Abstract | ||
---|---|---|
In intelligent systems, belief function is a common approach to modelling uncertain and imperfect information obtained constantly while knowledge bases are used to encapsulate multiple static information items. In the literature, many different approaches have been proposed for combining belief functions (resp. merging knowledge bases) when there are multiple belief functions (resp. knowledge bases) generated from different sources depicting the same issue of interest. However, the connection between belief function combination and knowledge base merging is not adequately explored. In this paper, we aim to study the correspondence between the two lines of research. By introducing a numerical characteristic function for each knowledge base, we show that there is one to one correspondence between a set of combination rules of belief functions and merging methods of weighted knowledge bases. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.ijar.2018.09.012 | International Journal of Approximate Reasoning |
Keywords | Field | DocType |
Knowledge bases,Evidential characteristic function,Merging,Evidence theory,Evidential reasoning | Bijection,Intelligent decision support system,Characteristic function (probability theory),Theoretical computer science,Artificial intelligence,Knowledge base,Perfect information,Merge (version control),Machine learning,Mathematics | Journal |
Volume | Issue | ISSN |
104 | 1 | 0888-613X |
Citations | PageRank | References |
0 | 0.34 | 18 |
Authors | ||
1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jianbing Ma | 1 | 242 | 15.73 |