Automatic Text Summarization on Social Media

A novel method of Chinese text content analysis and mining based on statistical models

“…Hao, Lili et al ( 2009) [13] used chi-square statistics to extract keywords for each month and combined with other rules to extract hot words for that month; they also used binomial test to test whether there was a significant difference between the probability of each word occurring in the first category and the second category where it appeared most frequently, and if so, the word was used as a keyword in the first category. Feng et al (2011) [14] and Feng et al ( 2012) [15] assumed that the documents came from a mixture of distributions in each category, and the generation model of documents under each category was a plain Bayesian model or a plain Bayesian model under tree constraints; each word might be a keyword with different probability of occurrence under each category or other words with the same probability of occurrence under each category. They introduced a potential indicator vector to indicate whether each word is a keyword or other words, and extracted keywords by Bayesian model selection.…”

“…Huang et al (1996) [13] proposed a processing segmentation standard for Chinese natural language that enables the realization of language adaptation, computational feasibility and data uniformity. Zhang et al (2020) [14] constructed an innovative text summarization model, which combines BERT, reinforcement learning, sequence-to-sequence and other technologies. Zhang et al (2019) [15] proposed a novel method ssp2vec to predict the contextual words based on the feature substrings of the target words for learning Chinese word embeddings.…”

Event Monitoring and Intelligence Gathering Using Twitter Based Real-Time Event Summarization and Pre-Trained Model Techniques

“…Ma [41] proposed a pre-trained model T-BERTSum for text summarization, which captures the key words of the topic information of social media, understands the meaning of the sentence and judges the topic of the message discussion, and then generates a high-quality section Summary. Kerui [42] uses BERT, Seq2seq and reinforcement learning to form a text summary model. Garg [43] uses T5, one of the most advanced pre-trained models, to perform a summary task on a data set with 80,000 news articles, and the results indicate that the summary generated by T5 has better quality than those generated by other models.…”

“…In summary, in our work, social media (such as Twitter), pre-trained models, and text summaries (including extractive and abstractive methods) are the three key elements of event summarization. From the above literature review, the text summaries of specific event tweets on social media have gradually become one of the most popular summarization research topics in recent years [1][2][3]41,42]. However, extractive summarization cannot obtain many key sentences in the tweet-form data, resulting in unsatisfactory quality of the generated summaries.…”

“…Furthermore, due to our task requires extensive domain knowledge beyond the meaning of individual words and sentences, in this work we limit our focus to abstractive summarization algorithms. On the other hand, in the recent references of text summarization, there are more and more work using pre-training language models [35][36][37][38][39][40][41][42][43][44]. Among them, BERT is the pre-trained model mainly used by most research teams.…”