Position recommendation method and apparatus

Abstract

Embodiments of the invention disclose a position recommendation method and apparatus, which relate to the technical field of internet and can ensure the quality of position recommendation while ensuring that time information in check-in metadata is not lost. The specific scheme is that: the method comprises: obtaining check-in information of user equipment to a position; calculating user similarity between a predetermined user and other user according to the check-in information; obtaining a prior probability of check-in of the predetermined user in a predetermined position according to the user similarity; calculating position similarity between the predetermined position and other position according to the check-in information; constructing a continuous time probability density function of check-in of the predetermined user in the predetermined position according to the user similarity and the position similarity; calculating an access probability of access of the predetermined user to each predetermined position within a preset time period according to the prior probability and the continuous time probability density function; and according to the access probability, recommending an access position for the predetermined user within the preset time period according to a predetermined rule. The position recommendation method and apparatus provided by the embodiments of the invention are used for position recommendation.

Description

technical field [0001] The present invention relates to the technical field of the Internet, in particular to a position recommendation method and device. Background technique [0002] Location-based services and online social networks are two mainstream applications of the mobile Internet. The integration of these two mainstream applications produces location-based social networks (location-based social networks), referred to as location-based social networks. In the location social network, users can establish social links (sociallinks), and check-in (check-in) their own location, especially some interesting locations, through mobile devices with positioning functions (such as mobile phones with GPS). Such as restaurants, shops, museums, etc.; write a review to share the experience of visiting (checking in) a location. [0003] Since location social networks contain a large number of locations, location recommendation technology can make it easier for users to find locati...

AI Technical Summary

Problems solved by technology

If the fine-grained time period method is used to alleviate the loss of time information, the quality of location recommendation will be low because the user check-in metadata contained in the fine-grained time period is very sparse.

[0006] Therefore, the current location recommendation technology cannot ensure the integrity of the time information in the check-in metadata while taking into account the quality of the location recommendation

Images