Citation :

Jaspreet Sodhi, Dr.Anuradha Misra , 2026. "AI Enabled Adaptive Digital System " ESP International Journal of Advancements in Science & Technology (ESP-IJAST)  Volume 4, Issue 2: 18-21.

Abstract :

Hash tags play a pivotal role in increasing the visibility, discoverability, and engagement of social media content, particularly in the tourism sector where users search for travel inspiration using thematic tags. This project presents a multi-label hash tag recommendation system for tourism related social media posts focused on Indian destinations. A key innovation is the use of weak supervision techniques to overcome the challenge of absent labeled data, a common limitation in real world applications. Two structured datasets containing tourism metadata such as site descriptions, geographical zones, ratings, and seasonal recommendations were used to generate synthetic social media style posts through template based natural language generation. A preprocessing pipeline including normalization, stop word removal, and lemmatization was applied, followed by TF-IDF factorization for feature extraction. A hybrid labeling approach combined rule based keyword matching with semantic similarity using cosine similarity between TF-IDF representations and prototype sentences. Manual filtering mechanism reduced noise and missioned hash tags. Finally, a logistic regression One-vs.-Rest multi-label model was trained and evaluated using micro, macro, and sample based F1 scores, demonstrating effective hash tag prediction.

References :

[1] J. Coelho, P. Nit, and P. Madera, "A personalized travel recommendation system using social media analysis," 2018 IEEE International Congress on Big Data (Big Data Congress), pp. 260–263, 2018.

[2] S. M. S, A. Krishna, and A. P. R. S, "Travel Destination Recommendation System based on Machine Learning by Analysing Integra Review Comments and Hash tags," 2024 5th IEEE Global Conference for Advancement in Technology (GCAT), pp. 1–10, 2024.

[3] Cities dataset from https://www.kaggle.com/datasets/kirtandwivedi02/most-traveled-cities-in-india.

[4] Places dataset from https://www.kaggle.com/datasets/saketk511/travel-dataset-guide-to-indias-must-see-places.

[5] A. H. Caldron, M. G. Pérez, F. J. Garcia Clemente, G. M. Pérez, "Design of a recommender system based on users' behaviour and collaborative location and tracking", Journal of Computational Science, vol. 12, pp. 83-94, Jan 2016.

[6] P. Martins, P. Madera, Personalizing Places of Interest Using Social Media Analysis, Milwaukee, WI, 2015.

Keywords :

Machine learning, Natural Language Processing, Recommendation System, Tourism