Court: District Court, D. Massachusetts; November 11, 2014; Federal District Court
A patent dispute involves Kenexa BrassRing, Inc. suing HireAbility.com, LLC, Main Sequence Technology, Inc., and Sendouts, LLC for infringement of three patents—U.S. Patent No. 5,999,939 ('939 patent), U.S. Patent No. 6,996,561 ('561 patent), and U.S. Patent No. 7,958,059 ('059 patent). These patents describe systems and methods for accurately transferring user data, specifically from resumes to a hiring database. The '939 patent was issued on October 19, 2004, with priority dating back to December 21, 1997; the '561 patent followed on February 7, 2006, as a continuation of an abandoned application related to the '939 patent; and the '059 patent was issued on June 7, 2011, as a continuation of the '561 patent. The technology aims to streamline job application processes by extracting data from resumes, allowing user verification and modification, and storing the information in a structured database.
Kenexa's lawsuit alleges willful infringement and seeks treble damages. The defendants assert various defenses, including non-infringement and patent invalidity. The claim construction phase is currently addressing disputes over ten specific terms related to the patents, including definitions of "nonuniformly formatted source data streams" and "permitting a job applicant to author a resume employing a word processing application." The defendants offer products and services that Kenexa claims infringe upon its patented systems and methods.
Kenexa initiated this lawsuit on May 25, 2012. HireAbility and Sendouts responded with answers and counterclaims regarding invalidity and non-infringement on July 20 and October 22, 2012, while Main Sequence filed its invalidity counterclaims on September 4, 2012. A Markman hearing concerning disputed claim terms was conducted on June 2, 2014.
In a related case, Kenexa filed a lawsuit against Taleo Corporation and Vurv Technology, Inc. on August 27, 2007, claiming infringement of the '939 and '561 patents. A brief opinion issued by the district court on November 18, 2010, established that Vurv products infringed all asserted claims and that Taleo products infringed nearly all claims. The trial focused on invalidity issues related to the '939 patent but ended in settlement after one day.
The legal framework emphasizes that claim term construction is a legal question. Courts are tasked with interpreting patent claims based on the understanding of a person skilled in the art at the time of the patent application’s filing. The interpretation process draws from the claims' language, the specification, prosecution history, and relevant extrinsic evidence. Claims define the invention and provide the right to exclude others. Disputed meanings may be clarified through construction, which is not always necessary if the plain meaning suffices. The arrangement of terms within claims can also influence their interpretation, as terms are typically used consistently throughout the patent and dependent claims suggest specific limitations not present in independent claims. The specification, integral to the claims, is essential for understanding their intended meaning.
The specification plays a crucial role in claim construction analysis, often serving as the primary guide to the meaning of disputed terms. The claims' scope is determined by the patentee's description of the invention, which may include special definitions or intentional disclaimers of claim scope. Claims must align with the specification and cannot exceed the invention described therein. While the specification informs claim interpretation, courts must avoid importing limitations from it into the claims, as claims define the scope of patent protection, not the embodiments in the specification. The existence of a single embodiment does not impose a limitation on claims, as those skilled in the art typically do not restrict definitions to exact representations. The focus should be on how a person of ordinary skill understands the terms.
Following the specification, the prosecution history serves as the next best indicator of term meaning, encompassing the complete record of proceedings before the PTO and prior art cited. It can clarify the inventor’s understanding and any limitations applied during prosecution. However, since it reflects an ongoing negotiation rather than a final outcome, it may lack the clarity of the specification and thus be less effective for claim construction.
Courts require patent applicants to clearly and unambiguously express their surrender of subject matter to disavow claim scope during prosecution, as established in Voda v. Cordis Corp. and Sorensen v. Int’l Trade Commission. Extrinsic evidence, which includes expert and inventor testimony, dictionaries, and learned treatises, serves to educate the court on the invention's field and assist in interpreting claim terms from the perspective of a person of ordinary skill in the art. However, extrinsic evidence can be biased and its relevance can vary, making it less reliable for interpreting patent claim scope unless considered alongside intrinsic evidence.
In a related motion, defendants sought to strike two expert reports by Michael D. Siegel, arguing that he was not identified as an expert, his reports did not comply with Federal Rule of Civil Procedure 26, and the opinions were irrelevant. The plaintiff countered that disclosure of Siegel was unnecessary, that the reports met Rule 26 requirements, and that they were relevant to claim construction. Ultimately, the court denied the motion to strike, noting that defendants failed to consult with the plaintiff prior to filing, violating Local Rule 7.1, and allowing the exhibits to be considered for their relevance in claim construction.
