e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 444 Gizem Kayar Yasar University Tolga Sümer TED University Furkan Soytürk TED University Galip Erkin Doruk TED University Cihan Çobanoğlu University of South Florida Sarasota-Manatee Explore Music Data to Enhance Customer Satisfaction Restaurant-like service areas have been adapting different technologies to enhance customer satisfaction for many years. In this LBR, we share our research idea about how to integrate music data and its analysis for this purpose. In the first part, we propose a voting system to carry your favorite song to the top of the list to be played next in your place. In the second part, we propose a recommendation system to find a place that suits your music requirements in your close proximity. Our preliminary survey results for the first part and the data analysis results for the second part shows that our approach has a promising potential for customer satisfaction. Key words: Data Analysis, Spotify API, Customer Satisfaction Gizem Kayar Computer Engineering Department Yaşar University Kampüs Caddesi, Bornova, İzmir Turkey Email: gizem.kayar@yasar.edu.tr Tolga Sümer, Furkan Soytürk, Galip Erkin Doruk Computer Engineering Department TED University Ziya Gokalp Cad. 48 Çankaya, Ankara Turkey Email: tolga.sumer@tedu.edu.tr, furkan.soyturk@tedu.edu.tr, galip.doruk@tedu.edu.tr Cihan Çobanoğlu College of Hospitality & Tourism Leadership (CHTL) University of South Florida Sarasota-Manatee 8350 N Tamiami Trail, Sarasota, FL 34243 United States Email: cihan@sar.usf.edu http://ertr.tamu.edu/ mailto:gizem.kayar@yasar.edu.tr mailto:furkan.soyturk@tedu.edu.tr galip.doruk@tedu.edu.tr cihan@sar.usf.edu e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 445 Gizem Kayar is an assistant professor and a faculty member at Yaşar University Computer Engineering Department. Her research focuses on Computer Graphics, Game Programming, VR/AR based gamified health applications, and integration of mentioned emerging technologies also in tourism. Furkan Soytürk is a TED University graduate and a freelance Computer Engineer. Tolga Sümer is a Computer Engineering student at TED University and a game programmer at OTTO Games, Turkey. Galip Erkin Doruk is a Computer Engineering student at TED University. Cihan Çobanoğlu is the McKibbon Endowed Chair Professor of the College of Hospitality & Tourism Leadership (CHTL) at the University of South Florida Sarasota-Manatee (USFSM), and he also serves as the director of the M3 Center for Hospitality Technology and Innovation and coordinator of International Programs for the College of hospitality & Tourism Leadership. His research involves the use and impact of technology in the hospitality industry. http://ertr.tamu.edu/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 446 Introduction Technology spread rate increases every single day and we are now very familiar with its usage in cafes, restaurants, etc.. We get used to see digitized menus, tabletop e-waiter and checkout screens or tabletop gamepads. The way we play music has also changed in a similar way. Old cassettes have been replaced by CDs, which is later substituted by MP3s, then Youtube playlists and finally by Spotify. Good music relieves stress, connects people and therefore, we believe that, has a great potential to enhance customer satisfaction and continuity. Being aware of the usage of Spotify technology, we propose an application using which people can find a place that matches their music needs in their close proximity, and vote for their favourite song to move it on top to be played next. Literature Review Improving customer satisfaction in service areas is one of the most popular research fields in tourism and hospitality. There are many researchers who focus on more traditional points, such as price, food quality, operational efficiency and physical conditions. On the other hand, some researchers investigate how to integrate emerging technologies. Although there are hundreds of valuable publications for both cases, we can mention only some of them here (e.g. Andaleeb and Conway, 2006, Ryu and Lee, 2017 or Barlan-Espino, 2017 for the traditional applications and Pantelidis, 2009, Koutroumanis, 2011 or Cavusoglu, 2019 for the latter), since they are beyond the scope of this LBR. For those who are interested in more comprehensive study, we can refer to Demicco et.al, 2015. http://ertr.tamu.edu/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 447 Methodology Voting System As mentioned before, our application has two parts. In the first part, we propose a voting system. Our application works as follows: • Enter a restaurant, register by tracking its QR-code • Open your application and see the currently active playlist of the restaurant • Vote for the song you want to listen next Meanwhile, we count for the votes and top-rated song is carried to the top of the list to be played next. For such an application, we need to handle both the front-end and back-end sides. At the server side, we store the information about the votes (Song ID, Voter ID, Playlist ID etc.), count the votes while the client-side sends the votes to the server, send the clients new nominations when each voting session ends and also handle security issues like ignoring votes from already-voted clients (done via checking the voter ID). At the client side, we view the votable songs and