Abstract: A pulley device includes a strap loop, a cord-securing component and cord respectively attached to the strap loop, a pulley component, the cord attached to the pulley component. The pulley component includes a housing, and may include a wheel that rotates within the housing. The wheel may have a circumferential groove or flat circumferential side to accommodate a cord or belt. The cord may form a loop, enabling the cord loop to be secured on the cord-securing component so that the pulley device may be worn on a person's wrist.

Abstract: A pulley device includes a strap loop, a cord-securing component and cord respectively attached to the strap loop, a pulley component, the cord attached to the pulley component. The pulley component includes a housing, and may include a wheel that rotates within the housing. The wheel may have a circumferential groove or flat circumferential side to accommodate a cord or belt. The cord may form a loop, enabling the cord loop to be secured on the cord-securing component so that the pulley device may be worn on a person's wrist.

Abstract: Information about physical places is gathered to develop quality scores for targeting advertisements to users of mobile devices sharing geographic location events with service providers. A scoring model incorporates selected attributes about places (e.g., created by a service provider vs. user-generated, category, keywords, census data providing population density by location, property values by area, popularity, special events, government vs. private area, whether place is associated with a brand, chain, or is independent) that have been gathered from multiple sources, verified, and normalized. Advertisements are targeted to users running mobile applications based on a location event shared with the service provider using the quality score for the place indicated in the location event in addition to other attributes of the place. In one embodiment, the system uses a machine learning algorithm to analyze conversion rates of targeted advertising to provide feedback to the scoring model.

Abstract: Information about physical places is gathered to develop quality scores for targeting advertisements to users of mobile devices sharing geographic location events with service providers. A scoring model incorporates selected attributes about places (e.g., created by a service provider vs. user-generated, category, keywords, census data providing population density by location, property values by area, popularity, special events, government vs. private area, whether place is associated with a brand, chain, or is independent) that have been gathered from multiple sources, verified, and normalized. Advertisements are targeted to users running mobile applications based on a location event shared with the service provider using the quality score for the place indicated in the location event in addition to other attributes of the place. In one embodiment, the system uses a machine learning algorithm to analyze conversion rates of targeted advertising to provide feedback to the scoring model.

Abstract: Information about physical places is gathered to develop quality scores for targeting advertisements to users of mobile devices sharing geographic location events with service providers. A scoring model incorporates selected attributes about places (e.g., created by a service provider vs. user-generated, category, keywords, census data providing population density by location, property values by area, popularity, special events, government vs. private area, whether place is associated with a brand, chain, or is independent) that have been gathered from multiple sources, verified, and normalized. Advertisements are targeted to users running mobile applications based on a location event shared with the service provider using the quality score for the place indicated in the location event in addition to other attributes of the place. In one embodiment, the system uses a machine learning algorithm to analyze conversion rates of targeted advertising to provide feedback to the scoring model.