融合地理标签数据的个性化信息服务应用研究

Research on Application of Personalized Information Service Combine with Geo-tag Data

[目的/意义]社交媒体网站的飞速发展为我们贡献了海量数据,通过对这些数据的进一步挖掘,可以实现个性化服务推荐。[方法/过程]本文利用地理标签中的丰富的元数据信息,结合基于密度的DBSCAN聚类算法和TF-IDF的统计方法,来提取和识别当地的景点区域,然后结合季节来计算景点的热度,最后运用基于混合过滤的推荐算法,为游客实现个性化旅游服务推荐。[结果/结论]通过Flickr网站爬取到的Geo-tagged数据集验证了本文提出方法的有效性。

[Purpose/Significance] The rapid development of social media websites has contributed a lot of data to us.Through further mining of these data,we can realized the personalized recommendation’s service.[Method/Process]This paper used the metadata information in the geo-tagged dataset,combined with the DBSCAN clustering algorithm based on density and the TF-IDF statistical method to extract and identify the local spots,and then calculated the hot of the local spots in combination with the seasons.At last,we used Hybrid filtering algorithm to achieve personalized travel recommend services.[Result/Conclusion]The Geo-tagged dataset from the Flickr website proved the effectiveness of the paper’s extraction method.