Conceptual
On the proceeded development of information technology, a lot more people have become to make use of online dating applications, and development might have been made worse by COVID-19 pandemic on these many years. Yet not, there was a sensation that off reading user reviews out of mainstream matchmaking programs was bad. To examine which phenomenon, i’ve used issue design so you’re able to mine bad ratings from mainstream relationships programs, and built a-two-stage servers learning design having fun with data dimensionality prevention and you can text message class in order to categorize user reviews off relationships apps. The analysis results show that: first, the reason why to your current negative reviews off relationship apps is actually generally focused in the battery charging system, phony account, membership and advertisements force system and coordinating procedure in the programs, advised involved improve pointers are proposed from the all of us; secondly, having fun with dominant part analysis to attenuate the fresh new dimensionality of text vector, and then playing with XGBoost design understand the reduced-dimensional data just after oversampling, a better category precision from reading user reviews can be acquired. Continue reading “Member review study out-of matchmaking apps centered on text message mining”