Type I and Type Ii Errors and Land Law

Only available on StudyMode
  • Download(s) : 302
  • Published : February 14, 2013
Open Document
Text Preview
AGUILAR, Janica Mara Y.
BERBOSO, Kevin Leo C.
CAYUBE, Katherine Kate Z.

Prof. Ria Sagum
Advisor

PHILEX: Philippine Land Law Expert Chatbot

Abstract

Keywords:Chatbot, Natural Language Processing (NLP), First-Order Predicate Logic, Natural Language Generation (NLG), Precision and Recall

The researchers tend to work on an expert system with integration of Philippine land laws. They chose to focus on land laws since one of the major problems of the Filipinos who seek law experts’ advice are due to land. The researchers deliberates the possible ways on how can they share knowledge and guidance that works like a real lawyer advising a client but have not taking an account to replace to seek a real land law experts guidance. Also, putting an expertise into a machine happens in many ways. So, the researchers find a way on constructing an expert system on Philippine land law using different techniques. In giving solution to the stated problem of the researchers, this study implements techniques such as Natural Language Processing (NLP) with sub-processes of Tokenizing, Part-of-Speech Tagging, and Parsing for input abstraction and validation, as well as Natural Language Generation (NLG) for the construction of answer/s. The developed system entitled, “PHILEX: Philipine Land Law Expert Chatbot” focused on providing answers to the user’s questions. Such users who are in need of information, assistance and knowledge and those who seek counselling in regards to any Philippine land law or land rights can benefit from the said expert system but must always know that the researchers do not intend to substitute the land law professionals. Upon various tests that the researchers conducted aiming to know the accuracy, specificity and sensitivity of the system, researchers came up with such conclusions. A technique, Precision and Recall, was used to test the system’s capability of answering the user’s questions relying from correct result...
tracking img