number of votes those songs get, view the full playlist, view currently playing song and the last voting’s winner, and send vote packets to the server (those packets will carry information like song ID, voter ID, playlist ID). Regarding technical details, we handle the login system credentials using Google Firebase and all the credentials are stored at their servers. The user interface is developed using cross-platform React-Native framework. The reason we prefer React-Native is its support in all platforms. http://ertr.tamu.edu/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 448 Recommendation System In the second part, we propose a recommendation system. This time, the user selects one of her/his own Spotify playlists and our algorithm tries to find the best matching restaurant among all the registered places in our system in her/his close proximity (e.g. 5 km2). Our algorithm checks the affinity of the songs in our playlist with the songs in the active playlist of all registered restaurants. For this purpose, we use EchoNest’s audio analysis data. This data includes acousticness, danceability, energy and similar features of each song. Our algorithm uses these features and find a mean average for each song for comparison purposes. Findings Voting System In order to support our claim, we have started our work with an experimental study on a group of 44 people (to see our anonymous survey, please visit: https://www.smartsurvey.co.uk/s/KDR5C/). Our study group consists of frequent restaurant/café visitors. We have informed all respondents about our Spotify-based voting system and asked them four questions. The results of our survey can be found in Fig. 1. Our initial findings show that more than 90% of the respondents want to see our application in their location and 77% of them are interested in voting. On the other hand, satisfaction ratio of question 3 is more than 78% which shows that this app can enhance customer satisfaction. Finally, we wanted to ask the respondents whether our app can cause any unwanted behaviour in between customers. 86% of the attendees answered it “No”. According to survey results, we believe that our application has a potential to be used in service areas. http://ertr.tamu.edu/ https://www.smartsurvey.co.uk/s/KDR5C/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 449 Figure 1. Our survey results. http://ertr.tamu.edu/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 450 Recommendation System We compared our method with popular K-Nearest Neighbors (KNN) algorithm. The accuracy rate of KNN is not always reliable because it uses only two most different features for comparisons. Instead of using only two most distinct audio features of the playlists, we decided to use every audio feature of songs to get more accurate comparison results. In this way, we compared two playlists with respect to each other by the average values of each audio features (Eqn. 1 and 2), obtained better matches for playlist genres and obtained more reliable results. x = |i-i1|+|j-j1|+|k-k1|+|l-l1|+|m-m1|+|n-n1|+|o-o1| Eqn. 1 y = (7-x) * 100 / 7 Eqn. 2 Value of ‘y’ is the similarity percentage of selected playlist with respect to compared playlist. i,j,k,l,m,n, and o stand for mean of danceability, energy, speechiness, acousticness, instrumentalness, liveness and valence values of selected playlist, and their subscripted versions stand for the same values of compared playlist in order. Conclusion In this LBR, we have presented a mobile application with two parts. In the first part, we have shown the potential of using music voting system in restaurants/cafes to increase the customer satisfaction. In the second part, we have proposed a way to analyse Spotify music data to find the best matching place with your music interest. The results show that our application can be used by any restaurant easily and has a great potential to drive customer satisfaction positively. Although the development phase is almost finished and preliminary results show the customer potential, our project is still ongoing. The next phase is to register http://ertr.tamu.edu/ e-Review of Tourism Research (eRTR), Vol. 17, No. 3, 2019 http://ertr.tamu.edu 451 many places in our system and to try our system with real customers to analyze its true potential. References Andaleeb, S., Conway, C. (2006). Customer satisfaction in the restaurant industry: An examination of the transaction-specific model. Journal of Services Marketing 20(1):3- 11. Barlan-Espino, A.G. (2017) Operational Efficiency And Customer Satisfaction of Restaurants: Basis For Business Operation Enhancement. Asia Pacific Journal of Multidisciplinary Research, Vol. 5, No. 1, February 2017 Cavusoglu, M. (2019). An Analysis of Technology Applications in the Restaurant Industry. University of South Florida Scholar Commons, 1-5. DeMicco, F., Cobanoglu, C., Dunbar, J., Grimes, R., Chen, C., & Keiser, J. R. (2015). Restaurant Management: A Best Practices Approach. Dubuque, IA: Kendall/Hunt. Koutroumanis, D.A. (2011). Journal of Applied Business and Economics vol. 12(1) 2011. Pantelidis, I. (2009). High tech foodservice; an overview of technological advancements. CHME 18th Annual Research Conference,At: Eastbourne, UK, May 2009. Ryu, K., Lee, J.S. (2017) Journal of Hospitality & Tourism Research, Vol. 41, No. 1, January 2017, 66 –92. DOI: 10.1177/1096348013515919 http://ertr.tamu.edu/