-n-out burger” Query 2 (callout added):
`
`28.
`
`Google Search practices a method comprising assigning to each selected entry of
`
`said set of selected entries a scoring value according to a scoring function.
`
`11034001
`
`
`
`- 10 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 11 of 50 PageID 11Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 11 of 50 PageID 11
`
`
`
`29.
`
`Entries are displayed to a user of Google Search in order of relevance. On
`
`information and belief, these entries are assigned a scoring value according to a scoring function
`
`in order to be sorted and displayed in this way. For example, a list of results from the Maps
`
`subset of entries is sorted according to a scoring function which considers metrics such as the
`
`distance of each result from the user at the time of the query. See, e.g., Google Maps Result List
`
`(red arrow added):
`
`30.
`
`Google Search practices a method comprising assigning to each subset a
`
`representative value according to the scoring values assigned to its entries.
`
`31.
`
`For example, Google Search presents results in subsets that appear in different
`
`orders for different queries. On information and belief, Google Search orders each subset based
`
`on a representative value according to the scoring values assigned to its entries. See, e.g.,
`
`Google Search Results for “in-n-out burger” Query 1 (red arrows added):
`
`11034001
`
`
`
`- 11 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 12 of 50 PageID 12Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 12 of 50 PageID 12
`
`
`
`
`
`32.
`
`Google Search practices a method comprising producing a merged list of entries
`
`in a predetermined manner based on the representative value assigned to each result list.
`
`33.
`
`For example, when a user transmits a query involving a term associated with a
`
`physical location, Google Search produces a merged list of entries where in Maps results are
`
`generally displayed higher in the merged list than News results because, on information and
`
`belief, the Maps result list has been assigned a higher representative value. See, e.g., id. (red
`
`arrows added):
`
`11034001
`
`
`
`- 12 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 13 of 50 PageID 13Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 13 of 50 PageID 13
`
`
`
`
`
`34.
`
`Google Search practices a method comprising producing a merged list of entries
`
`in a predetermined manner based on the representative value assigned to each result list, wherein
`
`the representative value varies in accordance with predetermined manner.
`
`35.
`
`For example, result lists from each individual search engine are generally
`
`displayed as individual groups of results in the merged list presented to the user. On information
`
`and belief, the representative value of each result list, which determines the order of said groups,
`
`11034001
`
`
`
`- 13 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 14 of 50 PageID 14Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 14 of 50 PageID 14
`
`
`
`varies in accordance with the nature of a user’s query. See, e.g., Google Search Result
`
`Groupings (red arrows added):
`
`
`
`36.
`
`Google has had knowledge of the ’704 Patent since at least January 1, 2013. On
`
`this date, two patents assigned to Google Inc. were issued, both of which cite the ’704 Patent as a
`
`reference: U.S. Patent Nos. 8,346,791 and 8,346,792. The ’704 Patent is also cited as a reference
`
`by two later patents, U.S. Patent Nos. 8,521,725 (issued August 27, 2013) and 9,152,714 (issued
`
`October 6, 2015), both of which name Google Inc. as assignee. Furthermore, two of the co-
`
`inventors of the ’704 Patent, Rajat Mukherjee and Prabhakar Raghavan, are listed as inventors on
`
`at least seven other patents or publications which are assigned to Google: U.S. Patent Nos.
`
`11034001
`
`
`
`- 14 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 15 of 50 PageID 15Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 15 of 50 PageID 15
`
`
`
`8,782,071 (issued July 15, 2014), 9,251,168 (issued February 2, 2016), 10,079,785 (issued
`
`September 18, 2018), 10,402,889 (issued September 3, 2019), and 10,917,371 (issued February
`
`9, 2021); and U.S. Patent Application Publication Nos. 2014/0258466 (published September 11,
`
`2014) and 2016/0371425 (published December 22, 2016). Two of these patents, U.S. Patent
`
`Nos. 8,782,071 and 10,402,889, relate directly to search technology. Furthermore, Prabhakar
`
`Raghavan now holds the position of Senior Vice President at Google, where he “is responsible
`
`for Google Search,” among other products.1 Given that four of its own patents cite the ’704
`
`Patent as a reference, two of the co-inventors of the ’704 Patent are listed as inventors on
`
`numerous later patents assigned to Google, and one of the co-inventors of the ’704 Patent is now
`
`in a senior leadership role at Google, Google knew or was willfully blind to the fact that its
`
`conduct was infringing by, at the very latest, July 15, 2014.
`
`37.
`
`Valtrus is informed and believes, and thereon alleges, that Google actively,
`
`knowingly, and intentionally has induced infringement of the ’704 Patent by, for example,
`
`offering for public use Google Search with the intent to encourage and facilitate infringing uses
`
`of that service in the Northern District of Texas, in the United States, and throughout the world.
