针对传统的时空轨迹相似性度量算法中存在的计算复杂度高且不适于增量计算的问题，提出了基于相似哈希计算用户时空轨迹相似度的方法，同时使用逆轨迹频率（ ITF）度量位置流行度对轨迹相似性的影响，将用户的历史轨迹编码为二进制轨迹指纹，并根据海明距离判断轨迹指纹之间的相似性，使得相似性计算可以在线性时间内完成；此外，改进了地点相似性算法，并将轨迹相似度和地点相似度相结合提出了基于地点和轨迹相似性的地点推荐算法。实验结果表明，本文的推荐方法在准确率、召回率和覆盖率方面能够取得较好的推荐效果，验证了所提方法的有效性。%In order to solve the problem of the high computational complexity and inapplicability to incremental com⁃puting of traditional spatial⁃temporal trajectory similarity measurements, in this paper we propose a simHash⁃based method to measure the similarity between different users'spatial⁃temporal trajectories, which also consider the influ⁃ence of location popularity on the trajectories'similarities by using the locations'inverse trajectory frequency ( ITF) . With this method, users' trajectories are initially transformed into binary trajectory fingerprints. We use the Ham⁃ming distance to determine the similarity of the users' trajectories, and the similarity calculation can be finished within linear time. In addition, we propose an improved location similarity algorithm and combine the location simi⁃larity with the trajectory similarity to generate interesting location recommendations. Compared with the existing method, the experimental results verify the effectiveness of the proposed method and demonstrate that it has better performance with respect to precision, recall, and coverage.
展开▼
