From one star to three stars : Upgrading legacy open data using crowdsourcing

摘要

Despite recent open data initiatives in many coun-tries, a significant percentage of the data provided is in non-machine-readable formats like image format rather than ina machine-readable electronic format, thereby restricting theirusability. This paper describes the first unified framework forconverting legacy open data in image format into a machine-readable and reusable format by using crowdsourcing. Crowdworkers are asked not only to extract data from an image of achart but also to reproduce the chart objects in spreadsheets.The properties of the reconstructed chart objects give theirdata structures including series names and values, which areuseful for automatic processing of data by computer. Sinceresults produced by crowdsourcing inherently contain errors,a quality control mechanism was developed that improves theaccuracy of extracted tables by aggregating tables created bydifferent workers for the same chart image and by utilizingthe data structures obtained from the reproduced chart objects.Experimental results demonstrated that the proposed frameworkand mechanism are effective.