`
`38.
`
`As a result of Google’s infringement of the ’704 Patent, Valtrus has been
`
`damaged. Valtrus is entitled to recover damages sustained as a result of Google’s wrongful acts
`
`in an amount subject to proof at trial.
`
`39.
`
`In addition, Google’s infringing acts and practices have caused and are causing
`
`immediate and irreparable harm to Valtrus.
`
`40.
`
`Valtrus is informed and believes, and thereon alleges, that Google’s infringement
`
`of the ’704 Patent has been and continues to be willful. As noted above, Google has long had
`
`
`1 https://research.google/people/PrabhakarRaghavan/
`
`11034001
`
`
`
`- 15 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 16 of 50 PageID 16Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 16 of 50 PageID 16
`
`
`
`knowledge of the ’704 Patent and its infringement of the ’704 Patent. Google has deliberately
`
`continued to infringe in a wanton, malicious, and egregious manner, with reckless disregard for
`
`Valtrus’s patent rights. Thus, Google’s infringing actions have been and continue to be
`
`consciously wrongful.
`
`SECOND CLAIM
`
`(Infringement of U.S. Patent No. 6,738,764)
`
`41.
`
`Valtrus re-alleges and incorporates herein by reference Paragraphs 1-40 of its
`
`Complaint.
`
`42.
`
`The ’764 Patent, entitled “Apparatus and method for adaptively ranking search
`
`results,” was duly and lawfully issued on May 18, 2004. A true and correct copy of the ’764
`
`Patent is attached hereto as Exhibit 2.
`
`43.
`
`The ’764 Patent names Jianchang Mao, Mani Abrol, Rajat Mukherjee, Michel
`
`Tourn, and Prabhakar Raghavan as co-inventors.
`
`44.
`
`The ’764 Patent has been in full force and effect since its issuance. Valtrus owns
`
`by assignment the entire right and title in and to the ’764 Patent, including the right to seek
`
`damages for any infringement thereof.
`
`45.
`
`The ’764 Patent “relates generally to computerized techniques for identifying
`
`relevant documents. More particularly, this invention describes computerized techniques for
`
`adaptively ranking documents identified in response to a search query.” Ex. 2 at 1:6-10.
`
`46.
`
`The ’764 Patent states that “it is impossible to predict a priori, for any corpus of
`
`documents and any associated search engine, on which queries the static method [of ranking
`
`search results] is satisfactory and on which other queries the adaptive method is satisfactory.”
`
`Id. at 1:51-55.
`
`11034001
`
`
`
`- 16 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 17 of 50 PageID 17Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 17 of 50 PageID 17
`
`
`
`47.
`
`The ’764 Patent explains that it is “highly desirable to provide a technique that
`
`selectively emphasizes a static method or an adaptive method to achieve optimal search results
`
`for a given query.” Id. at 1:56-59. “The invention provides improved search results by
`
`adaptively ranking, based upon the prior behavior of users, documents returned from a text
`
`search engine. More particularly, the prior behavior of users is utilized to determine the rate at
`
`which to apply adaptive correction for a given query.” Id. at 2:30-35.
`
`48.
`
`The method described by the ’764 Patent provides a novel approach for
`
`selectively emphasizing a static or an adaptive method of ranking search results based, at least in
`
`part, on the prior behavior of other users. By adaptively ranking documents based, at least in
`
`part, on other users’ behavior and building “a list of queries associated with a viewed document,”
`
`id. at 4:30, this method improves on the performance and operation of then-existing computer
`
`systems, which were unable “to predict a priori, for any corpus of documents and any associated
`
`search engine, on which queries the static method is satisfactory and on which other queries the
`
`adaptive method is satisfactory.” Id. at 1:52-55. The ’764 Patent describes specific steps for
`
`accomplishing this improvement through a number of different implementations. For example,
`
`the ’764 Patent sets out in detail a series of processing operations associated with an embodiment
`
`of the invention. Id. at 4:2-3. These operations include producing “relevance search results
`
`based upon a query,” logging a user’s pattern of viewing search result documents, producing a
`
`database “which stores a list of queries associated with a viewed document,” forming a feature
`
`vector for each viewed document which “characterizes attributes and query words associated”
`
`with it, and calculating a similarity score for a user’s query using said feature vectors. Id. at 4:4-
`
`46. These and other detailed descriptions provide step-by-step instructions for carrying out the
`
`method and thereby improve the function of a computer using the method. This method
`
`11034001
`
`
`
`- 17 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 18 of 50 PageID 18Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 18 of 50 PageID 18
`
`
`
`significantly reduces computational overhead and improves the speed of return of the search
`
`results.
