Abstract
R has gained explicit text mining support with the tm package enabling statisticians to answer many interesting research questions via statistical analysis or modeling of (text) corpora. However, we typically face two challenges when analyzing large corpora: (1) the amount of data to be processed in a single machine is usually limited by the available main memory (i.e., RAM), and (2) the more data to be analyzed the higher the need for efficient procedures for calculating valuable results. Fortunately, adequate programming models like MapReduce facilitate parallelization of text mining tasks and allow for processing data sets beyond what would fit into memory by using a distributed file system possibly spanning over several machines, e.g., in a cluster of workstations. In this paper we present a plug-in package to tm called tm.plugin.dc implementing a distributed corpus class which can take advantage of the Hadoop MapReduce library for large scale text mining tasks. We show on the basis of an application in culturomics that we can efficiently handle data sets of significant size.
Highlights
In the information age statisticians are confronted with an ever increasing amount of data stored electronically (Gantz et al 2008)
This paper has introduced an approach applying the distributed programming paradigm MapReduce to advance feasibility and performance of suitable text mining tasks in R
We showed that distributed memory systems can be effectively employed within this model to preprocess large data sets by adding layers to existing text mining infrastructure packages
Summary
In the information age statisticians are confronted with an ever increasing amount of data stored electronically (Gantz et al 2008). In a recent publication in Science, Michel et al (2011) use 15% of the digitized Google books content (4% of all books ever printed) to study the diffusion of regular English verbs and to probe the impact of censorship on a person’s cultural influence over time. This led to the advent of a new research field called Culturomics, the application of high-throughput data collection and analysis to the study of human culture
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have