A Multilingual BERT-based classification of reviews for enhanced visitors’ experience analysis

A Multilingual BERT-based classification of reviews for enhanced visitors’ experience analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A Multilingual BERT-based classification of reviews for enhanced visitors’ experience analysis RICCARDO RICCIARDI, Marica Manisera This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4495603/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 Aug, 2025 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Cultural organizations today can rely on online platforms to study users’ opinions and the most discussed topics related to both general and specific cultural offerings. Despite data acquisition tools, managing unstructured databases remains a hurdle. To overcome this, we propose a classification model that transforms unorganized data into a structured thematic database. The specific case pertains to the Italian city of Brescia. We build a language model that classifies online reviews into four semantic areas defined by the key attractions of the city. We fine-tuned the pre-trained Multilingual BERT model in a multiclassification task. The model shows promising results based on traditional performance metrics. Additionally, clusters of reviews have been detected by applying the HDBSCAN algorithm on their vector representations produced by the model. As a transformation of the chi-square statistic, the Keyness statistic has been employed to extract cluster-specific keywords, which have proven to be highly consistent with the characteristics and offerings of the key cultural attractions, further confirming the good performance of the model. Results show that the proposed model can be profitably employed by policymakers and managers of cultural tourism institutions to understand textual data and derive relevant insights about visitors’ experience at specific attractions of interest. Physical sciences/Mathematics and computing/Computational science Physical sciences/Mathematics and computing/Information technology Physical sciences/Mathematics and computing/Statistics Cultural Tourism Natural Language Processing BERT Fine-tuning Document clustering Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 26 Aug, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 06 Jan, 2025 Reviews received at journal 01 Nov, 2024 Reviewers agreed at journal 05 Oct, 2024 Reviews received at journal 04 Oct, 2024 Reviewers agreed at journal 03 Oct, 2024 Reviewers invited by journal 03 Oct, 2024 Editor assigned by journal 05 Sep, 2024 Editor invited by journal 07 Jul, 2024 Submission checks completed at journal 03 Jul, 2024 First submitted to journal 29 May, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4495603","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":332916023,"identity":"c3c6290b-548f-4e8a-afd5-0a7f88b66be3","order_by":0,"name":"RICCARDO RICCIARDI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIie3PsU7DMBCA4bMsucsRr0Sp+gwnZUAoEnmVRh06kp2BIKSy9AHgRZg6uLqhIw9Ah2bJxBAWVKQM2ILSDjHKyOB/SKxEn84HEAr9x+TvU1QAZN8sjT2Yw69BRE1/iM+cwO8MOmjAN+biQdbt5yqbQMJ37b68uo42+JGU5TaHkTZ9ZMwqjZfNPIWouI+XNLt84rPn5JGaovJc7FyiAjRcVCgWEkgSOYLEU98ujojuSG4twcaR/C8iT6awI8oR4b+YSpOxmacKhdtlQ7HdLkO7y0JK6iWa6/c3k000jup2391Q9ML1K3bbXOv1rnfMITXgSygUCoWG9gUDsk6ECC+x0AAAAABJRU5ErkJggg==","orcid":"","institution":"University of Brescia","correspondingAuthor":true,"prefix":"","firstName":"RICCARDO","middleName":"","lastName":"RICCIARDI","suffix":""},{"id":332916024,"identity":"c90c453e-0b99-4b69-b147-7c890742775b","order_by":1,"name":"Marica Manisera","email":"","orcid":"","institution":"University of Brescia","correspondingAuthor":false,"prefix":"","firstName":"Marica","middleName":"","lastName":"Manisera","suffix":""}],"badges":[],"createdAt":"2024-05-29 08:45:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4495603/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4495603/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-09418-9","type":"published","date":"2025-08-26T15:57:18+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":90345048,"identity":"f5599c7f-f28c-4530-a3c4-a1ee89312f74","added_by":"auto","created_at":"2025-09-01 16:09:47","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1032712,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4495603/v1_covered_02650c9b-f82d-4473-9f34-c4e0afaf9c93.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Multilingual BERT-based classification of reviews for enhanced visitors’ experience analysis","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Cultural Tourism, Natural Language Processing, BERT, Fine-tuning, Document clustering","lastPublishedDoi":"10.21203/rs.3.rs-4495603/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4495603/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCultural organizations today can rely on online platforms to study users’ opinions and the most discussed topics related to both general and specific cultural offerings. Despite data acquisition tools, managing unstructured databases remains a hurdle. To overcome this, we propose a classification model that transforms unorganized data into a structured thematic database.\u003c/p\u003e\n\u003cp\u003eThe specific case pertains to the Italian city of Brescia. We build a language model that classifies online reviews into four semantic areas defined by the key attractions of the city. We fine-tuned the pre-trained Multilingual BERT model in a multiclassification task. The model shows promising results based on traditional performance metrics. Additionally, clusters of reviews have been detected by applying the HDBSCAN algorithm on their vector representations produced by the model. As a transformation of the chi-square statistic, the Keyness statistic has been employed to extract cluster-specific keywords, which have proven to be highly consistent with the characteristics and offerings of the key cultural attractions, further confirming the good performance of the model. Results show that the proposed model can be profitably employed by policymakers and managers of cultural tourism institutions to understand textual data and derive relevant insights about visitors’ experience at specific attractions of interest.\u003c/p\u003e","manuscriptTitle":"A Multilingual BERT-based classification of reviews for enhanced visitors’ experience analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-31 18:52:01","doi":"10.21203/rs.3.rs-4495603/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-01-06T11:10:42+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-01T13:54:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"169449185086970862160198337886397843891","date":"2024-10-05T18:44:40+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-10-04T06:51:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"138829159625288544440549906577842855907","date":"2024-10-03T23:47:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-10-03T18:42:58+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-09-05T05:56:39+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-07-07T18:52:16+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-07-04T02:33:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-05-29T08:44:21+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2181eaa4-bd92-4587-8219-78298162bcde","owner":[],"postedDate":"July 31st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":35247408,"name":"Physical sciences/Mathematics and computing/Computational science"},{"id":35247409,"name":"Physical sciences/Mathematics and computing/Information technology"},{"id":35247410,"name":"Physical sciences/Mathematics and computing/Statistics"}],"tags":[],"updatedAt":"2025-09-01T16:06:31+00:00","versionOfRecord":{"articleIdentity":"rs-4495603","link":"https://doi.org/10.1038/s41598-025-09418-9","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-08-26 15:57:18","publishedOnDateReadable":"August 26th, 2025"},"versionCreatedAt":"2024-07-31 18:52:01","video":"","vorDoi":"10.1038/s41598-025-09418-9","vorDoiUrl":"https://doi.org/10.1038/s41598-025-09418-9","workflowStages":[]},"version":"v1","identity":"rs-4495603","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4495603","identity":"rs-4495603","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}