The Application Of Natural Language Processing In Mining The Characteristics Of Concept Convey

Natural Language Processing(NLP)is widely used in different research fields,including text segmentation,speech recognition,text information processing,artificial intelligence,etc.The processing of the nouns or phrases in text is an important part of NLP research.One of the main goals of noun or phrase research is to explore the possibility of combining traditional keywords and syntactic methods with semantic methods to improve the quality of information processing and analysis.Massive open online courses(MOOCs)provides a wealth of learning resources for the public,and MOOC discussion forum becomes the main communication channel for deep interaction between students and teachers.Therefore,the appropriate length of student-assistant discussion in the discussion area is especially valuable to the questioners and the observers of the MOOC community.In this paper,NLP technology is used to mine the information of student-assistant question-and-answer(Q&A)transcripts(student's questions,corresponding to assistant's corresponding answers)in MOOC discussion area,and a natural language processing method for mining concept convey characteristics of Q&A transcripts is proposed.In this paper,Stanford Word Segmenter is used to segment Q&A transcripts,and Stanford POS Tagger is used to mark words in Q&A transcripts,extract the nouns(i.e.concepts)in Q&A transcripts.The concepts extracted from Q&A transcripts are obtained by using the language knowledge base HowNet/WordNet to get the corresponding upper concepts of each concept.Then,combining UCInet(a network analysis tool)and CRIE/Coh-Metrix(a text analysis tool),the concept convey process is considered as a directed graph model.The average path length,semantic depth,out/in degree,readability and LSASS1(overlapping degree of upper and lower sentences)are selected as five indicators to analyze the characteristics of concept convey,and then excavate the characteristics of concept convey in Q&A transcripts.Based on a natural language processing method for mining concept convey characteristics of Q&A transcripts,the student-assistant Q&A transcripts of MOOC/edX discussion area were collected to study the conceptual transfer characteristics of students-assistant long/short follow discussion in MOOC discussion area.In addition,Q&A transcripts between scientists-listeners(group of scientists)in different science lectures and spokespersons-reporters(group of spokespersons)in different press conferences were collected as two control groups.Three comparative experiments between scientists and spokespersons,student-assistant(long/short)and scientists and spokespersons,student-assistant(long)and student-assistant(short)were designed to study the relationship between student-assistant(long/short)posts and the concept convey characteristics in Q&A transcripts of scientists and spokespersons.It was found that the assistant teachers in MOOC discussion area should use sentences with low repetition of the above and the following sentences to promote the discussion in the discussion area,and keep the concepts as closely related as that of the political spokesperson in the press conference,and appropriately apply more abstract concepts that of the scientists in the science lecture.This paper presents a natural language processing method for mining the concept convey characteristics of Q&A transcripts,which can extract and analyze the characteristics of concept convey.The experimental results can provide practical guidance for MOOC discussion area assistants to promote discussion.