AUTOMATED MEMORY CREATION AND RETRIEVAL FROM MOMENT CONTENT ITEMS
The present embodiments relate to automated memory creation and retrieval from moment content items. In some implementations, the automated memory creation and retrieval system can obtain moment content items from user-designated sources with a single user perspective or multiple user perspectives. The moment content items can be assigned tags and arranged in chronological order. The arranged moment content items can be clustered into memory content items based on clustering conditions. Once memory content items are created, they can be arranged into a memory hierarchy made up of connected nodes. In some implementations, the memory content items are also linked together based on similarity in various dimensions in a memory graph. The automated memory creation and retrieval system can receive search criteria for memories from a user interface and provide the user with memories from matched nodes in the memory hierarchy or linked memories in the memory graph.
1 . A method for organizing memory content items into one or more memory hierarchies, the method comprising:
collecting memory content items, each memory content item comprising one or more moment content items and tagged with semantic identifiers for the memory content item;
arranging the memory content items, according to a chronology, into an ordered set of leaf nodes;
generating the one or more memory hierarchies by iterating through one or more levels, of the one or more memory hierarchies, starting with a level containing the set of leaf nodes, by grouping nodes in the current level into higher-order nodes at a next level by:
identifying sets of sibling nodes that share a threshold amount of the semantic identifiers, wherein the threshold amount is less in each subsequent iteration through the one or more levels;
creating a higher order node for each identified set; and
linking the nodes that make up each set as child nodes of the corresponding created higher-order node; and
providing the one or more memory hierarchies.
2 . The method of claim 1 ,
wherein the iterating through the one or more levels completes when no new nodes were created in a previous iteration; and
wherein, upon the completion of the iterations, each of the highest order nodes created in the previous iteration is established as a root node for a corresponding one of the one or more hierarchies.
3 . The method of claim 1 , wherein each node can be linked as a child to no more than one higher order node.
4 . The method of claim 1 further comprising generating a graph memory structure by:
arranging the memory content items into memory content item pairs; and
iterating through the memory content items pairs wherein, for at least one of the pairs, the method further comprises:
determine one or more tags, in a dimension, that the memory content items in a pair have in common; and
connecting the memory content items in the pair with an edge for the dimension.
5 . The method of claim 4 , wherein the dimension is one of a set of dimensions in which edges are formed in the graph, the set of dimensions comprising: a location dimension, a time dimension, a people dimension, and an activities dimension.
6 . The method of claim 4 , further comprising providing one or more memories as output to a user by:
creating a pseudo memory based on search criteria;
mapping the pseudo memory into the one or more memory hierarchies;
providing one or more memories based on where the pseudo memory is mapped into the one or more memory hierarchies;
receiving a selection of one of the provided one or more memories;
determining, in the graph memory structure, at least one memory connected to the selected one of the provided one or more memories; and
further providing an indication of the determined at least one memory as the output to the user.
7 . The method of claim 1 , wherein the semantic identifiers include: people depicted in one or more of the memory content items, an activity occurring in one or more of the memory content items, a location associated with the memory content item, or any combination thereof.
8 . A computing system for organizing memory content items into one or more memory hierarchies, the computing system comprising:
one or more processors; and
one or more memories storing instructions that, when executed by the one or more processor, cause the computing system to perform a process comprising:
collecting memory content items, each memory content item comprising one or more moment content items and tagged with semantic identifiers;
arranging the memory content items, according to a chronology, into an ordered set of leaf nodes;
generating the one or more memory hierarchies by iterating through one or more levels, of the one or more memory hierarchies, starting with a level containing the set of leaf nodes, by grouping nodes in the current level into higher-order nodes at a next level by:
identifying sets of sibling nodes that share a threshold amount of the semantic identifiers, wherein the threshold amount is less in each subsequent iteration through the one or more levels;
creating a higher order node for each identified set; and
linking the nodes that make up each set as child nodes of the corresponding created higher-order node; and
providing the one or more memory hierarchies.
9 . The computing system of claim 8 ,
wherein the iterating through the one or more levels completes when no new nodes were created in a previous iteration; and
wherein, upon the completion of the iterations, each of the highest order nodes created in the previous iteration is established as a root node for a corresponding one of the one or more hierarchies.
10 . The computing system of claim 8 , wherein each node can be linked as a child to no more than one higher order node.
11 . The computing system of claim 8 , wherein the process further comprises generating a graph memory structure by:
arranging the memory content items into memory content item pairs; and
iterating through the memory content items pairs wherein, for at least one of the pairs, the process further comprises:
determine one or more tags, in a dimension, that the memory content items in a pair have in common; and
connecting the memory content items in the pair with an edge for the dimension.
12 . The computing system of claim 11 , wherein the dimension is one of a set of dimensions in which edges are formed in the graph, the set of dimensions comprising: a location dimension, a time dimension, a people dimension, and an activities dimension.
13 . The computing system of claim 11 , wherein the process further comprises providing one or more memories as output to a user by:
creating a pseudo memory based on search criteria;
mapping the pseudo memory into the one or more memory hierarchies;
providing one or more memories based on where the pseudo memory is mapped into the one or more memory hierarchies;
receiving a selection of one of the provided one or more memories;
determining, in the graph memory structure, at least one memory connected to the selected one of the provided one or more memories; and
further providing an indication of the determined at least one memory as the output to the user.
14 . The computing system of claim 8 , wherein the semantic identifiers include: people depicted in one or more of the memory content items, an activity occurring in one or more of the memory content items, a location associated with the memory content item, or any combination thereof.
15 . A machine-readable storage medium having machine executable instructions stored thereon that, when executed by one or more processors, cause the one or more processors to perform a method for organizing memory content items into one or more memory hierarchies, the method comprising:
collecting memory content items, each memory content item comprising one or more moment content items and tagged with semantic identifiers;
arranging the memory content items, according to a chronology, into an ordered set of leaf nodes;
generating the one or more memory hierarchies by iterating through one or more levels, of the one or more memory hierarchies, starting with a level containing the set of leaf nodes, by grouping nodes in the current level into higher-order nodes at a next level by:
identifying sets of sibling nodes that share a threshold amount of the semantic identifiers, wherein the threshold amount is less in each subsequent iteration through the one or more levels;
creating a higher order node for each identified set; and
linking the nodes that make up each set as child nodes of the corresponding created higher-order node; and
providing the one or more memory hierarchies.
16 . The machine-readable storage medium of claim 15 wherein the iterating through the one or more levels completes when no new nodes were created in a previous iteration.
17 . The machine-readable storage medium of claim 15 , wherein each node can be linked as a child to no more than one higher order node.
18 . The machine-readable storage medium of claim 15 , wherein the method further comprises generating a graph memory structure by:
arranging the memory content items into memory content item pairs; and
iterating through the memory content items pairs wherein, for at least one of the pairs, the method further comprises:
determine one or more tags, in a dimension, that the memory content items in a pair have in common; and
connecting the memory content items in the pair with an edge for the dimension.
19 . The machine-readable storage medium of claim 18 , wherein the dimension is one of a set of dimensions in which edges are formed in the graph, the set of dimensions comprising: a location dimension, a time dimension, a people dimension, and an activities dimension.
20 . The machine-readable storage medium of claim 15 , wherein the method further comprises providing one or more memories as output to a user by:
creating a pseudo memory based on search criteria;
mapping the pseudo memory into the one or more memory hierarchies; and
providing one or more memories based on where the pseudo memory is mapped into the one or more memory hierarchies.