`
`49.
`
`Valtrus is informed and believes, and thereon alleges, that Google has infringed
`
`and unless enjoined will continue to infringe one or more claims of the ’764 Patent, in violation
`
`of 35 U.S.C. § 271, by, among other things, using, selling, and offering for sale, without
`
`authority or license, Google products that use the claimed method of adaptively ranking search
`
`results in an infringing manner. Google practices every step of at least claim 1 of the ’764 Patent
`
`in the United States, including one or more steps that it practices in the Northern District of
`
`Texas.
`
`50.
`
`For example, the ’764 accused product, Google Search, embodies every limitation
`
`of at least claim 1 of the ’764 Patent, literally or under the doctrine of equivalents, as set forth
`
`below. The further descriptions below, which are based on publicly available information, are
`
`preliminary examples and are non-limiting.
`
`51.
`
`Google Search practices a method of ranking search results comprising the
`
`elements described below.
`
`52. When a user transmits a query, Google Search ranks search results. See, e.g.,
`
`How Search works – Ranking results2 (red rectangular annotation added):
`
`
`2 https://www.google.com/search/howsearchworks/how-search-works/ranking-results/
`
`11034001
`
`
`
`- 18 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 19 of 50 PageID 19Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 19 of 50 PageID 19
`
`
`
`53.
`
`Google Search practices a method comprising producing a relevance score for a
`
`document in view of a query.
`
`54.
`
`For example, Google Search algorithms “look at many factors,” including the
`
`“relevance and usability of pages,” to determine how to rank search results. In order for said
`
`algorithms to rank results, Google Search assigns a relevance score to results. See, e.g., Google
`
`Search consumer information3 (red underline added):
`
`55.
`
`Google Search practices a method comprising calculating a similarity score for
`
`said query utilizing a feature vector that characterizes attributes and query words of a different
`
`query associated with said document.
`
`56.
`
`For example, Google Search uses a “neural network-based technique for natural
`
`language processing (NLP) pre-training” known as BERT, which can “consider the full context
`
`
`3 https://support.google.com/websearch/answer/7585859?hl=en
`
`11034001
`
`
`
`- 19 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 20 of 50 PageID 20Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 20 of 50 PageID 20
`
`
`
`of a word by looking at the words that come before and after it.” See Pandu Nayak,
`
`Understanding searches better than ever before4 (red rectangular annotations added):
`
`57.
`
`BERT extracts feature vectors, which can characterize attributes and query words
`
`of other queries associated with the same site or document for other users. Feature vectors can
`
`then be used to determine the degree of similarity between sites and assign a score accordingly.
`
`58.
`
`Google Search practices a method comprising assigning a rank value for said
`
`document based upon said relevance score and said similarity score.
`
`59.
`
`For example, Google Search assigns a rank value to each document produced by a
`
`user’s query based on the relevance and similarity scores produced by algorithms including
`
`BERT. See, e.g., Google Search consumer information (red underline added):
`
`
`4 https://blog.google/products/search/search-language-understanding-bert/
`
`11034001
`
`
`
`- 20 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 21 of 50 PageID 21Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 21 of 50 PageID 21
`
`
`
`60.