Proposed constructions for disputed terms in the '939, '561, and '059 patents are outlined, comparing plaintiff’s and defendants’ interpretations.
Key terms include:
1. **Nonuniformly Formatted Source Data Streams**: The plaintiff defines this as "sets of data that are unstructured relative to a structured database," emphasizing variability in format. Defendants suggest a simpler definition: "data streams that vary in format from one stream to the next," referencing the '561 patent's specification, which highlights resumes as examples of loosely structured documents.
2. **Source Data String**: The plaintiff describes it as a "string of text from a source document," while defendants argue for a "plain and ordinary meaning."
3. **Target Data String**: Plaintiff's interpretation is "data obtained from the source data string and corresponding to a form," whereas defendants also advocate for the plain meaning.
4. **Extracting Selected Ones of Said Source Data Strings**: The plaintiff defines this as "identifying and isolating strings of text," with defendants supporting the plain meaning.
5. **User**: Both parties agree on the plain meaning, identifying the user as "the person who verifies the accuracy of the target data strings."
6. **Storing Data Corresponding to Modified and/or Accepted Data**: The plaintiff's construction involves "storing in a database said strings from said form fields," with defendants again favoring the plain meaning.
7. **Enabling Each User to Modify and/or Accept Target Data Strings**: Plaintiff outlines this as allowing users to interactively modify or accept displayed data to ensure accurate information transfer, while defendants also propose the plain meaning.
8. **Supplemental Inquiry Form**: Defined by the plaintiff as a form for receiving additional data beyond what is extracted, with defendants concurring on the plain meaning.
9. **Networks or Components that Can Transfer Data Between Independent Computer Systems**: The plaintiff's definition includes "a data connection between two or more devices," with the defendants agreeing on the plain meaning.
10. **Remote Communication Interface**: The plaintiff describes this as providing job applicants the opportunity to author resumes using a computer application with specific features, while defendants support the plain meaning.
The term "nonuniformly formatted source data streams" appears in the preambles of claims in both the '939 and '561 patents, where it is crucial for facilitating accurate data transfer into a structured database. The claims outline steps involving the receipt of digital data from multiple users, underscoring the need for clarity in the interpretation of the terms.
The Board of Patent Appeals and Interferences (BPAI) indicated that "nonuniformly formatted" data pertains to the diversity of user-provided documents. Defendants argue against defining nonuniformity in relation to a structured database, asserting that the preamble of the claims should not be construed as it merely states intended use and is not limiting. Conversely, the plaintiff argues for a construction that includes "unstructured relative to a structured database," asserting that this term in the preamble is a limitation that clarifies the claimed invention and connects to "source data streams" in the claims' body. The plaintiff cites a Delaware district court ruling that defines “nonuniformly formatted source data streams” as varying formats among resumes, with the same file types possibly being used.
Both parties seek the Court's construction of the term within claims 1 through 6 of the '561 patent, focusing on two components: the meanings of "nonuniformly formatted" and "data streams." They agree that "nonuniformly formatted" indicates variability in format but dispute whether it should include being "unstructured relative to a structured database." The Court finds little support for this added limitation in the claim or specification. It notes that while claims mention “a structured database,” they do not compare “source data streams” to it. The specification suggests that resumes are generally unstructured compared to the database, implying that the additional comparative phrase is unnecessary. Thus, the Court will define "nonuniformly formatted" simply as "vary in format from one stream to the next."
Regarding "data streams," the plaintiff proposes "sets of data," while the defendant suggests the ordinary term "data streams." The Court recognizes that laypersons may struggle with the technical interpretation of "data streams," indicating that a construction could aid the jury.
The term "sets of data" lacks clarity and may cause confusion if interpreted as "sets of audio" in the context of "audio streams." The Court is not obligated to add language for claim construction if the ordinary meaning is clear to a layperson, as indicated in O2 Micro. The specification provides a "source data stream" example as a "document/file," which may sufficiently clarify the term "data stream" for a jury, leading the Court to decline further construction of "data streams."
The parties disagree on whether the preamble to claims requires construction. A preamble serves as a limitation only if it describes essential structure or steps or is necessary for the claim's meaning. It can be limiting if it provides antecedent basis, is essential for understanding limitations, or is vital to the claim. Conversely, it is not limiting if the claim body defines a complete invention, using the preamble merely for purpose or intended use.
