Information Retrieval from Job Posts based on K-means++ Clustering Algorithm
Keywords:
Information Retrieval, Machine Learning, K- means, K-means++, Elbow Method, Silhouette Analysis, Dis- counted Cumulative Gain Information Retrieval, Machine Learn- ing, K-means, K-means++, Elbow Method, Silhouette Analysis, Discounted Cumulative GainAbstract
The research paper deals with two main sections: firstly, the experiment comparison between k-means and k-means++ have been done using Elbow method and Silhouette method. Since, K-means++ is better than K-means, this research tries to justify that K-means++ has higher performance than K-means. Secondly, K-means++ has been used for Search and Information Retrieval system. Information Retrieval is an activity to obtain information system resources that are relevant to an information need from a collection of those resource. This research is useful to retrieve relevant documents that match a given query. When user add input such as industry type, job types, skills, and state, it will automatically calculate average and display the ranking. Subjective evaluation with DCG(Discounted Cumulative Gain) is done in order to measure ranking quality of information retrieval.
