We propose an indicator to estimate the effects of transports, culture, crowd, managerial, accommodation dimensions on price determination for Airbnb accommodations. The indicator is defined using a proportional odds model in two steps. In the first phase, we estimate for each accommodation the probability to belong to a specific category of the price moving from very-low to very-high. Then, we estimate the average concentration index to identify which is the price class more likely for each observation and which dimension can better explain the price. Then, we assign the price class to each observation using the median of the probabilities of the model’s fitted values and identify the most significant dimension. Afterward, we aggregate the concentration index calculated for each observation for the district of Trastevere in Rome, with the aim to identify its most relevant dimensions. The results highlight a significant impact of the managerial and accommodation dimensions.
Si propone un indicatore per stimare l’effetto delle dimensioni del trasporto, cultura, affollamento, management e caratteristiche dell’alloggio sulla determinazione del prezzo per gli alloggi Airbnb. L’indicatore e definito utilizzando un ` proportional odds model come spiegato nel Vector Generalized Additive Model. L’indicatore viene costruito in due fasi. Nella prima fase si stima la probabilita di ogni alloggio di appartenere ad una delle cinque categorie di prezzo che vanno da molto basso a molto alto. Successivamente, si stima l’indice di concentrazione medio per identificare la fascia di prezzo in cui ricade l’osservazione e quale dimensione possa spiegare meglio il prezzo. Poi, la fascia di prezzo viene assegnata a ciascuna osservazione utilizzando la mediana del vettore di probabilita e si identificano le dimensioni piu significative. Successivamente, l’indice di concentrazione ` calcolato per ciascuna osservazione viene aggregato per il quartiere di Trastevere di Roma, con l’obiettivo di individuare la dimensione piu rilevanti per il quartiere. ` I risultati hanno evidenziato un impatto significativo della dimensione management e caratteristiche dell’alloggio.
A statistical model to identify the price determinations: the case of Airbnb
Contu, Giulia;Frigau, Luca;Zammarchi, Gianpaolo;Mola, Francesco
2021-01-01
Abstract
We propose an indicator to estimate the effects of transports, culture, crowd, managerial, accommodation dimensions on price determination for Airbnb accommodations. The indicator is defined using a proportional odds model in two steps. In the first phase, we estimate for each accommodation the probability to belong to a specific category of the price moving from very-low to very-high. Then, we estimate the average concentration index to identify which is the price class more likely for each observation and which dimension can better explain the price. Then, we assign the price class to each observation using the median of the probabilities of the model’s fitted values and identify the most significant dimension. Afterward, we aggregate the concentration index calculated for each observation for the district of Trastevere in Rome, with the aim to identify its most relevant dimensions. The results highlight a significant impact of the managerial and accommodation dimensions.File | Dimensione | Formato | |
---|---|---|---|
A statistical model to identify the price determinations the case of Airbnb.pdf
accesso aperto
Tipologia:
versione editoriale (VoR)
Dimensione
397.75 kB
Formato
Adobe PDF
|
397.75 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.