Available on ORBi since
13 June 2022
Article (Scientific journals)
Using meaning instead of words to track topics
Poumay, Judicaël  ; Ittoo, Ashwin 
2022 • In Natural Language Processing and Information Systems
Peer reviewed
 

Files


Full Text
Topic_Tracking_in_Hierarchical_Topic_Modelling__Final_ (2).pdf
Author postprint (276.23 kB) Creative Commons License - Attribution

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
topic model; topic tracking; hierarchical topic
Abstract :
[en] The ability to monitor the evolution of topics over time is extremely valuable for businesses. Currently, all existing topic tracking methods use lexical information by matching word usage. However, no studies has ever experimented with the use of semantic information for tracking topics. Hence, we explore a novel semantic-based method using word embeddings. Our results show that a semantic-based approach to topic tracking is on par with the lexical approach but makes different mistakes. This suggest that both methods may complement each other.
Disciplines :
Computer science
Author, co-author :
Poumay, Judicaël ;  Université de Liège - ULiège > HEC Recherche > HEC Recherche: Business Analytics & Supply Chain Management
Ittoo, Ashwin ;  Université de Liège - ULiège > HEC Liège : UER > UER Opérations
Language :
English
Title :
Using meaning instead of words to track topics
Publication date :
2022
Journal title :
Natural Language Processing and Information Systems
Publisher :
Springer, Valencia, Spain
Special issue title :
27th International Conference on Applications of Natural Language to Information Systems
Peer reviewed :
Peer reviewed

Statistics


Number of views
31 (7 by ULiège)
Number of downloads
10 (2 by ULiège)

Bibliography


Similar publications



Contact ORBi