A Client-Centric Evaluation System to Evaluate Guest’s Satisfaction on Airbnb Using Machine Learning and NLP

Abstract

Understanding the determinants of satisfaction in P2P hosting is crucial, especially with the emergence of platforms such as Airbnb, which has become the largest platform for short-term rental accommodation. Although many studies have been carried out in this direction, there are still gaps to be filled, particularly with regard to the apprehension of customers taking into account their category. In this study, we took a machine learning-based approach to examine 100,000 customer reviews left on the Airbnb platform to identify different dimensions that shape customer satisfaction according to each category studied (individuals, couples, and families). However, the data collected do not give any information on the category to which the customer belongs to. So, we applied natural language processing (NLP) algorithms to the reviews in order to find clues that could help us segment them, and then we trained two regression models, multiple linear regression and support vector regression, in order to calculate the coefficients acting on each of the 6 elementary scores (precision, cleanliness, check-in, communication, location, and value) noted on Airbnb, taking into account the category of customers who evaluated the performance of their accommodation. The results suggest that customers are not equally interested in satisfaction metrics. In addition, disparities were noted for the same indicator depending on the category to which the client belongs to. In light of these results, we suggest that improvements be made to the rating system adopted by Airbnb to make it suitable for each category to which the client belongs to.

Effect of word embedding vector dimensionality on sentiment analysis through short and long texts

“…In addition, the number of parameters for a word embedding or a model that relies on word embeddings, such as recurrent neural networks, is usually a linear or quadratic function of dimensionality, which affects training time and computational costs [40]. Our study showed that this dimensionality is 100d for short texts, such as tweets or comments related to blog posts and 300d for long texts, such as IMDb comments or reviews left on Airbnb [41]. Indeed, mapping each word of the corpus on such a large number of dimensions, and even more so if the corpus is large, could increase the complexity of the model, slow down the training speed and add inferential latency, which has as a direct consequence, the impossibility to deploy the model on real tasks.…”

Predicting Bitcoin Trends Through Machine Learning Using Sentiment Analysis with Technical Indicators

“…Reference [37] used seven types of machine learning models to classify airline customer feedback sentiments into three classes and revealed that random forest had the best performance with an F1-score of 0.86. In addition to the studies reviewed above, machine learning has exhibited excellent efficiency in prediction tasks and is being actively applied in several domains [38][39][40][41].…”

LSTM, VADER and TF-IDF based Hybrid Sentiment Analysis Model

“…Regression is a method of modeling a variable (called target) as a function of independent predictors (called features), where the algorithm involved tries to find causal relationships between the variables [52].…”