Distance-based algorithms are widely used for data classification problems. The k-nearest neighbour classification (k-NN) is one of the most popular distance-based algorithms. This classification is based on measuring the distances between the test sample and the training samples to determine the final classification output. The traditional k-NN classifier works naturally with numerical data. The main objective of this paper is to investigate the performance of k-NN on heterogeneous datasets, where data can be described as a mixture of numerical and categorical features. For the sake of simplicity, this work considers only one type of categorical data, which is binary data. In this paper, several similarity measures have been defined based on a combination between well-known distances for both numerical and binary data, and to investigate k-NN performances for classifying such heterogeneous data sets. The experiments used six heterogeneous datasets from different domains and two categories of measures. Experimental results showed that the proposed measures performed better for heterogeneous data than Euclidean distance, and that the challenges raised by the nature of heterogeneous data need personalised similarity measures adapted to the data characteristics.

No abstract

Highlights The historical development of the public COSMOS database is provided. COSMOS NG is a knowledge hub to share toxicity data and in silico tools. COSMOS NG has broad chemical coverage, with a focus on cosmetics. Chemical and toxicological data are quality assured through inclusion criteria. In silico TTC, profiling and read-across workflows are illustrated.

With the increased use of online social networking sites, data retrieval from social networking profiles is becoming a major tool for business. What makes social networking profile data different is its semi-structured format. The structure and the presentation of profile data change all the time. In social networking there is a lack of research into automated data retrieval from semi-structured web pages. Our approach is based on automated retrieval of the profile's attributes and list of top friends from MySpace by examining and extracting the relevant tokens in the parsed HTML code. The tokens were placed into a repository and Breadth First Search algorithm was used. The approach was implemented and tested with a profile which resulted in over 800 top friend profiles and attributes being extracted. This implementation process highlighted that MySpace profile structures vary depending on profile type and the way in which the user has customised the profile.

No abstract