# US 11,907,133 B2 · Standardized addresses generation from address substrings > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,907,133 B2 - Document ID: US11907133B2 - Status: Granted - Kind: B2 - Application number: 17/877,443 - Filed: 2022-07-29 - Published: 2023-05-25 - Granted: 2024-02-20 - Earliest claimed: 2021-11-19 - Assignees: SafeGraph, Inc. - Applicants: SafeGraph, Inc. - Inventors: Vera Sazonova - Web page: https://patentgrep.com/patent/US11907133B2 ## CPC - G06F 12/10: Address translation - G06F 40/45: Example-based machine translation; Alignment - G06F 16/215: Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors - G06F 40/103: Formatting, i.e. changing of presentation of documents (automatic justification G06F40/189; automatic line break hyphenation G06F40/191) - G06F 40/279: Recognition of textual entities - G06F 40/295: Named entity recognition - G06F 2212/7201: Logical to physical mapping or translation of blocks or pages - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Standardized address generation from address substrings includes receiving an address string for a place-of-interest, one-to-many mapping at least one of a plurality of address substrings of the address string to respective address components, concatenating the address substrings using a template that specifies an order of concatenating the address substrings, and making the concatenated address substrings available for further use. ## Claims ### Claim 1 A computer-implemented method, including: receiving an address string for a place-of-interest, wherein the address string has a plurality of address substrings; one-to-one mapping respective address substrings in the address string to respective address components; one-to-many mapping at least one of the address substrings to two or more address components, and partitioning a one-to-many mapped address substring into address substrings that correspond to respective address components; using a template that specifies an order of concatenating the address substrings, and concatenating the address substrings according to the specified order; and making the address substrings concatenated in the specified order available for further use, wherein a model processes the address substrings as input through a convolutional neural network and generates an output including one-to-one mapping of each of the address substrings to respective address components, wherein the convolutional neural network uses trigram layers and residual connections to mix information from words on either side of a target word so that an output word is of a same dimensionality as the target word. ### Claim 2 (depends on 1) The method of claim 1, wherein the address components include a primary number address component, a street name address component, a street suffix address component, a street postdirection address component, a unit type address component, a unit number address component, a city address component, a state address component, a country address component, and a postal code address component. ### Claim 3 (depends on 1) The method of claim 1, further including: prior to using the template, normalizing the address substrings according to address suffix abbreviation rules. ### Claim 4 (depends on 1) The method of claim 1, further including: prior to using the template, using positions of one-to-one mapped address substrings in the address string to detect a non-compliant address string; mapping the non-compliant address string to a particular pre-defined non-compliant signature in a plurality of pre-defined non-compliant signatures; and applying a resolution specific to the particular pre-defined non-compliant signature to transform the non-compliant address string to a compliant address string. ### Claim 5 (depends on 4) The method of claim 4, wherein a pre-defined non-compliant signature in the plurality of pre-defined non-compliant signatures includes at least two address substrings positioned adjacent to each other in the address string and each address substring is one-to-one mapped to a street suffix address component. ### Claim 6 (depends on 4) The method of claim 4, wherein a pre-defined non-compliant signature in the plurality of pre-defined non-compliant signatures includes at least two address substrings positioned adjacent to each other in the address string and each address substring is one-to-one mapped to a same address component. ### Claim 7 (depends on 4) The method of claim 4, wherein a pre-defined non-compliant signature in the plurality of pre-defined non-compliant signatures includes at least two address substrings at random positions in the address string and each address substring is one-to-one mapped to a same address component. ### Claim 8 (depends on 4) The method of claim 4, wherein a pre-defined non-compliant signature in the plurality of pre-defined non-compliant signatures includes at least two address substrings in the address string separated by a separator token indicating an intersection location. ### Claim 9 (depends on 8) The method of claim 8, wherein the separator token includes at least one of an ‘&’ symbol or an ‘and’ substring. ### Claim 10 (depends on 8) The method of claim 8, wherein at least one address substring in the at least two address substrings further comprises at least two address substrings positioned adjacent to each other wherein a first address substring is one-to-one mapped to a street name address component and a second address substring is one-to-one mapped to a street suffix address component. ### Claim 11 (depends on 1) The method of claim 1, wherein the template uses coding accuracy support system (or CASS) address format. ### Claim 12 (depends on 1) The method of claim 1, wherein the template is used on a per-jurisdiction basis. ### Claim 13 A system including one or more processors coupled to memory, the memory loaded with computer instructions that when executed on the processors, implement actions comprising: receiving an address string for a place-of-interest, wherein the address string has a plurality of address substrings; one-to-one mapping respective address substrings in the address string to respective address components; one-to-many mapping at least one of the address substrings to two or more address components, and partitioning a one-to-many mapped address substring into address substrings that correspond to respective address components; using a template that specifies an order of concatenating the address substrings, and concatenating the address substrings according to the specified order; and making the address substrings concatenated in the specified order available for further use, wherein a model processes the address string as input through a convolutional neural network and generates a substring-by-substring output that classifies each of the output substrings as the address substrings and non-address substrings, wherein the convolutional neural network uses trigram layers and residual connections to mix information from words on either side of a target word so that an output word is of a same dimensionality as the target word. ### Claim 14 (depends on 13) The system of claim 13, wherein the address components include a primary number address component, a street name address component, a street suffix address component, a street postdirection address component, a unit type address component, a unit number address component, a city address component, a state address component, a country address component, and a postal code address component. ### Claim 15 (depends on 13) The system of claim 13, further implementing actions comprising: prior to using the template, using positions of one-to-one mapped address substrings in the address string to detect a non-compliant address string; mapping the non-compliant address string to a particular pre-defined non-compliant signature in a plurality of pre-defined non-compliant signatures; and applying a resolution specific to the particular pre-defined non-compliant signature to transform the non-compliant address string to a compliant address string. ### Claim 16 A non-transitory computer readable storage medium impressed with computer program instructions, the instructions, when executed on a processor, implement a method comprising: receiving an address string for a place-of-interest, wherein the address string has a plurality of address substrings; one-to-one mapping respective address substrings in the address string to respective address components; one-to-many mapping at least one of the address substrings to two or more address components, and partitioning a one-to-many mapped address substring into address substrings that correspond to respective address components; using a template that specifies an order of concatenating the address substrings, and concatenating the address substrings according to the specified order; and making the address substrings concatenated in the specified order available for further use, wherein a model processes the address string as input through a convolutional neural network and generates a substring-by-substring output that classifies each of the output substrings as the address substrings and non-address substrings, wherein the convolutional neural network uses trigram layers and residual connections to mix information from words on either side of a target word so that an output word is of a same dimensionality as the target word. ### Claim 17 (depends on 16) The non-transitory computer readable storage medium of claim 16, the implementing the method further comprising: prior to using the template, using positions of one-to-one mapped address substrings in the address string to detect a non-compliant address string; mapping the non-compliant address string to a particular pre-defined non-compliant signature in a plurality of pre-defined non-compliant signatures; and applying a resolution specific to the particular pre-defined non-compliant signature to transform the non-compliant address string to a compliant address string.