Spatial Concept Query Based on Lattice-Tree

As a basic method of spatial data operation, keyword query can provide meaningful information to meet user demands by searching textual datasets. How accurately understand users’ intentions and efficiently retrieve results from big are always the focus research. Spatial their complex correlation between features not only enrich connotation objects but also bring difficulties efficient recognition retrieval similar objects. Because there lot many-to-many relationships massive features, most existing research that employ tree-like table-like structures index inefficient in retrieving In this paper, firstly, we define concept (STC) as group with same keywords limited region order present features. Then attempt introduce lattice model maintain related STCs propose hybrid structure, lattice-tree, for data. Lattice-tree employs R-tree location objects, it embeds structure into specific tree nodes organize STC set large number relationships. Based on this, novel query, named Top-k (TkSCQ), answer multiple The empirical study is carried out two sets Yelp Amap. Experiments lattice-tree verify its feasibility demonstrate embed 3 5 levels. TkSCQ evaluate results, keywords, volume, so on, baseline based IR-tree Fp-tree, inverted-tree Fpindex-tree, developed compare Experimental has better efficiency cases, especially case amounts queries, where performance much than Fpindex-tree.