Deep learning machine vision to analyze localities for comparative spending analyses
A system, method, and computer-readable storage medium are disclosed that execute machine vision operations to categorize a locality. At least one embodiment accesses a map image of a locality, where the map image includes geographical artefacts corresponding to entities within the locality; analyzes the map image to detect the entities in the locality using the geographical artefacts; assigns entity classes to detected entities in the locality; assigns a locality score to the locality based on entity classes included in the locality; retrieves street view images for one or more of the detected entities in the locality; and analyzes street view images of the detected entities to assign one or more further classifications to the detected entities. Other embodiments include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the method.
1. A computer-implemented method for using machine vision to categorize a locality to conduct comparative spending analyses, the method comprising:
assigning locality profile scores to a plurality of localities based on entity classes included in the localities, wherein the locality profile scores are derived through neural network analyses of map images for the localities;
grouping localities having similar locality profile scores into separate locality groups;
extracting entities in a locality group;
retrieving historical data for entities in the locality group;
grouping spending of the entities in the locality group by product areas;
ranking spending of the entities in the locality group; and
providing the spending per product area of the entities and ranks of the spending of the entities in the locality group for a comparative spending analysis.
2. The computer-implemented method of claim 1 , wherein grouping localities having similar locality profile scores comprises:
determining a Euclidean distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Euclidean distance metric below a predetermined threshold.
3. The computer-implemented method of claim 1 , wherein grouping localities having similar locality profile scores comprises:
determining a Manhattan distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Manhattan distance metric below a predetermined threshold.
4. The computer-implemented method of claim 1 , wherein
entities within a group are divided into one or more subgroups based on entity identification data in the historical data.
5. The computer-implemented method of claim 1 , further comprising:
accessing a map image of a locality, wherein the map image includes geographical artefacts corresponding to entities within the locality;
analyzing the map image to detect the entities in the locality using the geographical artefacts;
assigning entity classes to detected entities in the locality; and
assigning the locality profile score to the locality based on entity classes included in the locality.
6. The computer-implemented method of claim 5 , wherein
analyzing the map image includes
providing the map image to a neural network, wherein the neural network is configured to separate portions of the map image having geographical artefacts from portions of the map image that do not have geographical artefacts;
reconstructing the map image using the portions of the map image having geographical artefacts;
detecting entities in the reconstructed map using the geographical artefacts;
identifying entity classes of the detected entities by accessing a database for the locality to identify entities using textual portions of the reconstructed map image;
wherein assigning a locality profile score includes
clustering entities with similar entity classes within the locality, wherein each cluster has a cluster type corresponding to the entity classes within the cluster; and
assigning a locality profile score to the locality based on the clustered entities within the locality.
7. The computer-implemented method of claim 6 , wherein
locality profile scores are assigned to the locality using a percentage based analysis, wherein a percentage is assigned to each cluster type based on a number of entities included in the cluster type with respect to a total number of clustered entities.
8. A system comprising:
a processor;
a data bus coupled to the processor; and
a non-transitory, computer-readable storage medium embodying computer program code, the non-transitory, computer-readable storage medium being coupled to the data bus, the computer program code interacting with a plurality of computer operations and comprising instructions executable by the processor and configured for:
assigning locality profile scores to a plurality of localities based on entity classes included in the localities, wherein the locality profile scores are derived through neural network analyses of map images for the localities;
grouping localities having similar locality profile scores into separate locality groups;
extracting entities in a locality group;
retrieving historical data for entities in the locality group;
grouping spending of the entities in the locality group by product areas;
ranking spending of the entities in the locality group; and
providing the spending per product area of the entities and ranks of the spending of the entities in the locality group for a comparative spending analysis.
9. The system of claim 8 , wherein grouping localities having similar locality profile scores comprises:
determining a Euclidean distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Euclidean distance metric below a predetermined threshold.
10. The system of claim 8 , wherein grouping localities having similar locality profile scores comprises:
determining a Manhattan distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Manhattan distance metric below a predetermined threshold.
11. The system of claim 8 , wherein
entities within a group are divided into one or more subgroups based on entity identification data in the historical data.
12. The system of claim 8 , wherein the instructions are further configured for:
accessing a map image of a locality, wherein the map image includes geographical artefacts corresponding to entities within the locality;
analyzing the map image to detect the entities in the locality using the geographical artefacts;
assigning entity classes to detected entities in the locality; and
assigning the locality profile score to the locality based on entity classes included in the locality.
13. The system of claim 12 , wherein
analyzing the map image includes
providing the map image to a neural network, wherein the neural network is configured to separate portions of the map image having geographical artefacts from portions of the map image that do not have geographical artefacts;
reconstructing the map image using the portions of the map image having geographical artefacts;
detecting entities in the reconstructed map using the geographical artefacts;
identifying entity classes of the detected entities by accessing a database for the locality to identify entities using textual portions of the reconstructed map image;
wherein assigning a locality profile score includes
clustering entities with similar entity classes within the locality, wherein each cluster has a cluster type corresponding to the entity classes within the cluster; and
assigning a locality profile score to the locality based on the clustered entities within the locality.
14. The system of claim 13 , wherein
locality profile scores are assigned to the locality using a percentage based analysis, wherein a percentage is assigned to each cluster type based on a number of entities included in the cluster type with respect to a total number of clustered entities.
15. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:
assigning locality profile scores to a plurality of localities based on entity classes included in the localities, wherein the locality profile scores are derived through neural network analyses of map images for the localities;
grouping localities having similar locality profile scores into separate locality groups;
extracting entities in a locality group;
retrieving historical data for entities in the locality group;
grouping spending of the entities in the locality group by product areas;
ranking spending of the entities in the locality group; and
providing the spending per product area of the entities and ranks of the spending of the entities in the locality group for a comparative spending analysis.
16. The non-transitory, computer-readable storage medium of claim 15 , wherein grouping localities having similar locality profile scores comprises:
determining a Euclidean distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Euclidean distance metric below a predetermined threshold.
17. The non-transitory, computer-readable storage medium of claim 15 , wherein grouping localities having similar locality profile scores comprises:
determining a Manhattan distance metric between locality profile scores of the plurality of localities; and
grouping localities having a Manhattan distance metric below a predetermined threshold.
18. The non-transitory, computer-readable storage medium of claim 15 , wherein entities within a group are divided into one or more subgroups based on entity identification data in the historical data.
19. The non-transitory, computer-readable storage medium of claim 15 , wherein the instructions are further configured for:
accessing a map image of a locality, wherein the map image includes geographical artefacts corresponding to entities within the locality;
analyzing the map image to detect the entities in the locality using the geographical artefacts;
assigning entity classes to detected entities in the locality; and
assigning the locality profile score to the locality based on entity classes included in the locality.
20. The non-transitory, computer-readable storage medium of claim 19 , wherein
analyzing the map image includes
providing the map image to a neural network, wherein the neural network is configured to separate portions of the map image having geographical artefacts from portions of the map image that do not have geographical artefacts;
reconstructing the map image using the portions of the map image having geographical artefacts;
detecting entities in the reconstructed map using the geographical artefacts;
identifying entity classes of the detected entities by accessing a database for the locality to identify entities using textual portions of the reconstructed map image;
wherein assigning a locality profile score includes
clustering entities with similar entity classes within the locality, wherein each cluster has a cluster type corresponding to the entity classes within the cluster; and
assigning a locality profile score to the locality based on the clustered entities within the locality.