Abstract
<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Background:</b> There is a growing need for Chinese state-owned enterprises (CSOEs) to utilize Twitter, as an effective communicative tool in the professional business context, to build a credible image to the global community. <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Literature review:</b> Little attention has been paid to measuring the discursive construction of message credibility through corporate Twitter. Therefore, based on the theoretical insights of message credibility from existing literature on communication and information science, our study has conceptually developed a broad framework to measure the message credibility of CSOEs’ Twitter discourse from two general aspects (content and form), four separate levels ({thematic}, {intrinsic}, {contextual}, and {representational}), and nine specific dimensions (<capability>, <morality>, <objectivity>, <authority>, <accuracy>, <informativeness>, <timeliness>, <consistency>, and <persuasiveness>). With the help of Natural Language Processing (NLP) and corpus tools (MAT, CLA, TAALES, GAMET, SÉANCE, and TAACO), the framework has been practically operationalized by a total of 62 discursive features, including 18 content-based themes (thematic features) and 44 form-based features. <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Research questions:</b> 1. What themes do CSOEs develop, and how do they express these themes to establish message credibility in their tweets? 2. Which dimensions of message credibility are significantly highlighted in CSOEs’ tweets? 3. Which enterprises establish the highest message credibility in their tweets? <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Methodology:</b> We collected tweets during the year 2020 from the official Twitter accounts of 15 CSOEs and applied our operationalized framework to conduct nine separate One-way ANOVAs, a principal component analysis (PCA), and a mean-value based descriptive statistics comparison, respectively. <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Results:</b> First, CSOEs developed themes including strength, power, cooperation, and legitimacy, among others, and used discursive features including nominalizations, mentions/@ , word length, time adverbials, hashtags/#, and semantic overlaps, among others when expressing these themes to establish message credibility. Second, CSOEs significantly highlighted the <capability>, <authority>, <informativeness>, and <consistency> dimensions of message credibility in their tweets. Last, China National Machinery Industry Co. (Sinomach), China Datang Co. (CDC), China Railway Engineering Co. (CREC), and China State Construction Engineering Co. (CSCEC) were found to have established the highest message credibility in their tweets. <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Discussion and conclusion:</b> Our study may be the first to generate an NLP-cum-corpus-operationalized framework to quantitatively measure the discursive realization of message credibility in the context of business communication on social media. It also provides some practical insights into how relevant business professions can utilize certain discursive resources to establish message credibility in the B2C communication on social media.