`
`Google has had knowledge of the ’764 Patent at least since May 22, 2007. On
`
`this date, U.S. Patent No. 7,222,127, which cites the ’764 Patent as a reference, was issued with
`
`Google Inc. as assignee. The ’764 Patent is also cited as a reference by the following later
`
`patents: U.S. Patent Nos. 7,716,225 (issued May 11, 2010), 7,870,147 (issued January 11, 2011),
`
`8,051,096 (issued November 1, 2011), 8,140,524 (issued March 20, 2012), 8,346,791 (issued
`
`January 1, 2013), 8,346,792 (issued January 1, 2013), 8,359,309 (issued January 22, 2013),
`
`8,396,865 (issued March 12, 2013), 8,498,974 (issued July 30, 2013), 8,521,725 (issued August
`
`27, 2013), 8,615,514 (issued December 24, 2013), 8,661,029 (issued February 25, 2014),
`
`8,694,511, 8,694,374 (both issued April 8, 2014), 8,832,083 (issued September 9, 2014),
`
`8,843,536 (issued September 23, 2014), 8,874,555 (issued October 28, 2014), 8,909,655 (issued
`
`December 9, 2014), 8,924,379 (issued December 30, 2014), 8,938,463 (issued January 20, 2015),
`
`8,959,093 (issued February 17, 2015), 8,972,391, 8,972,394 (both issued March 3, 2015),
`
`9,002,867 (issued April 7, 2015), 9,009,146 (issued April 14, 2015), 9,092,510 (issued July 28,
`
`2015), 9,110,975 (issued August 18, 2015), 9,183,499 (issued November 10, 2015), 9,223,868
`
`(issued December 29, 2015), and 9,623,119 (issued April 18, 2017), all of which name Google
`
`Inc. as assignee. Furthermore, three of the co-inventors of the ’764 Patent, Rajat Mukherjee,
`
`Prabhakar Raghavan, and Michel Tourn, are listed as inventors on at least eight other patents or
`
`publications which are assigned to Google: U.S. Patent Nos. 8,738,612 (issued May 27, 2014),
`
`8,782,071 (issued July 15, 2014), 9,251,168 (issued February 2, 2016), 10,079,785 (issued
`
`September 18, 2018), 10,402,889 (issued September 3, 2019), and 10,917,371 (issued February
`
`9, 2021); and U.S. Patent Application Publication Nos. 2014/0258466 (published September 11,
`
`2014) and 2016/0371425 (published December 22, 2016). Three of these patents, U.S. Patent
`
`Nos. 8,738,612, 8,782,071, and 10,402,889, relate directly to search technology. Furthermore,
`
`11034001
`
`
`
`- 21 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 22 of 50 PageID 22Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 22 of 50 PageID 22
`
`
`
`Prabhakar Raghavan now holds the position of Senior Vice President at Google, where he “is
`
`responsible for Google Search,” among other products.5 With more than 30 patents assigned to
`
`Google over the course of a decade citing the ’764 Patent as a reference, three of the co-inventors
`
`of the ’764 Patent listed as inventors on numerous later patents assigned to Google, and at least
`
`one of the co-inventors of the ’764 Patent in a senior leadership role at Google, Google knew or
`
`was willfully blind to the fact that its conduct was infringing since at least the earliest issue date
`
`of these patents.
`
`61.
`
`Valtrus is informed and believes, and thereon alleges, that Google actively,
`
`knowingly, and intentionally has induced infringement by, for example, offering for public use
`
`Google Search with the intent to encourage and facilitate infringing uses of that service in the
`
`Northern District of Texas, in the United States, and throughout the world.
`
`62.
`
`As a result of Google’s infringement of the ’764 Patent, Valtrus has been
`
`damaged. Valtrus is entitled to recover damages sustained as a result of Google’s wrongful acts
`
`in an amount subject to proof at trial.
`
`63.
`
`In addition, Google’s infringing acts and practices have caused and are causing
`
`immediate and irreparable harm to Valtrus.
`
`64.
`
`Valtrus is informed and believes, and thereon alleges, that Google’s infringement
`
`of the ’764 Patent has been and continues to be willful. As noted above, Google has long had
`
`knowledge of the ’764 Patent and its infringement of the ’764 Patent. Google has deliberately
`
`continued to infringe in a wanton, malicious, and egregious manner, with reckless disregard for
`
`Valtrus’s patent rights. Thus, Google’s infringing actions have been and continue to be
`
`consciously wrongful.
`
`
`5 https://research.google/people/PrabhakarRaghavan/
`
`11034001
`
`
`
`- 22 -
`
`
`
`

`

`
`
`Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 23 of 50 PageID 23Case 4:22-cv-00020-O Document 1 Filed 01/10/22 Page 23 of 50 PageID 23
`
`
`
`THIRD CLAIM
`
`(Infringement of U.S. Patent No. 6,816,809)
`
`65.
`
`Valtrus re-alleges and incorporates herein by reference Paragraphs 1-64 of its
`
`Complaint.
`
`66.
`
`The ’809 Patent, entitled “Hardware based utilization metering,” was duly and
`
`lawfully issued on November 9, 2004. A true and correct copy of the ’809 Patent is attached
`
`hereto as Exhibit 3.
`
`67.
`
`68.
`
`The ’809 Patent names Edgar Circenis as inventor.
`
`The ’809 Patent has been in full force and effect since its issuance. Valtrus owns
`
`by assignment the entire right and title in and to the ’809 Patent, including the right to seek
`
`damages for any infringement thereof.
`
`69.
`
`The ’809 Patent’s technical field includes “methods that use central processor
`
`metering to determine processor utilization for billing and other purposes.” Ex. 3 at 1:6-8.
`
`70.
`
`The ’809 Patent states that, “[i]n a computer system having hardware that may be
`
`partitioned, gathering processor utilization data from a hardware system requires
`
`communications between the metering application and all operating systems running within the
`
`hardware. The need for communication with different operating systems poses significant
`
`challenges because operating syst