In the relevant patent claims, "nonuniformly formatted source data stream" does not have an antecedent reference, as seen in claim 1 of the '939 patent, where the first mention stands alone. However, further analysis shows that terms in the preamble must be construed if essential for understanding the claim body or necessary for the claim's meaning. The ability to transfer data into a structured database from nonuniformly formatted source data streams is a critical aspect of the inventions. Variations in formatting must be addressed for successful data transfer, as illustrated by an example from the specification regarding standardizing date formats in resumes. Thus, the modifier "nonuniformly formatted" in the preamble is essential for understanding and should be construed similarly to its usage in the claims.
The Court will define “nonuniformly formatted source data streams” as “data streams that may vary in format from one stream to the next,” applicable to both the claims' body and preamble. The term “source data string” is present in all claims of the '939 patent, claims 1 through 6 of the '561 patent, and claims 1 through 6 and 20 through 23 of the '059 patent. In claim 1 of the '939 patent, it refers to digital data from multiple user source data streams, each containing multiple discernible source data strings. The claims instruct to extract “source data strings” from these streams, generate “target data strings,” and insert them into structured forms. For instance, in a resume, "Jane" and "Doe" would be “source data strings” extracted from the source data stream.
The plaintiff proposes that “source data string” should be interpreted as “a string of text from a source document,” supported by claims and specifications, while suggesting alternatives like “file” or “source data stream.” They cite the '939 patent specification, explaining how the data extractor isolates discernible source data strings and determines correlations between them and database fields. The plaintiff supports their definition with technical dictionary definitions of “string” as a sequence of characters representing human-readable text.
Conversely, defendants argue for the term's plain meaning, asserting that the plaintiff's definition is overly narrow, as it restricts data to text form, is inconsistent with the patent's use of “source text strings,” and illogically implies that not all source data strings are text strings. They further argue that the term “source” should encompass more than just “documents,” including “audio streams” and “audio files.” The Court finds that "source data string" is not self-evident for a jury, indicating a need for clearer construction, but critiques the plaintiff's proposal for being overly reliant on technical definitions while misrepresenting the scope of “sources” in the patents.
The patents exhibit some inconsistency regarding the terms “data” and “text,” but overall, the specifications indicate that “source data strings” must be in text form. The claimed invention requires that these strings be displayed for user interaction, suggesting they should be readable and editable. Even references to audio streams in the '561 specification stipulate that they must be converted to text prior to processing. The Court will define “source data string” as “a string of text from a source data stream.”
Regarding “target data string,” this term appears in multiple claims across several patents. The plaintiff defines it as “a string of text based on a source data string and corresponding to a form field” or a “database field,” emphasizing that variations of date formats in source data strings must be converted to a uniform target format. The plaintiff argues that the specifications support the idea that a target data string must relate to both a source data string and a field on a form displayed to users.
Conversely, the defendants propose a broader definition, asserting it is simply “data obtained from the source data stream” and argue that a target data string does not need to correspond to a form or field. They claim that since “form” and “field” are part of dependent claims, their absence in independent claims suggests that these elements should not be included in the definition of a target data string. However, the Court finds the defendants' proposal confusing and unsupported, clarifying that target data strings are generated from portions of the source data stream, but are not part of it. The Court will define “target data string” as “a string of text based on a source data string,” ensuring clarity and consistency throughout the patent.
The term “extracting selected ones of said source data strings” is consistently found in claim 1 across three patents, including a related phrase in claim 21 of the '561 patent and another in claim 24 of the '059 patent. The plaintiffs propose that this term means “identifying and isolating strings of text from a source document,” while for claim 24, it specifies “from a received resume.” The specifications clarify that the data extractor isolates discernible strings in a resume but may also produce incomplete or erroneous correlations. Plaintiffs argue that extraction is distinct from the creation of target data strings, while defendants assert that no construction is necessary and criticize the plaintiffs' interpretations as overly narrow. They reference the DEFT text data extraction system, which performs more than simple identification and isolation. The Court will uniformly interpret “source data string” as “a string of text from a source data stream.” It acknowledges that “extracting” and “generating” are separate operations, and thus adopts the plaintiffs' interpretation of “extracting” as “identifying and isolating.” Consequently, the relevant phrases will be construed to reflect this understanding.
The term “user” appears in various claims across multiple patents. For example, claim 1 of the '939 patent discusses supplying digital data from multiple users and allowing them to modify or accept target data strings. Claim 20 of the '059 patent similarly describes receiving data from job applicants and enabling modifications to target data strings.
