000			02648nam a22002417a 4500
003			ft8899
005			20251218093303.0
008			251218b ||||| |||| 00| 0 eng d
041			_aengtag
050			_aQA76.6 B35 2025
082			_a.
100	1		_aBaje, Ronan C.; Caladio, Jerome Z.; Tenio, Jonald R.
245			_a Designing an adaptive dynamic approach applied in text summarization
264		1	_a. _b. _cc2025
300			_bUndergraduate Thesis: (Bachelor of Science in Computer Science) - Pamantasan ng Lungsod ng Maynila, 2025
336			_2text _atext _btext
337			_2unmediated _aunmediated _bunmediated
338			_2volume _avolume _bvolume
505			_aABSTRACT: In line with the trends in machine learning, training of models, and artificial intelligence, the concept the researchers marked is text summarization. This condenses texts into concise summaries given the important aspects with the pool of texts as keywords by using two approaches of dynamic programming: (1) memorization and (2) tabulation. The focus in using dynamic programming keeps the application as lightweight in memory space as possible and is the reason for its use to enhance the existing algorithms into a new design. Dynamic programming for this research is a well-refactored series of codes in combination of memorization and tabulation techniques combined into an adaptive tabulation system updating the hard-coding ways of existing algorithms in initializing base cases for summarizing texts. This also solves the limitation in regular expressions on inconsistent summaries and saving redundant words of different capitalizations. This method modernizes programming of text summarizers using three principles : (1) normalization, (2) shallow parsing, and (3) segmentation. This research demonstrated the process of summarizing document texts input in giving the essence in its chunks of text. This proves useful and beneficial for future researchers whether they train models or use machine learning using some metrics: (1) ROUGE-L., (2) BERT Score, and (3) Reference-free Score. The results show consistent summarization of texts among a variety of research, articles and storytelling ranging 80-90% in precision and accuracy. The streamlines the existing algorithms from hard-coding systems that can only provide precise and accurate results of 99.99%, for texts that are specified on certain types of messages. For the next step, the application can be managed to operate in different file formats and modify it into reading other types of literature and human writing.
526			_aF
655			_aacademic writing
942			_2lcc _cMS
999			_c37372 _d37372