A review of documented findings on Text Summarization for Children news rendering

News rendering involves the representation of News in a suitable and usable manner essentially for better understanding. Over the years there has been severe interest in the ways by which this representation is being formulated. These concerns are majorly issues around the target audience especially children. Many approaches to re-render news objects to children have failed as they alter the semantics of the body of the contents. The difficulty has been a gap that necessitated the study. We therefore present a basic parsing technique and a Text Summarization system to meet with the identified difficulties. The basic parsing model can summarize texts based on age-graded variation (Lower School level and the Middle school level). The system can summarize texts of about 20,000 characters in length for children that belong to lower school level and 500,000 characters in length for children that belong to middle school level.

After testing some of the collected datasets, the study reveals that the average processing time for summarizing input texts on the platform for children in lower school level is about 0.2036ms while on the basic level is about 1264ms. The result indicates that the system adheres to the proposed approach, it summarizes inputs from different sources and processes request very fast. After comparison with some of the existing summarizers based on level of grammar correctness, parameter information content, length of summary that can be processed and structured of content that is generated, deductions shows that the proposed approach is a tool to reckon with for children news rendering activities.

Keywords: Text Summarization, News rendering, Machine learning, Natural Language Processing.