Big Data, Linked Data
Classification Research at the Junction
In Montréal, on 2 November, Richard Smiraglia participated in the Association for Information Science and Technology’s Classification Research workshop, “Big Data, Linked Data: Classification Research at the Junction.”
Eleven papers were presented exploring the interaction between data-mining and knowledge organization, including papers on objectivity or truthfulness of big data, semantic similarity for mining tweets, social contextualization of big data, and warrant as theoretical vocabulary for big data classification. Several papers explored statistical techniques for data-mining large bibliographic files using knowledge organization systems to create mining pathways.
Richard’s paper “Big Classification: Using the Empirical Power of Classification Interaction” used a random sample of the nine million Universal Decimal Classification numbers from the OCLC WorldCat (from the Knowledge Space Lab project) to demonstrate statistically how deconstructed classification components can be used to predict the co-occurrence of structural bibliographic elements. For more information you can read the preliminary program