Achieving Classification and Clustering in One Shot Lesson Learned from Labeling Anonymous Datasets

摘要

This paper presents an algorithm LadsComplete which is able to automatically assign label for HTML tabular web data, depending on syntactical similarities between elements of the table. We categorize columns into three types: Disjoint Set Column (DSC), Repeated Prefix / Suffix Column (RPS) and Numeric Column (NUM). For labeling DSC column, our method rely on hits count from web search engine. Experimental results from large number of sites in different domains and subjective evaluation show that the proposed algorithm works fairly well. We hypothesize that our algorithm LadsComplete will do a good job for autonomous label assignment. We are NOT aware of any such prior work that address to connect two orthogonal research viz. wrapper generation and label extraction for value added services such as online comparison shopping.