Web Page Classification on News Feeds Using Hybrid Technique for Extraction

Abstract

At the initial stage of web world, the number of websites hosted was handful, and from the users end, it was easy to maintain log file which consists of information like web pages URL or domain name, but as the number of web hosting increased gradually, it was found hard for the users to maintain such log files. Thus, the requirement exists which helped the users to search the information from the website easily which is now renowned as “Search Engines”; the only limitation found is that the users must be sound enough to give searching keywords in order to search relevant information, but in many cases, users have obtained irrelevant information from the web, and hence, looking into the current scenario of the Internet world, the number of websites has grown drastically holding various web pages within it. These web pages are observed to be published in structured or semi-structured manner which comprises of various multimedia contents, [(Nethra et al. in J Soft Comput 4:692–696, 2014) 1; (Kardan et al. in A novel approach for Keyword extraction in learning objects using text mining & WordNet. pp. 788–792, 2011) 2; (Menaka and Radha in Int J Adv Res Comput Sci Softw Eng 352:24–28, 2013) 3] so the chance to fetch wrong information also increased, and hence, there is a need to auto-categorize the web pages into some predetermined sections. The key point in this research is to recognize and allocate the news feeds into fixed sections of news like business, sports which enhance the reader’s accessibility towards relevant news by traversing appropriate category as per his/her choice, and this is done by adopting hybrid technique of URL analysis and content context analysis. The paper emphasis on a proposed model to perform classification on news feed related to various fields such as sports, health which starts with web crawling of URLs, scraping of news contents followed by the analysis carried out on account of generating keywords, weight calculation, and then at last identify the relevant category on the basis of contents fetched among various Indian news web portal like “The Times of India”, “Hindustan Times”, “The Guardian” [(Nethra et al. in J Soft Comput 4:692–696, 2014) 1; (Jena and Kamila in Int J Appl Innov Eng Manage 2, 2013) 4; (Ercan and Cicekli in Int J Inf Process Manage 43:1705–1714, 2007) 5].