Parus Holdings Inc v. Google LLC

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Parus Holdings Inc. (Delaware)
  • Defendant: Google, LLC (Delaware)
  • Plaintiff’s Counsel: Mintz, Levin, Cohn, Ferris, Glovsky and Popeo, P.C.
• Case Identification: 6:21-cv-00571, W.D. Tex., 08/11/2021
• Venue Allegations: Plaintiff alleges venue is proper in the Western District of Texas because Defendant Google maintains a permanent physical presence and regular, established places of business in the district, and has allegedly committed acts of patent infringement there.
• Core Dispute: Plaintiff alleges that Defendant’s Google Assistant service and the associated hardware products on which it operates infringe patents related to voice-based systems for searching the internet and controlling remote devices.
• Technical Context: The technology at issue falls within the domain of voice-activated digital assistants, a key interface for consumer interaction with the internet and smart devices.
• Key Procedural History: This First Amended Complaint alleges that all three asserted patents are related and share a common specification. Plaintiff alleges Defendant had constructive notice of the asserted patents as of 2014 due to patent marking.

Case Timeline

Date	Event
2000-02-04	Earliest Priority Date for '705, '941, and '455 Patents
2004-04-13	U.S. Patent No. 6,721,705 Issues
2008-06-10	U.S. Patent No. 7,386,455 Issues
2011-02-01	U.S. Patent No. 7,881,941 Issues
2014-01-01	Alleged Constructive Notice of Patents to Defendant
2016-01-01	Alleged Launch of Infringing Google Assistant Service
2021-08-11	Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 6,721,705 - "Robust Voice Browser System and Voice Activated Device Controller"

The Invention Explained

• Problem Addressed: The patent describes a technical landscape where accessing internet information via portable devices like PDAs or early web-phones was cumbersome and unreliable (Compl. ¶17; ’705 Patent, col. 1:25-54). These devices had limited capabilities, and the rapidly changing structure and content of websites meant that systems designed to access them would often fail (Compl. ¶17; ’705 Patent, col. 2:25-36). Furthermore, voice-based systems faced the unique hurdle of user intolerance for latency, requiring immediate responses to be useful (Compl. ¶18; ’705 Patent, col. 2:43-46).
• The Patented Solution: The invention proposes a system that decouples a simple voice-enabled user device from the complexity of web browsing. The system uses a back-end server architecture that maintains a database of information sources (e.g., web sites) and assigns a rank to each source (’705 Patent, Abstract). When a user makes a voice request, the system accesses the source with the highest rank to retrieve the information (’705 Patent, col. 4:25-34). The system also includes a polling mechanism to periodically check the web sites for functionality and response speed, automatically adjusting the rankings to ensure reliability and speed (’705 Patent, col. 16:35-67).
• Technical Importance: This approach aimed to make voice-based internet browsing on limited-resource devices practical by offloading the complex processing to a robust server system that could adapt to the dynamic nature of the web (’705 Patent, col. 2:32-36).

Key Claims at a Glance

• The complaint asserts at least independent claim 1 (Compl. ¶27).
• The essential elements of independent claim 1 include:
  • An Internet voice browsing system with a CPU-based media server and a CPU-based web browsing server.
  • A database containing a list of web sites.
  • A rank number assigned to each web site and stored in the database.
  • The web browsing server accessing the web site with the highest rank to retrieve information.
  • The media server generating and transmitting an audio message representing the retrieved information.
  • A polling mechanism that periodically sends a message to each web site and receives a response.
  • The polling mechanism is configured to decrease the rank of a polled web site under certain conditions: if no response is received, if an unexpected response is received, or if its response time is longer than another polled web site's response time.
• The complaint reserves the right to assert other claims (Compl. ¶84A).

U.S. Patent No. 7,881,941 - "Robust Voice Browser System and Voice Activated Device Controller"

The Invention Explained

• Problem Addressed: As a continuation of the '705 patent, this patent addresses the same set of problems related to providing reliable and timely voice-based web search results from resource-constrained devices (Compl. ¶17).
• The Patented Solution: The ’941 patent focuses on a specific method for retrieving information. The method involves providing an instruction set that includes a plurality of pre-selected website addresses. When a user issues a speech command, the system accesses the first website in the plurality; if the information is not found, it sequentially accesses the other websites until the information is located or the list is exhausted ('941 Patent, Abstract; col. 20:45-67). This formalizes a fallback procedure for information retrieval.
• Technical Importance: This patented method provides a structured, sequential approach to querying multiple sources, aiming to increase the probability of successfully retrieving information for a given voice command (Compl. ¶21; '941 Patent, col. 20:45-48).

Key Claims at a Glance

• The complaint asserts at least independent claim 1 (Compl. ¶48).
• The essential elements of independent claim 1 (a method claim) include:
  • Providing a computer connected to the internet, a speech recognition engine, and a speech synthesis engine.
  • Providing a voice-enabled device to receive speech commands from users.
  • Providing an instruction set in a database with a plurality of pre-selected website addresses.
  • Providing a user's speech command to the speech recognition engine.
  • The computer accessing at least one of the websites from the instruction set to obtain information.
  • The computer first accessing a first website, and if the information is not found, sequentially accessing the plurality of websites until the information is found or the plurality has been accessed.
  • The speech synthesis engine producing and transmitting an audio message with the retrieved information.
• The complaint reserves the right to assert other claims (Compl. ¶84A).