Plaintiff argues for the plain and ordinary meaning of "user," while Defendants suggest it refers specifically to "the person who verifies the accuracy of the target data strings." This definition contradicts claim 20 of the '561 patent, which differentiates between the "job applicant" who verifies data and the "user" who accesses the verified data. Consequently, the Court rejects Defendants' proposal. Although Defendants assert that some construction is necessary and claim that Plaintiff previously indicated that a "user" is the "author" of the source data stream, Plaintiff has not maintained this position in the current context. Instead, Plaintiff emphasizes that the claims detail the actions of the user, making the term comprehensible to a layperson. The Court concurs, stating that if the ordinary meaning of a term is clear and resolves the dispute, no further construction is required.
Regarding the '939 patent, claim 1 includes the steps for users to modify or accept target data strings and to store corresponding data in a database. Plaintiff asserts that no construction is necessary, while Defendants propose defining "storing" as user-modified data, contending that the storing step must follow the enabling step. Defendants also argue that the claim is ambiguous due to the reference to "said strings." Plaintiff counters that Defendants' proposal is redundant, as the claim already states that data must be modified or accepted. Despite some confusion from the repeated use of "said," the context clarifies that the data stored in the database is modified or accepted. The Court finds that further construction is neither necessary nor helpful and will only attribute ordinary meanings to the terms. Lastly, the Court notes that a patent typically does not require a specific order of steps unless the claim explicitly indicates such an order.
Method claims typically do not require steps to be performed in a specific order unless explicitly stated in the claim language or if a narrow construction is necessary. In the case at hand, claim 1 of the '939 patent implies an order where the “storing” step must occur after the “enabling” step, as indicated by the repeated use of the term “said” referring to prior steps. This logical and grammatical structure necessitates the sequence, as supported by previous case law. The patent aims to ensure accurate data processing, logically requiring that data is modified before being stored.
Claim 1 outlines a method involving the display of a structured form with fields for data strings, enabling users to modify or accept target data strings related to the source data stream. The defendants propose a construction that emphasizes user choice in modifying or accepting these strings to ensure accurate data transfer, suggesting that “enabling” implies interactivity and necessitates a decision-making process. They argue that the term “and/or” is ambiguous and should be separated into distinct actions. In contrast, the plaintiff argues against the need for construction, claiming that the defendants’ proposal introduces unnecessary complexity and limitations, asserting that the original phrasing is clear and commonly understood.
The proposal to define “enabling” as “allowing” is deemed unnecessary since the term is clearly understood and chosen by the patent drafters. The inclusion of “interactively” is considered unwarranted, as the claimed invention inherently involves user-system interaction. Additionally, a purposive statement regarding “accurate transfer” is unnecessary; courts cannot add limitations not found in the claim language, as established in Phillips and DSW cases. The ordinary meaning of “enabling each user... to modify and/or accept said target data strings” will be adopted.
Regarding the term “supplemental inquiry form” in claims 6 of the '939 and '561 patents, there is a disagreement between the Plaintiff and Defendants on its construction. The Plaintiff suggests it be defined as “a form having fields for receiving data in addition to the data actually extracted from the source data stream,” consistent with the construction in the Taleo case. The Defendants propose it as “a form having fields for receiving data from a user (that may be different from or in addition to) the data actually obtained from the source data stream.” The disagreement centers on whether the supplemental data can differ from the extracted data and whether it must specify that the supplemental data comes from a user. The Plaintiff argues that the supplemental inquiry form cannot request contradictory information to what has been obtained, while the Defendants assert that “different” indicates the form seeks additional information not captured in the prior extraction step.
The parties generally agree that the supplemental inquiry form (1) does not duplicate previously obtained information and (2) does not accept data that contradicts earlier data. To reflect this understanding, the phrase "in addition to" is deemed sufficient. The plaintiff observes that while some claims specify data submissions by a "user," the specific claims in question do not include this detail. The defendants argue that the specification indicates the supplemental inquiry form "asks the user one or more supplemental questions," leading to user-completed fields. Although a reasonable interpretation suggests that users enter supplemental information, the claim language must prevail. Consequently, the Court will not incorporate a "from a user" limitation in the term "supplemental inquiry form." Additionally, the plaintiff uses "extracted" and the defendants use "obtained" to describe data derived from a source stream. The Court prefers "obtained" due to prior constructions. Thus, "supplemental inquiry form" will be defined as "a form having fields for receiving data in addition to the data actually obtained from the source data stream."
