Small Data Analysis for Bigger Data Analysis

Abstract

The terms “data science” and “big data” become very popular these days. Importance of these concepts are popularly recognized mainly due to the success of AI technologies. Especially, machine learning (ML) technologies such as deep learning have been applied practically in these years and equipment using these ICT technologies becomes very sophisticated. Thus, our life becomes more convenient. Huge amount of data is required in order to apply ML technologies into practical use. As a result, “big data” and “big data analysis” are recognized quite important. Even with such an environment, “small data (or non-big data)” and “small data analysis” remain important. Small data and small data analysis have advantages such as ease of data collection, ease of data analysis/mining, and appropriateness for experimental analysis in the style of trial and error, especially for domain-specific exploratory analysis. In this paper, we discuss advantages of small data analysis in comparison with big data analysis based on our experience of analysis mainly of those data obtained in our educational practices. We conclude that it is an efficient and effective method for developing data analysis methods to start from small data and expanding them in their size and variety.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 6 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.