Knowledge Engineering Assignment 2: Introduction to Knowledge Engineering Related Fields

Natural Language Processing

Knowledge engineering is a research field formed from the construction of expert systems, and has now become an interdisciplinary comprehensive discipline.His main research fields include soft computing, natural language processing, logic and reasoning, etc. [1].This article mainly focuses on the field of natural language processing for relevant introductions.
I. Overview of Natural Language Processing
NLP for short, it is a science that integrates linguistics, computer science, and art.Therefore, it is very closely related to the study of linguistics, but has important differences.The research object of natural language processing is not the natural language in daily life, but the development of computer systems, especially the software systems, which can effectively realize natural language communication.Hence it is part of computer science [2].
At present, there are several classic tasks in the field of natural language processing: stemming, lemmatization, word vectorization, part-of-speech tagging, named entity disambiguation, named subject recognition, sentiment analysis, text semantic similarity analysis andText summary, etc.
Second, stem extraction
This paper mainly selects stem extraction for a detailed introduction.Stemming is the process of removing the inflection or derivative form of words and converting the original words into stems [3].(For example, in English: the stems of "beautiful" and "beautifully" are both "beauti"; "stemmer", "stemming" and "stemmed" are based on the stem "stem".)
Three.Porter Stemming
The classic approach to this task is Martin Porter's Porter stemming algorithm.Dr. Porter also received the 2000 Tony Kents Award for his work in stemming and information retrieval.
This method first requires us to define a file containing a to store stemThe class to extract the array of words before the formal processing of the algorithm can begin.The first step of the algorithm is to deal with plurals, and words ending in "ed" and "ing".The second step is to find out if there is a word that contains a vowel and ends in "y"; after finding it, change "y" to "i".The third step is to map words with double suffixes, such as preferred; to single suffixes.The fourth step is to deal with suffixes such as "-ic", "-full", and "-ness".Then remove "-ant", "-ence", "-e", etc. at the end of the word where appropriate.Finally, a stem() method is used to obtain the stem of the word transformation.

Fourth, the latest method
Since the advent of the Porter stemming method, new stemming methods have been emerging, and there are improved methods based on the Porter stemming method [4]; otherThere are some new and smarter methods, such as the method of n-gram parsing.The method exploits the context of a word to extract the correct stem, which undoubtedly greatly improves the practicality.

References
[1] Huang Ronghuai, Li Maoguo, Sha Jingrong, Knowledge Engineering: A New Important Research Field [E], Electronic Education Research, 2004, (10): 1-7
[2] Li Changyun, Wang Zhibing, Intelligent perception technology and its application in electrical engineering, University of Electronic Science and Technology of China Press, 2017.05, p. 163)
[3] Common 10 Natural Language Processing Technologies, September 2, 2021,https://blog.csdn.net/Harrytsz/article/details/120053267
[4] Widjaja M, Seng H. Implementation of Modified Porter Stemming Algorithm to Indonesian Word Error Detection Plugin Application[J]. Int J Hum CultStud, 2015, 6(2):139.