Abstract | ||
---|---|---|
The Web hosts a huge variety of multi-cultural taxonomies. They encompass product catalogs of e-commerce, general-purpose knowledge bases and numerous domain-specific category systems. The \"common denominator\" of all these is their enormous diversity, which makes it infeasible to combine multiple taxonomies for ad-hoc tasks. To support the alignment of independently created Web taxonomies, we introduce the ACROSS framework. For mapping categories across different taxonomies, ACROSS harnesses instance-level features as well as distant supervision from an intermediate source like multiple Wikipedia editions. Our experiments with heterogeneous taxonomies for different domains demonstrate the viability of our approach and improvement over state-of-the-art baselines. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1145/2908131.2908164 | WebSci |
Keywords | Field | DocType |
Multicultural Knowledge Bases, Knowledge Taxonomies, Alignment Methods, Integer Linear Programming, Integer Quadratic Programming, Label Propagation | Astronomy,Information retrieval,Computer science,Label propagation,Baseline (configuration management),Integer programming,Quadratic integer programming,Artificial intelligence,Fraction (mathematics),Machine learning | Conference |
Citations | PageRank | References |
3 | 0.45 | 27 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Natalia Boldyrev | 1 | 5 | 1.16 |
Marc Spaniol | 2 | 897 | 61.13 |
Gerhard Weikum | 3 | 12710 | 2146.01 |