1 Introduction

Econometric linguistics explores the relationship between word types and word examples, and evaluates vocabulary richness based on word types and word examples, but cannot analyze the temporal evolution of vocabulary. The vocabulary growth model fits the relationship between word types and word examples growth through a function, describing the trend of changes in word types and word examples ratio. The mainstream models include Guiraud, Heaps, and Hubert models, which are suitable for analyzing lexical changes in corpus with temporal order. Macao, as a hub of Chinese and Western cultures, has implemented the "Three Literatures and Four Languages" policy, but there is relatively little research on the evolution of vocabulary. The media is the main channel for information dissemination, and news media objectively reflects social phenomena and development trends. This article constructs a news corpus for the Macau Times and uses a vocabulary growth model to analyze the changes in Macau news vocabulary, verifying the effectiveness of the model.

2 Vocabulary Growth and Macau Vocabulary Research

Explored the application of word types and examples in evaluating language proficiency, determining author identity, and pointed out its limitations in diachronic analysis. This article introduces a vocabulary growth model that predicts the growth relationship between word types and examples through mathematical functions, effectively reflecting vocabulary richness. It also provides examples to illustrate its value in analyzing the evolution of vocabulary in speeches and government work reports. At the same time, an overview of the current status of research on Chinese vocabulary in Macau was provided, pointing out that as a cultural hub of East and West, there have been studies on the hidden language of the gambling industry, local characteristic words, and the formation of language life in Macau. However, there is a lack of corpus based analysis of the relationship between the growth of vocabulary over time and social hotspots. Studying the evolution of vocabulary in Macau can help validate the applicability of the vocabulary growth model in the field of news, fill the gap in research on the evolution of Macau vocabulary, and provide a more in-depth analysis of language development dynamics.

3. Constructing the Corpus of Macau Times

This chapter introduces the process of constructing the corpus of the Macau Times. Firstly, the comprehensive newspaper "Macau Times" was chosen as the research object, as it covers a wide range of fields such as law, sports, people's livelihood, and international hotspots, making it suitable for vocabulary analysis. Secondly, use the pkuseg word segmentation toolkit developed by Peking University to perform word segmentation on traditional Chinese corpus to ensure accuracy. After segmentation, remove words containing Arabic numerals and punctuation marks, and consider polysyllabic structures as words. Finally, the statistical information of the Macao Times corpus is presented, including the number of articles, examples of words, types of words, and number of words from 2011 to 2021.

4. Vocabulary growth of Macau Times

This chapter explores three models of vocabulary growth in The Macau Times: Guiraud, Heaps, and Hubert. The Guiraud model is based on the ratio of word types to the number of word examples, the Heaps model is an exponential prediction model, and the Hubert model distinguishes between general and specialized vocabulary. By using the Scipy4 framework and nonlinear least squares optimization, the Guiraud model parameters r=98.028625, the Heaps model parameters a=2.785034, C=0.727058, and the Hubert model scale parameter p=0.316723 were obtained. The model fitting results show that the Heaps model has the best prediction performance and is closest to the actual observed values, while the Guiraud model has the worst performance. Therefore, the subsequent analysis will use the results of the Heaps model.

Analysis of Vocabulary in Chinese Media in Macau

The vocabulary analysis of the Macau Times from 2011 to 2020 shows that the number of newly added words fluctuates between 13000 and 72000, influenced by changes in news reporting themes. In 2011, the predicted values of the Heaps model were slightly lower than the actual observed values, reflecting the emergence of social hot topics such as "nuclear power plant", "Zhuhai North Station", and "Osama bin Laden". In 2012, the predicted value was higher than the observed value, with a total of 16398 new words added, reflecting society's attention to people's livelihoods, historical and cultural protection, space events, and financial innovation. In 2013, the predicted values were higher than the observed values, with 15140 new words added, related to the rapid development of technology and the Legislative Council elections. In 2014, the observed values were lower than the predicted values, and 13505 new words were added, focusing on livelihood issues and patriotic related vocabulary. In 2015, the predicted value was greater than the observed value, with 17878 new words added, reflecting the decline in gambling industry revenue and the effectiveness of economic diversification. In 2016, the predicted values were lower than the observed values, with 65524 new words added, mainly in the fields of sports and military. In 2017, the predicted value was lower than the observed value, and the number of new words added was influenced by the change of the Macau Times from a weekly to a daily publication, as well as the implementation of smart city construction policies. In 2018, the predicted values were lower than the observed values, with 38690 new words added, related to the promotion of food culture, emergency mechanisms, and the development of artificial intelligence. In 2019, the predicted values were lower than the observed values, resulting in 28273 new words closely related to technological development. In 2020, the predicted value was greater than the observed value, and 28071 new words were added, mainly in the fields related to the COVID-19 epidemic, with a slight decline in vocabulary richness. These data reflect the development and changes of Macau society in areas such as economy, technology, culture, and public health.

6 Verification Procedure for Vocabulary Growth in Macau

This chapter verifies the difference between the predicted and observed values of vocabulary growth in Macau through randomization methods. Generate new text by rearranging the order of all words in the corpus, while keeping the total number of word examples and types unchanged. By calculating the Z-Score of the difference between predicted and observed values, determine the impact of changes in news content on the model's prediction performance. The experimental results indicate that the prediction error of random text is much smaller than that of real text, and supporting changes in news content is the main reason for the error. The Heaps model has good reliability in processing different corpora.

7 Conclusion

Based on more than ten years of corpus from the Macau Times, Guiraud, Heaps, and Hubert models were used to analyze the temporal evolution of Macau vocabulary, and it was found that the Heaps model had the best fitting effect. When policies are stable and world hot news is stable, the predicted value of the Heaps model is higher than the observed value, reflecting a high rate of word repetition and few new words added; When major hot events or new policies are introduced, if the observed values are higher than the predicted values, it indicates that there are many new words added. By randomly shuffling the text order of the corpus, the effectiveness of the Heaps model in the analysis of news media corpus over time was further validated. This study reveals that the language life in Macau focuses on local livelihoods, national affairs, and international current events, showcasing the social characteristics of the intersection of Chinese and Western cultures in Macau.