1 Introduction

This article introduces the importance of constructing China's image, using the theory of extended meaning units and sentiment analysis techniques. By analyzing 500 index rows in the Coronavirus corpus, the semantic rhyme intensity is calculated, aiming to reveal the image of China's anti epidemic efforts in the eyes of foreign media and promote the enhancement of China's international influence.

2 Literature Review

This chapter explores the construction of China's image within the framework of the extended meaning unit theory. The extended semantic unit model proposed by Sinclair includes five components: nodal words, collocation, class connection, semantic tendency, and semantic rhyme. Among them, semantic rhyme is the key factor that expresses the speaker's attitude towards the pragmatic context. The frequency based polarity ratio method is a common method for calculating semantic rhyme strength, but it has limitations, such as not distinguishing the emotional strength of collocations in the same polarity category. Emotion analysis technology analyzes text emotions through an emotion lexicon, which contains a large number of emotion words and their corresponding values, and can be used for precise measurement of semantic rhyme intensity. At present, the application of sentiment analysis technology in semantic rhyme intensity calculation is relatively limited. This study applies sentiment analysis technology to the English corpus Coronavirus, using "China" as a node word, to calculate its semantic rhyme intensity in different extended meaning units, analyze the image of China's anti epidemic efforts in the eyes of foreign media, and provide inspiration for the external dissemination of China's image.

3 Research Design

Elaborated in detail on the design of Chinese image construction research based on semantic rhyme intensity algorithm. The research focuses on analyzing the semantic rhyme intensity of the word "China" in the corpus and its reflection of the image of China. The corpus is Coronavirus online corpus, which collects reports on the COVID-19 in 20 English speaking countries from 2020 to 2022. The researchers extracted the corpus from December 2022 to form a self built corpus with a total of 1543500 symbols. Corpus analysis includes importing R software, manually analyzing semantic tendencies, extracting class connections, recognizing emotional words, and calculating the overall and different category semantic rhyme strengths. The research aims to analyze the image of China in foreign media reports and understand China's international image in the context of the new era, especially in the fields of epidemic prevention and control, politics, military, diplomacy, and economy.

4 Research Results

The semantic tendency and emotional color of the word "China" in international news reporting were analyzed using the semantic rhyme intensity algorithm. Research has found that the term 'China' generally presents a weak positive semantic rhyme, but the emotional intensity of negative evaluations is generally higher than that of positive evaluations. Among different semantic orientations, the emotional response caused by "loosening policies" is the strongest, while the emotional intensity of "economic relations with other countries" is the weakest. The class connection form has a significant impact on the distribution of semantic rhyme, and different collocation forms exhibit different emotional strengths under the same semantic tendency. In the economic field, China has undergone a transition from a "dynamic zero tolerance" policy to a "loosening policy". Foreign media hold complex views on this, including optimistic expectations for economic recovery, as well as concerns about challenges and uncertainties. In terms of international relations, China has demonstrated its responsibility and commitment as a major power, while adhering to its own principles and bottom line. Foreign media have reported and evaluated this from different perspectives.

5 Conclusion

By using sentiment analysis techniques to optimize the measurement of semantic rhyme strength, the semantic rhyme strength values of Chinese node words were calculated, and their semantic rhyme features in different semantic orientations and class connections were analyzed. Research has found that in international news reports in December 2022, China overall presented a positive image, but negative evaluations were stronger than positive evaluations. The research provides inspiration for the dissemination of China's image to the outside world, and suggests restructuring the propaganda logic, innovating the discourse system, and telling the story of China well. At the same time, the limitations of the research were pointed out, such as the comprehensiveness of the sentiment lexicon and the rationality of sentiment tendency assignment. It was proposed that future research could apply sentiment analysis techniques to the calculation of semantic rhyme intensity for different text types, expand sentiment lexicon data, and improve the applicability of language sentiment analysis techniques.