US 11,887,719 B2
Food knowledge graph for a health tracking system
Hesamoddin Salehian, San Francisco, CA (US); Poojit Sharma, San Francisco, CA (US); Kent Frazier, San Francisco, CA (US); Surender Reddy Yerva, San Francisco, CA (US); Iman Barjasteh, San Francisco, CA (US); and Layla Martin, San Francisco, CA (US)
Assigned to MyFitnessPal, Inc., Austin, TX (US)
Filed by MYFITNESSPAL, INC., San Francisco, CA (US)
Filed on May 20, 2019, as Appl. No. 16/417,291.
Claims priority of provisional application 62/674,170, filed on May 21, 2018.
Prior Publication US 2019/0355465 A1, Nov. 21, 2019
Int. Cl. G16H 20/60 (2018.01); G09B 19/00 (2006.01); G06N 5/02 (2023.01)
CPC G16H 20/60 (2018.01) [G06N 5/02 (2013.01); G09B 19/0092 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A method of operating a health tracking system, the method comprising:
storing in a crowdsourced database of data records, a food knowledge graph defined by a data web comprised of a plurality of nodes and a plurality of connections between the nodes, wherein the nodes of the data web are defined by a plurality of labels describing consumable items and wherein the connections between the nodes are defined by a plurality of relationships between pairs of labels in the plurality of labels, at least some of the labels in the plurality of labels being generic names for consumable items, and wherein, for each pair of labels, one relationship is defined between a first label and a second label of said pair of labels, said one relationship identifying one of (i) the second label as an allergen of the first label, (ii) the second label as a substitute for the first label, and (iii) the second label as a flavor of the first label;
receiving a data record comprising at least one descriptive string regarding a first consumable item from a first health tracking device;
including the received data record within the crowdsourced database of data records;
matching the at least one descriptive string to at least one first label of the pairs of labels in the plurality of labels by implementing a machine learning model in order to match the at least one descriptive string to the at least one first label of the pairs of labels in the plurality of labels;
updating at least one first information field of the data record to associate the data record with the at least one first label to which the at least one descriptive string was matched;
identifying at least one second label in the plurality of labels that is related to the at least one first label based on the plurality of relationships;
updating at least one second information field of the data record to associate the data record with the identified at least one second label;
receiving a request for data records in the crowdsourced database from a second health tracking device;
matching the request for data records from the second health tracking device to the at least one first label or the at least one second label; and
providing the data record having the updated at least one first information field and the at least one second information field to the second health tracking device in response to the request being matched to the at least one first label or the at least one second label, wherein said provided data record is displayed on the second health tracking device along with a menu allowing the user to select a first option for recommended similar items for the provided data record, a second option for recommended substitutes for the provided data record, and a third option for recommended healthier items in lieu of the displayed data record;
after receiving a selection of the first option, generating a first list of one or more additional data records in the crowdsourced database that are associated with a first minimum threshold amount of same descriptive labels as the provided data record, and transmitting the first list to the second health tracking device by wireless transmission and immediately displaying the first list on the second health tracking device following receipt of the selection of the first option;
after receiving a selection of the second option, generating a second list of one or more additional data records in the crowdsourced database that are either (i) identified as substitutes for the provided data record in the food knowledge graph, or (ii) associated with a second minimum threshold amount of the same descriptive labels as the provided data record, wherein the second minimum threshold is greater than the first minimum threshold, and transmitting the second list to the second health tracking device by wireless transmission and immediately displaying the second list on the second health tracking device following receipt of the selection of the second option; and
after receiving a selection of the third option, generating a third list of one or more additional data records in the crowdsourced database that are associated with a third minimum threshold amount of the same descriptive labels as the provided data record and also have a health score in the crowdsourced database that is greater than that of the provided data record, and transmitting the third list to the second health tracking device by wireless transmission and immediately displaying the third list on the second health tracking device following receipt of the selection of the third option.