APLIKASI PENDETEKSI POLA KALIMAT HEADLINE COPYWRITING DENGAN METODE SHIFT REDUCE PARSING

Abstract

This research aims to produce an application that assesses copywriting headline sentences with correct grammar and word mismatches. Copywriting is an activity of a work through writing to made the readers get response that the writers want to convey and in it there is one of the constituent elements, namely the headline which is a sentence that will be seen first before reading the main content of the work and in the headline there is an RTO formula for making it.

In this research, the problem raised was implementing the shift reduce parsing method as a detection of sentence patterns and grammar in the headline copywriting from the preprocessing stage which provides part-of-speech (POS) labels using the HMM model with an accuracy value of 94.69% to the steps parsing and grading of sentences. In making this application, the SDLC waterfall model is carried out in a sequence of several processes in stages in designing and developing a system.

The result of this research is an application with the form of a web framework using ASP.net as a web interface. After the application has been built, it will be tested using the blackbox test which results in 98% successful parsing that goes according to design.