engleski

Design of a workbench and guidelines to improve the efficacy of advertising messages

Purpose – This study is aimed at proposing the design of an efficient workbench and guidelines to help content providers in tourism to address proper messages both in managing ordinary activities and in facing unpredictable events and external shocks. Recent Covid-19 related events have particularly stressed the importance of tools supporting, guiding and directing message providers in designing advertising campaigns to attract visitors providing proper information and contrasting the possible negative influence of misinformation. Social media and the Internet can give a wide picture of current trends. Nevertheless, without appropriate tools suitable to analyse a huge quantity of data, the research of information on trends and mood by advertising providers constitutes a challenge difficult to face unless limited to the analysis of selected websites with consequent possible loss of details. Such a choice, based on humans’ subjective selection of sources, could introduce bias compromising the efficiency of the message. Methodology – Several studies state that the composition of an advertising campaign should be a conjoint activity including both creative and scientific work, involving also methodologies and tools helping messages provider to improve the efficacy and the impact of advertising campaigns. The present study aims to introduce an innovative methodology based on combining ideas from Information Technology and Econometrics. In particular Artificial Intelligence methods, such as Machine Learning and sentiment analysis, through the Artificial Neural Networks (ANN) and Vector Autoregressive models (VAR). Findings – The language used and emotions elicited in tourism advertising messages have a strong impact on attracting visitors and tourists, particularly when international events create concerns in the population. Indeed, a multimodal approach, supported by tools from Econometrics, Machine Learning and Sentiment Analysis showed the efficacy in capturing the mood and the trend in prospect visitors to allow to redistribute tourist flows towards different destinations and within wider periods of the year. Contribution – The work gives a theoretical and practical contribution. Indeed, the combination of Econometrics, Artificial Neural Networks and Sentiment Analysis tools can constitute a scientific base to analyse trends and to compose advertising texts, websites and campaigns taking into account the impact of unexpected events and maximising Linguistics suggestions to improve the efficiency of messages. Besides, the tools also allow suggesting guidelines of the features and the structure that should be always considered. Last but not least, these tools could be used to select the main topics arising on the web and social networks around current issues perceived by visitors, allowing the selection of a list of Q&A practical indications which can be translated into effective virtual assistants particularly useful when the personnel is reduced due to unexpected events and constraints, as during Covid-19 periods.

Tourism ; Econometrics ; Machine Learning ; Sentiment Analysis ; Workbench ; Virtual Assistant

nije evidentirano