The objective is a neural-based feature selection in intelligent recommender systems. In particular, a hybrid neural genetic architecture is modeled based on human nature, interactions, and behaviour. The main contribution of this chapter is the development of a novel genetic algorithm based on human nature, interactions, and behaviour. The novel genetic algorithm termed “Buabin Algorithm” is fully integrated with a hybrid neural classifier to form a Hybrid Neural Genetic Architecture. The research presents GA in a more attractive manner and opens up the various departments of a GA for active research. Although no scientific experiment is conducted to compare network performance with standard approaches, engaged techniques reveal drastic reductions in genetic operator operations. For illustration purposes, the UCI Molecular Biology (Splice Junction) dataset is used. Overall, “Buabin Algorithm” seeks to integrate human related interactions into genetic algorithms as imitate human genetics in recommender systems design and understand underlying datasets explicitly.
TopIntroduction
Data explosion across business platforms has necessitated the call for efficient storage mechanisms. Research scientists estimate that close to 80% of a business’s data lies in amorphous data format. This means, management decisions are likely to be optimized should unstructured data formats be considered in decision-making processes. Since its inception, the Internet has grown to become one of the largest data repositories in the world. With every computer literate as potential author on the Web, new domains have been registered and corresponding Website information added or updated to increase readership/viewership. News producing giants such as CNN, BBC etc have taken great advantage of the Internet to disseminate news to a wider reading/viewing public. By transmitting news stories on the Internet, “breaking news” for example, can be accessed across the globe by the within a matter of minutes. Governments, institutions (public or private) and individuals, have also joined in the production of unstructured data by publishing pages about business, jobs, goods, and services. Text is easily to deployed, have low runtime impact on servers, and load faster. They use less hardware resources (e.g. server disk space) and much more information likely to be published—as compared to video, images, etc. The growing nature of the Internet has prompted researchers to mine the data therein to ascertain hidden trends and heuristics.
Classification based learning agents have been built to extract, mine and perform exploratory tasks on large text data in non-stationary platforms. Among approaches used, neural based methods have proved to be better candidates than their counterparts. Buabin (2011a, 2012) are typical examples of such systems. Buabin (2012) argues that, classification and recommender systems have major commonalities across their design. They both
- 1.
Operate in search related domains,
- 2.
Use search related mechanisms for finding optimal solutions to problems,
- 3.
Use learning algorithms to extract knowledge from the training set into their structures and most importantly train on the exemplars.
With the aim of improving classifier performance, researchers have extracted features prior to performing classification tasks. Although implemented in many data mining systems, feature selection has taken many forms, some of which are statistical approaches (Perkins, et al., 2003; Hermes & Buhmann, 2000; Dash & Liu, 1997; Forman, 2003; Krishnapuram, et al., 2004; Piramuthu, 2004; Yang, 1999; Liu & Yu, 2005; Song, et al., 2007; Weston, et al., 2003) and neural based approaches (Gabrilovich & Markovitch, 2004; Rakotomamonjy, 2003; Froehlich, 2002; Liu & Zheng, 2005; Taira & Haruno, 1999; Hardin, et al., 2004; Zhang, et. al., 2006; Huang & Wang, 2006; Chen, 2007; Warmuth, et al., 2003; Suna, 2004).