Multi-Patent Capsule

• Patent Identification: U.S. Patent No. 7,386,455, "Robust Voice Browser System and Voice Activated Device Controller," issued June 10, 2008.
• Technology Synopsis: This patent claims a system and method for controlling a remote system over a network like the Internet using voice commands. An instruction set stored in a database identifies the remote system, and a user's speech command prompts the computer to access that system and execute a pre-selected function ('455 Patent, Abstract; Compl. ¶22).
• Asserted Claims: The complaint asserts at least independent claim 1 (Compl. ¶69).
• Accused Features: The complaint accuses Google Assistant's functionality for controlling remote smart home devices, such as Google's Nest product line, of infringing the ’455 patent (Compl. ¶¶ 24, 71, 77).

III. The Accused Instrumentality

Product Identification

• The complaint names the "Google Accused Products," a collective term for Google hardware (Pixel phones, Nest devices, Chromecast, etc.) that implements the Google Assistant service (Compl. ¶24).

Functionality and Market Context

• The accused functionality centers on the Google Assistant voice service. Users issue voice commands to a device, which includes a microphone (Compl. ¶33). The system processes these commands, which can involve on-device AI or cloud-based processing, to understand the user's intent (Compl. ¶41). A diagram in the complaint illustrates this flow from Automatic Speech Recognition (ASR) to Text-to-Speech (TTS) output (Compl. ¶41, p. 12).
• For information requests, Google Assistant queries Google's search infrastructure, which relies on the "Googlebot" crawler to discover and index web pages from a "list of web addresses from past crawls and sitemaps" (Compl. ¶36). The system then ranks these pages algorithmically to determine the most relevant result (Compl. ¶39). The retrieved information is converted into an audio message and spoken back to the user (Compl. ¶41). The complaint includes a screenshot from Google's promotional materials for the Pixel 3, which highlights "your Google Assistant" as a primary feature (Compl. ¶34, p. 9).
• For device control, the system is alleged to execute functions on remote systems, such as compatible smart home devices (Compl. ¶77). The complaint positions the Google Assistant and the associated products as a major part of Google's consumer offerings (Compl. ¶15).

IV. Analysis of Infringement Allegations

’705 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
a CPU-based media server, said media server including at least a speech recognition engine, a speech synthesis engine...	The Accused Products include a media server with speech recognition (ASR) and synthesis (TTS) engines that convert speech commands to data and generate audio responses. A diagram shows this ASR/TTS flow.	¶31, ¶41, p. 12	col. 8:36-40
a database containing a list of web sites stored on magnetic media,	Google's web crawling bot, "Googlebot," begins its process with "a list of web addresses from past crawls and sitemaps," which constitutes the database of websites. This information is stored in Google's data centers.	¶36, ¶38	col. 4:60-64
a rank number assigned to each one of said web sites and stored in said database;	Google Assistant allegedly determines ranks for each website to decide where to extract responses, with the complaint referencing "ranking for voice" and Google's use of algorithms to rank sites.	¶39	col. 4:29-31
a CPU-based web browsing server...configured to access one of said web sites having the highest said rank number and to retrieve information...	The web browsing servers of the Accused Products access one of the web sites (e.g., extracting a "featured snippet" from Wikipedia) having the highest rank to retrieve information.	¶40	col. 4:31-34
a polling mechanism configured to periodically send a polling digital data message to each one of said web sites and to receive a response... said polling mechanism configured to decrease said rank number...if no response is received...if an unexpected response is received...[or] if a response time...is longer...	Google allegedly uses polling mechanisms to determine webpage quality and change site rank. The complaint alleges this includes using digital data messages and evaluating responses to decrease rank based on factors like quality, cloaking, and other guidelines.	¶43, ¶44	col. 20:5-21

Identified Points of Contention

• Scope Questions: A principal question is whether Google's unified search index, which is built by crawling the public web, constitutes "a database containing a list of web sites" with an "assigned...rank number" as contemplated by the patent. The patent's specification appears to describe a more curated, category-specific list of distinct web sites (e.g., a choice between three weather sites) rather than a single, comprehensive index of billions of pages ranked by a complex algorithm like PageRank (’705 Patent, Table 1, Fig. 2).
• Technical Questions: The complaint alleges that Google's system meets the specific "polling mechanism" limitations for decreasing rank. It will be a key factual question whether Google's crawling and ranking algorithms, which are designed to assess hundreds of signals for relevance and quality, perform the specific rank-demotion functions recited in the claim (i.e., decreasing rank for no response, an unexpected response, or a slower response time) in the manner described by the patent (’705 Patent, col. 20:5-21; Compl. ¶43).

