Fodé Camara

A Novel Term Weighting Scheme Model

The use of textual data has increased exponentially in recent years due to the networking infrastructure such as Facebook, Twitter, Wikipedia, Blogs, and so one. Analysis of this massive textual data can help to automatically categorize and label new content. Before classification process, term weighting scheme is the crucial step for representing the documents in a way suitable for classification algorithms. In this paper, we are conducting a survey on the term weighting schemes and we propose an efficient term weighting scheme that provide a better classification accuracy than those obtained with the famous TF-IDF, the recent IF-IGM and the others term weighting schemes in the literature.

A Parallelized Spark Based Version of mRMR

Nowadays, we are surrounded by enormous large-scale high dimensional data called big data and it is crucial to reduce the dimensionality of data for machine learning problems. That’s why feature selection plays a vital role in the process of machine learning because it aims to reduce high-dimensionality by removing irrelevant and redundant features from original data. However some characteristics of big data like data velocity, volume and data variety have brought new challenges in the field of feature selection. In fact, most of existing feature selection algorithms were designed for running on a single machine (centralized computing architecture) and do not scale well when dealing with big data. Their efficiency may significantly deteriorate to the point of becoming inapplicable. For this reason, there is an increasing need for scalable yet efficient feature selection methods. That’s why we present here a distributed and effective version of the mRMR (Max-Relevance and Min-Redundancy) algorithm to face real-world problems of data mining and evaluate the empirical performance of the proposed algorithms in selecting features in several public datasets. When we compared the efficiency and the scalability of our parallelized method in comparison with the centralized one we have found out that our parallelized method have given better results.

S-FPG: A parallel version of FP-Growth algorithm under Apache Spark™

Frequent Itemsets Mining (FIM) is an essential data mining task, with many real world applications such as market basket analysis, outlier detection, and so one. Many efficient single-node FIM algorithms such as the well-known FP-Growth algorithm have been proposed in the last two decades. However, as large-scale datasets are usually adopted nowadays, these algorithms become inefficient to mine frequent itemsets over big data. Scalable parallel algorithms hold the key to solving the problem in this context. However, the existing parallel versions of FP-Growth algorithm implemented with the disk-based MapReduce model are not efficient enough for iterative computation. In this paper, we propose an implementation of scalable parallel FP-Growth using the inmemory parallel computing framework Apache Spark™. Our experimental results demonstrated that the proposed algorithm can scale well and efficiently process large datasets.

A Secure Protocol to Maintain Data Privacy in Data Mining

Recently, privacy issues have becomes important in data mining, especially when data is horizontally or vertically partitioned. For the vertically partitioned case, many data mining problems can be reduced to securely computing the scalar product. Among these problems, we can mention association rule mining over vertically partitioned data. Efficiency of a secure scalar product can be measured by the overhead of communication needed to ensure this security. Several solutions have been proposed for privacy preserving association rule mining in vertically partitioned data. But the main drawback of these solutions is the excessive overhead communication needed for ensuring data privacy. In this paper we propose a new secure scalar product with the aim to reduce the overhead communication.