# US 8,788,507 B1 · Watermarking of structured results and watermark generation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 8,788,507 B1 - Document ID: US8788507B1 - Status: Granted - Kind: B1 - Application number: 13/296,451 - Filed: 2011-11-15 - Granted: 2014-07-22 - Earliest claimed: 2011-03-30 - Assignees: Google Inc. - Applicants: Ashish Venugopal; Jurij Ganitkevic; Franz Josef Och; David Robert Talbot; Jakob David Uszkoreit - Inventors: Ashish Venugopal; Jurij Ganitkevic; Franz Josef Och; David Robert Talbot; Jakob David Uszkoreit - Web page: https://patentgrep.com/patent/US8788507B1 ## CPC - G06F 16/9577: {Optimising the visualization of content, e.g. distillation of HTML documents} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract A way of generating a watermark for a structured result, such as a search result or a machine translation. A hash function is used to generate a bit sequence for each of a plurality of structured results. A ranking score is generated for each resulting bit sequence. The ranking score can be based on the detectability of the bit sequence compared to a randomly-generated bit sequence and the quality of each of the structured results. A structured result is selected as the watermarked structured result based upon the ranking score. ## Claims ### Claim 1 A method, comprising: receiving a plurality of alternative search results based upon a query from a user; computing a hash of at least part of each alternative search result to produce a bit sequence corresponding to each alternative search result; computing a first ranking score and a second ranking score for each alternative search result based upon its corresponding hash, the first ranking score for a specific alternative search result based upon a deviation of a distribution of binary digits in its corresponding hash from a binary distribution in which a parameter p=0.5 and the second ranking score for a specific alternative search result based upon a quality of the specific alternative search result; and selecting one of the plurality of alternative search results based upon the ranking score to obtain a watermarked search result, the watermarked search result having been determined to include machine-generated content that is not appropriate for machine learning by a language translation model. ### Claim 2 (depends on 1) The method of claim 1, wherein the first ranking score and the second ranking score comprise a single ranking score. ### Claim 3 A method, comprising: computing a hash of at least part of each of a plurality of alternative structured results to produce a bit sequence corresponding to each alternative structured result; computing a ranking score for each alternative structured result based upon a distribution of binary digits in its corresponding bit sequence; and selecting one of the plurality of alternative structured results based upon the ranking score to obtain a watermarked structured result, the watermarked structured result having been determined to include machine-generated content that is not appropriate for machine learning by a language translation model. ### Claim 4 (depends on 3) The method of claim 3, wherein the watermarked structured result is a search result. ### Claim 5 (depends on 3) The method of claim 3, wherein the watermarked structured result is a machine-generated translation. ### Claim 6 (depends on 3) The method of claim 3, wherein computing the hash to produce the bit sequence for a specific alternative structured result comprises: mapping the specific alternative structured result to a set of sub-results; computing a hash of least some of the sub-results; and concatenating at least some of the bit sequences for the sub-results for which the hashes are computed to produce the bit sequence corresponding to the specific alternative structured result. ### Claim 7 (depends on 3) The method of claim 3, wherein the ranking score for a specific alternative structured result is based upon a deviation of a distribution of binary digits in its corresponding hash from a binomial distribution with a parameter p=0.5. ### Claim 8 (depends on 3) The method of claim 3, wherein the ranking score is further based upon a quality measure of each alternative structured result. ### Claim 9 (depends on 3) The method of claim 3, wherein the ranking score for a specific alternative structured result is based on a number of occurrences of a given element in its hash in relation to a total number of elements in its hash. ### Claim 10 A system, comprising: a database storing a plurality of alternative structured results; a processor in connection with the database, the processor configured to: compute a hash of at least part of each alternative structured result to produce a bit sequence corresponding to each alternative structured result; compute a ranking score for each alternative structured result based upon a distribution of binary digits in its corresponding bit sequence; and select one of the plurality of alternative structured results based upon the ranking score to obtain a watermarked structured result, the watermarked structured result having been determined to include machine-generated content that is not appropriate for machine learning by a language translation model. ### Claim 11 (depends on 10) The system of claim 10, wherein the watermarked structured result is a search result. ### Claim 12 (depends on 10) The system of claim 10, wherein the watermarked structured result is a machine-generated translation. ### Claim 13 (depends on 10) The system of claim 10, wherein the processor is further configured to: map a specific alternative structured result to a set of sub-results; compute a hash of least some of the sub-results; and concatenate at least some of the bit sequences for the sub-results for which the hashes are computed to produce the bit sequence corresponding to the specific alternative structured result. ### Claim 14 (depends on 10) The system of claim 10, wherein the ranking score for a specific alternative structured result is based upon a deviation of a distribution of binary digits in its corresponding hash from a binomial distribution with a parameter p=0.5. ### Claim 15 (depends on 10) The system of claim 10, wherein the ranking score is further based upon a quality measure of each alternative structured result. ### Claim 16 (depends on 10) The system of claim 10, wherein the ranking score for a specific alternative structured result is based on a number of occurrences of a given element in its hash in relation to a total number of elements in its hash.