رها بصرایی

Consumers rely on social media opinions to make product choices and purchases. With the popularity of web-based platforms like Tripadvisor, consumers express their opinions and feelings about food quality, service, and other aspects affecting restaurants through comments. Hence, analyzing these comments can be valuable for others to choose a restaurant or to improve and develop their products and brands. Sentiment analysis utilizes text mining methods to extract, identify, and study emotions and subjective perceptions. Since consumers can use comments to choose a restaurant, this study seeks to provide sentiment analysis of their reviews on the Tripadvisor website about Iranian restaurants. This study is applied in nature, aiming to analyze and manually label 4000 comments from the Tripadvisor website regarding restaurants in ten tourist cities across Iran. It uses a standard extended long short-term memory algorithm for sentiment analysis, a deep learning neural network, and Python text mining packages for modeling. The results indicate that the F-Measure for all aspects exceeds 80%, indicating sufficient efficiency and accuracy of the aspect-based sentiment analysis model for restaurant reviews. The most significant features for customers of Iranian restaurants are the food and the atmosphere. This study represents one of the initial efforts to analyze comments posted on the Tripadvisor website concerning Iranian restaurants. Business owners in the tourism industry, especially restaurant owners, can use the proposed model to automatically and quickly analyze customer feedback, improve performance, and gain a competitive edge. The proposed model can also assist users of online platforms in analyzing the opinions of others, enabling them to make informed decisions more efficiently.

Green information technology is in the spotlight for organizations, helping them save money by using information technology (IT) to achieve the highest efficiency and thus reduce environmental impacts. One of the ways that can help organizations planning for deploying green IT is to evaluate green information technology maturity (GITM). Previous studies have referred to various criteria for green IT evaluation, most of which are qualitative criteria that are difficult to measure and evaluate in ambiguous conditions. The main objective of this study is to identify crucial criteria that affect the GITM level and to design a fuzzy inference system to assess the GITM level in any organization. While using a Mamdani Inference system, inputs can be verbal expressions or crisp values, and the output shows the level of maturity of green information technology. Since green IT knowledge is not modeled in previous studies, modeling it in the current study is a valuable step for organizations confused about various factors they should consider for going green. The main system criteria are the conditions of the data center, office environment, work practice, procurement, and corporate citizenship. Due to the generality of the model used for the knowledge base system development, organizations can use this system for the green IT maturity level determination. The presented inference system helps organizations understand their status of being IT green and plan for the following steps to accomplish their desired maturity level. The proposed inference system has been tested, validated, and used to determine the maturity level of Tehran municipality.