The term "remote communication interface" is referenced in several claims across three patents. For example, the '059 patent outlines processes involving receiving digital data from job applicants and sending structured forms through this interface. The specification describes a "remote communication interface" as consisting of the Internet and relevant hardware/software for communication, including various connection types like local area networks or modem links. Initially, the plaintiff suggested a plain meaning interpretation while the defendants proposed a more technical definition. Following oral arguments, the plaintiff revised their proposal to define "remote communication interface" as "network(s) or components that can transfer data between independent computer systems at separate locations."
Plaintiff argues that the term in question includes both geographic and operational separateness, citing definitions from specifications and technical dictionaries. Key definitions include "communications" as data transfer between computers, "interface" as the connection point for interaction between devices, and "remote" as something not in immediate vicinity but accessible via communications links. The plaintiff emphasizes that other patent claims use "communication interface" without requiring it to be "remote." Defendants propose a simpler definition as a "data connection between two or more devices," asserting that a "remote communication interface" does not necessitate Internet use or specific physical distance. They reference findings from the BPAI during the '561 patent reexamination regarding data transmission to remote devices being disclosed in prior art. The court indicates that neither proposed construction fully captures the term's meaning; defendants' definition omits the concept of "remote," while plaintiffs' interpretation excessively focuses on distance and independence, contrary to the specification that includes Local Area Networks as examples of remote communication interfaces. The court ultimately concludes that the term "remote communication interface" should be defined as “one or more component(s) that can transfer data from a distance between two or more computers or computer systems.”
Additionally, claims 24 through 33 of the '059 patent describe a method for accurately transferring resume information, highlighting the process of a job applicant creating a resume using a word processing application and transmitting it electronically to a processing computer system.
Microsoft Word is cited as an example of a "word processing application" in the specification. The plaintiff argues that the term's plain meaning is sufficient and claims that the defendants are improperly imposing limitations on the claims. If construction is deemed necessary, the plaintiff suggests that the term indicates the invention is designed to accept resumes created with a word-processing application. The defendants argue for a necessary construction, proposing a definition that includes the features of formatting, editing, and saving, asserting that resume generation is an interactive process requiring these capabilities. They support their position with dictionary definitions of "word processing" and "word processor."
However, the court finds that the defendants have not convincingly demonstrated that the term "word processing" requires functional description for lay comprehension. The court also notes that changing "permitting" to "providing an opportunity" adds an unnecessary layer of complexity not evident in the patent. Consequently, the court concludes that the term "permitting a first job applicant to author a resume employing a word processing application" will retain its plain and ordinary meaning.
The court denies the defendants' motion to strike and provides the following constructions for disputed claim terms:
1. "nonuniformly formatted source data streams" means "data streams that vary in format from one stream to the next."
2. "source data string" means "a string of text from a source data stream."
3. "target data string" means "a string of text based on a source data string."
4. "extracting selected ones of said source data strings" means "identifying and isolating selected strings of text from a source data stream."
5. The "enabling" step must precede the "storing" step.
6. "supplemental inquiry form" means "a form having fields for receiving data in addition to the data actually obtained from the source data stream."
7. "remote communication interface" means "one or more components that can transfer data from a distance between two or more computers or computer systems."
The parties had also requested construction for an additional term related to a structured form with multiple fields accommodating data strings.
The term in question is agreed to be defined as a structured form with multiple fields, where at least two fields can contain a data string. The Federal Circuit's decision in Phillips emphasized avoiding reliance on broad dictionary definitions at the outset of claim construction, as this can lead to overly expansive interpretations if the specification's implicit limitations are overlooked. Instead, courts are advised to focus initially on the patentee's use of the term within the claims, specification, and prosecution history. If no specific meaning emerges from this intrinsic evidence, then the common understanding of the words may suffice for construction. Claims 2 through 6 are dependent on claim 1, and while a court's claim construction is not binding on others, it may reference analyses from different district courts on the same patent for uniformity. Reasoned deference may be given to prior constructions of identical terms. The court notes potential challenges in implementing a construction that excludes the preambles of claims, suggesting that it may be more practical to include them. In supplemental discussions, the plaintiff raised concerns about defendants' arguments regarding the timing of resume creation. However, defendants clarified their position, indicating no significant dispute remains regarding the timing or location of the resume, thus rendering further construction unnecessary.
