Hellenic Open University Conferences, International Conference on Business & Economics of the Hellenic Open University 2015

Font Size: 
A statistical analysis of big web market data structure using a big real dataset of wines and the direct e-commerce social phenomenon of producer to consumer transaction
Ioannis Athanasiadis, Dimitris Ioannidis

Building: Titania
Room: Solon
Date: 2015-02-06 05:00 PM – 07:00 PM
Last modified: 2015-01-27


The web market structure nowadays it's facing the issue of “big-data”. The huge increasing evolution of transactions is creating the corresponding increasing of e-commerce data through e-shop and client. This paper provides research from a statistical view in the above new transaction methods of e-buying. Finding empirical results from some on-line big market sites like Amazon and GetGlue this paper extracts interesting statistical results using functional regression with the price functions as response variables introducing the “ranking” and the “sampling” of a big dataset.

Having a big dataset of real parameters of wines of a known Greek company we investigate and research some statistical results trying to predict the quality score depending on these parameters.

Furthermore it extends these results to other type of market as in electronic or food products. Especially in this area of food products, using a complete e-shop site created by the authors, we study and support the dynamic new emerging field of the producer-to-consumer direct e-buying transaction which eliminates all the intermediate involved such the middleman or wholesale merchants.

An appropriate questionnaire is answered as an in-depth interview by many producers of this market area giving interesting directions of the research.


e-commerce; big data; R language; statistical models; producer-to-consumer market; regression; wine quality