’941 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
providing a computer operatively connected to the internet...at least one speaker-independent speech recognition engine and at least one speech synthesis engine;	Google provides the Accused Products, which include a computer connected to the internet, a speech recognition engine, and a speech synthesis engine to process voice commands and generate audio.	¶51	col. 19:35-39
providing a voice enabled device operatively connected to said computer, said voice enabled device configured to receive speech commands from users;	Google provides voice-enabled devices (e.g., Pixel phones) with microphones that are connected to the computer system and receive speech commands from users.	¶54, ¶55	col. 19:40-42
providing at least one instruction set stored in a database... said instruction set including a plurality of pre-selected website addresses...	Google's crawling process begins with "a list of web addresses from past crawls and sitemaps," which constitutes the instruction set of pre-selected website addresses stored in a database.	¶57, ¶59	col. 19:43-48
said computer first accessing the first web site of the plurality of web sites and, if the information to be retrieved is not found...sequentially accesses...	The computer in the Accused Products allegedly accesses at least one of the websites to obtain information, and if not found at the first, "sequentially accesses the plurality of web sites until the information...is found."	¶63	col. 20:49-54
said speech synthesis engine producing an audio message containing any retrieved information...and transmitting said audio message to users via the voice enabled device.	The speech synthesis engine in the Accused Products produces an audio message with the retrieved information (e.g., reading a featured snippet) and transmits it to the user's device.	¶64, ¶65	col. 20:55-61

Identified Points of Contention

• Scope Questions: Similar to the '705 patent, a central issue is whether Google's search infrastructure functions as "a plurality of pre-selected web sites." The complaint alleges Google's crawling process starts from a list of sites, but the infringement theory may face questions about whether querying an integrated index is the same as accessing a discrete plurality of sites.
• Technical Questions: The claim requires a specific "sequential" access method: trying the first site, and if that fails, trying the next, and so on. It will be a factual question for the court whether Google's system, which retrieves results from its pre-compiled index, actually performs this "sequential" access of live websites at the time of a query, or if its method of result aggregation is technically distinct (Compl. ¶63).

V. Key Claim Terms for Construction

• The Term: "a database containing a list of web sites" (and "a plurality of pre-selected web sites")

• Context and Importance: This term appears in the independent claims of both the '705 and '941 patents and is foundational to the infringement theories. Its construction will determine whether Google's vast, algorithmically generated search index can be considered equivalent to the system described in the patents. Practitioners may focus on this term because the plaintiff's case depends on mapping this term onto Google's core search infrastructure.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The claims themselves do not specify a size or method of curation for the "database" or "list." The terms could be interpreted broadly to encompass any structured collection of website data, including a large-scale web index.
  • Evidence for a Narrower Interpretation: The specification repeatedly provides examples of a system with a small, curated list of websites for specific categories like "weather" or "stock" information (’705 Patent, Table 1, col. 5:27-46). Figure 2 of the patent, showing a database record, depicts discrete entries for individual websites with specific extraction commands, which may suggest a system architecturally different from a general-purpose web index (’705 Patent, Fig. 2).

• The Term: "sequentially accesses"

• Context and Importance: This term from claim 1 of the '941 patent defines the core of the claimed retrieval method. The infringement analysis will turn on whether Google's method of serving results from its index constitutes "sequentially" accessing websites in the manner claimed.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: "Sequentially" could be interpreted to mean any ordered or prioritized process. If Google's algorithm considers a primary source for a featured snippet and then considers other sources for supplemental results, a party could argue this constitutes a form of sequential access.
  • Evidence for a Narrower Interpretation: The claim language "if the information to be retrieved is not found at the first web site, the computer sequentially accesses the plurality of web sites" suggests a linear, step-by-step process of checking live websites one after another (’941 Patent, col. 20:49-54). This may support a narrower construction that requires a series of distinct access attempts, rather than a single query against a unified index.

VI. Other Allegations

• Indirect Infringement: The complaint does not contain separate counts for indirect or contributory infringement. The allegations are framed as Google directly infringing by making, using, and operating the accused systems (Compl. ¶¶ 27, 48, 69).
• Willful Infringement: The complaint does not use the word "willful" but does request increased damages pursuant to 35 U.S.C. § 284, which encompasses willfulness (Compl. p. 25, ¶D). The basis for this claim appears to be alleged pre-suit constructive notice of the patents "at least as of 2014" from patent marking, and post-suit knowledge upon the filing of the complaint (Compl. ¶¶ 9, 28, 49).

VII. Analyst’s Conclusion: Key Questions for the Case

• A core issue will be one of definitional scope: can the patent claims, which describe a system for selecting, ranking, and polling from a discrete list of third-party web sites, be construed to cover Google’s modern, monolithic search architecture that resolves queries against a single, vast, pre-crawled index of the web?
• A key evidentiary question will be one of technical operation: does the accused Google Assistant system perform the specific functions required by the claims? In particular, does Google’s ranking algorithm operate as the claimed "polling mechanism" that decreases rank based on specific negative feedback, and does its method of retrieving results from an index constitute the "sequentially accessing" of websites as claimed in the '941 patent?