The positive energy of netizens: development and application of fine-grained sentiment lexicon and emotional intensity model

Abstract

The outbreak of COVID-19 has led to a global health crisis and caused huge emotional swings. However, the positive emotional expressions, like self-confidence, optimism, and praise, that appear in Chinese social networks are rarely explored by researchers. This study aims to analyze the characteristics of netizens' positive energy expressions and the impact of node events on public emotional expression during the COVID-19 pandemic. First, a total of 6,525,249 Chinese texts posted by Sina Weibo users were randomly selected through textual data cleaning and word segmentation for corpus construction. A fine-grained sentiment lexicon that contained POSITIVE ENERGY was built using Word2Vec technology; this lexicon was later used to conduct sentiment category analysis on original posts. Next, through manual labeling and multi-classification machine learning model construction, four mainstream machine learning algorithms were selected to train the emotional intensity model. Finally, the lexicon and optimized emotional intensity model were used to analyze the emotional expressions of Chinese netizens. The results show that POSITIVE ENERGY expression accounted for 40.97% during the COVID-19 pandemic. Over the course of time, POSITIVE ENERGY emotions were displayed at the highest levels and SURPRISES the lowest. The analysis results of the node events showed after the outbreak was confirmed officially, the expressions of POSITIVE ENERGY and FEAR increased simultaneously. After the initial victory in pandemic prevention and control, the expression of POSITIVE ENERGY and SAD reached a peak, while the increase of SAD was the most prominent. The fine-grained sentiment lexicon, which includes a POSITIVE ENERGY category, demonstrated reliable algorithm performance and can be used for sentiment classification of Chinese Internet context. We also found many POSITIVE ENERGY expressions in Chinese online social platforms which are proven to be significantly affected by nod events of different nature.

Construction of an Emotional Lexicon of Patients With Breast Cancer: Development and Sentiment Analysis

“…We randomly selected the annotation results of 327/4079 (8%) words from the corpora of patients with BC for the annotator consistency test [ 80 ], and the Fleiss κ value was 0.491 (95% CI 0.482-0.501), with moderate strength of agreement, which showed that the annotation results were consistent and the consistency was acceptable.…”

“…See Multimedia Appendix 3 for the number and examples of emotional words based on 8 emotional categories in the emotional lexicon of patients with BC. We randomly selected the annotation results of 1471 of 14,709 sentiment words at a rate of 8% for the annotator consistency test [ 80 ], and the Fleiss κ value was 0.439 (95% CI 0.437-0.441), with moderate strength of agreement, which showed that the annotators’ annotation results were consistent and the consistency was acceptable.…”

“…Drawing lessons from the emotional classification standard of emotional words in the C-LIWC and the polysemy of Chinese words (ie, a word may belong to different emotional categories) [ 44 , 48 ], the following provisions were made referring to previous research on manual lexicon construction [ 44 ]: (1) the words that could not be classified would be marked as “none”; (2) if the same word was marked by more than 5 annotators in 2 or more categories, this word would be marked as belonging to 2 or more categories based on the research of Wu et al [ 44 ] and Zhou and Yang [ 79 ]; and (3) if more than 5 annotators marked a word as “none,” and the number of annotators who marked this word in other categories was less than 5, the word would be excluded from the emotional lexicon. Next, we used a random number table to randomly select the annotation results at a rate of 8% for the annotator consistency test [ 80 ].…”

A Rumor Propagation Model Considering Media Effect and Suspicion Mechanism under Public Emergencies

“…By collecting the basic public opinion discussion data of the 'Taocheng Hengshui Middle School Incident' for 15 consecutive days, taking the Weibo Topic readership number, the number of discussions and the number of original people as the life cycle division which is shown in Figure 2, the emotional words were extracted through the emotion dictionary [46], and some of the emotion words, under the artificial annotation of specific comments, were taken as the emotion base words [47]. Then, calculate the semantic similarity [48].…”