Topic Subject Creation Using Unsupervised Learning for Topic Modeling

  •  Rashid Mehdiyev    
  •  Jean Nava    
  •  Karan Sodhi    
  •  Saurav Acharya    
  •  Annie Ibrahim Rana    


We address the problem of topic mining and labelling in the domain of retail customer communications to summarize the subject of customers inquiries. The performance of two popular topic mining algorithms - Non-Negative Matrix Factorization (NMF) and Latent Dirichlet Allocation (LDA) – were compared, and a novel method to assign topic subject labels to the customer inquiries in an automated way was proposed. Experiments using a retailer’s call center data verify the efficacy and efficiency of the proposed topic labelling algorithm. Furthermore, the evaluation of results from both the algorithms seems to indicate the preference of using Non-Negative Matrix Factorization applied to short text data.

This work is licensed under a Creative Commons Attribution 4.0 License.
  • ISSN(Print): 1913-8989
  • ISSN(Online): 1913-8997
  • Started: 2008
  • Frequency: semiannual

Journal Metrics

WJCI (2022): 0.636

Impact Factor 2022 (by WJCI):  0.419

h-index (January 2024): 43

i10-index (January 2024): 193

h5-index (January 2024): N/A

h5-median(January 2024): N/A

( The data was calculated based on Google Scholar Citations. Click Here to Learn More. )