An Approach for Content Retrieval from Web Pages Using Clustering Techniques

Abstract

Mining the content from an information database provides challenging solutions to the industry experts and researchers, due to the overcrowded information in huge data. In web searching, the information retrieved is not an appropriate, because it gives ambiguous information for the user query, and the user cannot get relevant information within the stipulated time. To overcome these issues, we propose a new methodology for information retrieval EPCRR by providing the top most exact information to the user, by using the collaborative clustered automated filter which makes use of the collaborative data set and filter works on the prediction by providing the highest ranking for the exact data retrieved. The retrieval works on the basis of recommendation of data which consists of relevant data set with highest priority from the cluster of data which is on high usage. In this work, we make use of the automated wrapper which works similar to the meta crawler functionality and it obtains the content in the semantic usage data format. Obtained information from the user to the agent will be ranked based on the Enabled Pile clustered data with respect to the metadata information from the agent and end-user. The information is given to the end-user with the top most ranking data within the stipulated time and the remaining top information will be moved to the data repository for future use. The data collected will remain stable based on the user preference and works on the intelligence system approach in which the user can choose any information under any instances and can be provided with suitable high range of exact content. In this approach, we find that the proposed algorithm has produced better results than existing work and it costs less online computation time.

Exploring the Landscape of Hybrid Recommendation Systems in E-Commerce: A Systematic Literature Review

“…The synergy between these approaches creates a stable foundation, forming the cornerstone for subsequent enhancements [64]. The consistent reliance on collaborative filtering and content-based filtering reflects a strategic choice made by researchers to build upon a strong foundation while incorporating additional algorithms to fine-tune and optimize system performance [64], [65], [66], [67]. This strategic combination, as observed in the majority of instances, reflects an efficient approach to building recommendation systems that have successfully navigated a spectrum of challenges [66].…”

Design and development of text extraction and retrieval using style of documents in web searching

“…The design for identifying the web text is analyzing the characteristic of text, single pass, and web article text extraction [12]. Automatic annotation of query result from web database is classified into automatic annotation, decorative tag detector, cluster based shifting and Simple probabilistic , Tag Path(TP), Combining tag and vale Similarity (CTV'S), Local interface Schema (LIS), New Combining tag and vale Similarity (NCTV'S), Integrated Interface Schema(IIS) [8], [9] and [12]. Automatic annotation and wrapper generation is based on six annotators namely schema value, table, query based, frequency based, in-text prefix/suffix, common knowledge annotator.…”

“…Support Vector Machine (SVM) is used to identify the classifying of non-content blocks. Hierarchical clustering is used to find the similarity measure value of DOM tree [9]. There are two kinds of automatic page annotation search results namely tabling annotator and query based annotator [13].…”