Separating Wheat from the Chaff – A Relationship Ranking Algorithm

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9989)


We address the problem of ranking relationships in an automatically constructed knowledge graph. We propose a probabilistic ranking mechanism that utilizes entity popularity, entity affinity, and support from text corpora for the relationships. Results obtained from preliminary experiments on a standard dataset are encouraging and show that our proposed ranking mechanism can find more informative and useful relationships compared to a frequency based approach.


Knowledge Graph Ent Input Target Entity Domain-specific Corpus Large Human Effort 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Castelli, V., Raghavan, H., Florian, R., Han, D.-J., Luo, X., Roukos, S.: Distilling and exploring nuggets from a corpus. In: SIGIR, p. 1006 (2012)Google Scholar
  2. 2.
    Hoffart, J., Seufert, S., Nguyen, D.B., Theobald, M., Weikum, G.: Kore: Keyphrase overlap relatedness for entity disambiguation. In: CIKM 2012, pp. 545–554 (2012)Google Scholar
  3. 3.
    Li, X., Li, C., Yu, C.: Entity-relationship queries over wikipedia. ACM Trans. Intell. Syst. Technol. 3(4), 70:1–70:20 (2012)CrossRefGoogle Scholar
  4. 4.
    Nagarajan, M. et al.: Predicting future scientific discoveries based on a networked analysis of the past literature. In: KDD, pp. 2019–2028 (2015)Google Scholar
  5. 5.
    Zhang, Y., Cheng, G., Qu, Y.: JIST 2013, Selected Papers, chapter Towards ExploratoryRelationship Search: A Clustering-Based Approach, pp. 277–293 (2014)Google Scholar

Copyright information

© Springer International Publishing AG 2016

Authors and Affiliations

  1. 1.IBM Watson, Almaden Research CentreSan JoseUSA

Personalised recommendations