Abstract:
In one embodiment, techniques are provided to establish and use semantic associations between location profiles and ambient profiles. One or more location profiles are selected from a location database. A first plurality of ambient profiles is selected for a first area surrounding one or more geographic locations of the location profiles. One or more patterns are extracted from the first plurality of ambient profiles and are used to generate associations between location profiles and ambient profiles in an association database which semantically associates location profiles with ambient profiles independent of geographic location. The associations may be used, among other things, to service requests from mobile devices and/or update ambient profiles or location profiles.
Abstract:
In one embodiment, techniques are provided to establish and use semantic associations between location profiles and ambient profiles. One or more location profiles are selected from a location database. A first plurality of ambient profiles is selected for a first area surrounding one or more geographic locations of the location profiles. One or more patterns are extracted from the first plurality of ambient profiles and are used to generate associations between location profiles and ambient profiles in an association database which semantically associates location profiles with ambient profiles independent of geographic location. The associations may be used, among other things, to service requests from mobile devices and/or update ambient profiles or location profiles.
Abstract:
In one example embodiment, an analysis application implements a technique for measuring a property of interest of an input dataset of location samples. The analysis application may process the input dataset in one or more multi-stage pipelines to produce values that measure a metric for the input dataset. The values that measure the metric for the input dataset may be compared with values that measure the metric for one or more labeled datasets that are generated by the analysis application, for example, using a feedback loop and sampling from a plurality of delta data sets. Each labeled datasets may have different values that measure the metric and corresponding measures of the property of interest. Based on the comparison, the analysis application may determine the measure of the property of interest for the input dataset and such measure may be returned, for example, to a remote device.
Abstract:
In one example embodiment, an analysis application implements a technique for measuring a property of interest of an input dataset of location samples. The analysis application may process the input dataset in one or more multi-stage pipelines to produce values that measure a metric for the input dataset. The values that measure the metric for the input dataset may be compared with values that measure the metric for one or more labeled datasets that are generated by the analysis application, for example, using a feedback loop and sampling from a plurality of delta data sets. Each labeled datasets may have different values that measure the metric and corresponding measures of the property of interest. Based on the comparison, the analysis application may determine the measure of the property of interest for the input dataset and such measure may be returned, for example, to a remote device.
Abstract:
In various embodiments, techniques are provided for determining one or more zones in which mobile devices are presently located and identifying or updating characteristics of on or more zones. Samples that include beacon information and/or sensor information collected by mobile devices are aggregated and dynamically organized into sample classes that are associated with zero, one or more zones. A venue is characterized by a set of zones and associated tags, which may be informed based on samples for the venue, a venue group to which the venue belongs, or all venues. To determine if a mobile device is located in one or more zones, the samples are compared to zone characteristics, and based thereon (and optionally history information) one or more zones are selected having determined likelihoods, and at least a zone having the highest likelihood of the one or more selected zones is returned.
Abstract:
In one embodiment, techniques are provided to establish and use semantic associations between location profiles and ambient profiles. One or more location profiles are selected from a location database. A first plurality of ambient profiles is selected for a first area surrounding one or more geographic locations of the location profiles. One or more patterns are extracted from the first plurality of ambient profiles and are used to generate associations between location profiles and ambient profiles in an association database which semantically associates location profiles with ambient profiles independent of geographic location. The associations may be used, among other things, to service requests from mobile devices and/or update ambient profiles or location profiles.
Abstract:
In one embodiment, techniques are provided to establish and use semantic associations between location profiles and ambient profiles. One or more location profiles are selected from a location database. A first plurality of ambient profiles is selected for a first area surrounding one or more geographic locations of the location profiles. One or more patterns are extracted from the first plurality of ambient profiles and are used to generate associations between location profiles and ambient profiles in an association database which semantically associates location profiles with ambient profiles independent of geographic location. The associations may be used, among other things, to service requests from mobile devices and/or update ambient profiles or location profiles.