Skip to main content

Arikui - A Dubious User Detection System for Online Dating in Japan

James Neve ; Ivan Palomares (2018) — 2018 IEEE International Conference on Systems, Man, and Cybernetics (SMC)

Citation activity

Total citations · Google Scholar
Unavailable
Average per calendar year
Citation count unavailable

Counts reflect Google Scholar’s coverage and do not measure research quality. Source, calculation and limitations

View Original Source (opens in a new tab)

DOI: 10.1109/smc.2018.00395 (opens in a new tab)

Country:   Japan
Tags:   Class Imbalance, Dating Platform Spam, Dubious User Detection, Early Detection, Japan, Machine Learning, Online Dating, Pairs App
What do these research terms mean?
Preprint
A manuscript shared before formal peer review and publication. Check whether a later published version is available.
Dataset
A collection of data or examples for others to inspect or reuse. It can appear in Library search and topic mapping, but RSRC does not use dataset records as evidence in Research Insights.
Dissertation or thesis
Research submitted for an academic degree. This describes its format, not its reliability.
Journal article
An article published in a journal. This label alone does not establish peer review, study quality, or how well the findings apply elsewhere.
Qualitative research
Examines experiences, meanings, or processes, often through interviews or observations. It can explain how something happens without estimating how common it is.
Quantitative research
Uses numerical measurements to describe patterns or test relationships. A relationship between two measurements does not by itself show that one causes the other.
Systematic review
Uses a planned, documented method to find and assess research addressing a question. Its conclusions still depend on the included studies and what the search covered.
Meta-analysis
Statistically combines results from multiple studies. Combining studies does not remove weaknesses in their design or make unlike populations interchangeable.
Not classified
This record has no recognized label in this filter. It does not mean the publication used no method, or that no research exists.

Definitions draw on DataCite resource types; Cochrane review methods; NLM: association and causation. RSRC’s dataset and classification rules are explained in our methodology.

Citation tools
Download RIS Download BibTeX 


              
              

            

          

        


      

      

    

        
Transparency

        
Evidence and review status

        
This page contains AI-generated content. No human content review or subject-matter-expert review is recorded.

    


    

                    

                
Source basis

                

                    Downloaded PDF
                

            

                    

                
Source updates

                

                    No notice found at last check
                

            

                    

                
AI-generated page content

                

                    Yes
                

            

                    

                
Automated checks

                

                    Passed
                

            

                    

                
Administrative approval

                

                    Yes
                

            

                    

                
Human content review

                

                    Not recorded
                

            

                    

                
Subject-matter-expert review

                

                    Not recorded
                

            

            


    

        
            How this was prepared
        
        

                            

                    
Source basis

                    

                    
RSRC downloaded and privately stored a copy of the paper for internal analysis. The PDF is not offered to viewers from this page.


                                            
    
                                                            
  • 
                                    PDF added to RSRC:
                                    
                                
    • 
                                                    

                    
                                        

                

                            

                    
Source updates

                    

                    
No incoming update notice was found in the dated Crossref response. Coverage is incomplete, particularly for corrections and expressions of concern; this is not a guarantee that the source is valid or unchanged.


                                            
    
                                                            
  • 
                                    Last source-status attempt:
                                    
                                
    • 
                                                    

                    
                                        

                

                            

                    
AI-generated page content

                    

                    
AI-generated research notes displayed on this page: Synopsis, Identified gaps, Methods, Limitations, Future work. The paper itself is not described as AI-generated.


                                            
    
                                                            
  • 
                                    Document analysis recorded:
                                    
                                
    • 
                                                            
  • 
                                    Synopsis generation recorded:
                                    
                                
    • 
                                                            
  • 
                                    Page record updated:
                                    
                                
    • 
                                                    

                    
                                        

                

                            

                    
Automated checks

                    

                    
The current, source-bound synopsis passed the recorded versioned publication checks.


                                            
    
                                                            
  • 
                                    Checks completed:
                                    
                                
    • 
                                                    

                    
                                            

                            
                                View passed checks (3)
                            
                            
    
                                                                    
  • Length, completeness, repetition, refusal, boilerplate, and active-markup screening
    • 
                                                                    
  • Numerical claims checked against the available source text
    • 
                                                                    
  • English-source lexical grounding check
    • 
                                                            

                        

                                        

                

                            

                    
Administrative approval

                    

                    
An authenticated administrator approved the bibliographic record for public Library display. This is not a review of every research claim.


                                            
    
                                                            
  • 
                                    Approved for public display:
                                    
                                
    • 
                                                    

                    
                                        

                

                            

                    
Human content review

                    

                    
No human review is recorded for the AI-generated content displayed on this page.


                    
                                        

                

                            

                    
Subject-matter-expert review

                    

                    
RSRC has not recorded review of this content by a subject-matter or methods expert.


                    
                                        

                

                    

    


    

        Review-state definitions

        Found a possible error?
        Request a correction.
    



    



              

        
                  

            
Synopsis

            
This publication introduces Arikui, a dubious-user detection system designed for a Japanese online dating service. The authors frame the problem as distinguishing accounts used for purposes beyond dating—such as sales or multi-level marketing—from genuine users. The approach combines three data streams: profile data, early user behavior within the first hours after registration, and external social-media data (Facebook) to assess a user’s likelihood of being a spammer. The system aims to enable rapid, early-stage detection (within about 24 hours of registration) so that dubious users can be flagged before exchanging contact details or meeting, thereby reducing potential harm to other users and easing workloads for customer service. The technical solution comprises three components: a profile filter, a behavior filter, and a data-processing pipeline implemented as Python models on cloud infrastructure. The profile filter uses a random forest classifier trained on structured profile fields and free-text data, with a dataset that includes labeled dubious users. The behavior filter analyzes temporal patterns of activity, using a convolutional neural network to capture how behavior evolves over time, such as heightened activity shortly after joining and early initiation of conversations. The authors report that the profile filter achieved high cross-validated accuracy on their data, and the behavior filter achieved slightly higher accuracy under cross-validation. Combined, the system feeds alerts to customer-service personnel who review and mark accounts as dubious or normal. The authors acknowledge limitations inherent in their setting, including unbalanced class distributions and imperfect data quality, which can produce false positives. They suggest improvements such as larger training sets and longer temporal analysis, while emphasizing the system’s practical value in reducing exposure to dubious users within Eureka’s Pairs service.

          

        
                  

            
Identified Gaps

            
The study identifies unresolved operational and measurement gaps: the true prevalence of dubious users is unknown because unreported users may remain undetected; severe class imbalance produces many false positives at practical thresholds; and the available training set was too small for image-based neural-network analysis or separate gender-specific behavioral models.

          

        
                  

            
Methods

            
Using data from Japan's Pairs dating app, the authors built a deployed early-warning system combining a random-forest profile filter and a convolutional neural-network behavioral filter. The profile model used 30,000 customer-service-labeled dubious users and 30,000 normal users, split 90/10 for training and testing. It incorporated profile, Facebook, and text-derived features. The behavioral model used hourly profile updates, likes, and message activity during the first 24 hours. Customer-service staff reviewed system alerts.

          

        
                  

            
Limitations

            
Customer-service labels likely missed some dubious users, meaning the normal-user data may contain mislabeled cases. Balanced cross-validation accuracy did not translate directly to live deployment because dubious users were a small, unknown fraction of users, creating substantial false positives. The 30,000 dubious-user examples constrained model options, including separate models by gender and image analysis. The true number of undetected spammers could not be estimated.

          

        
                  

            
Future Work

            
Increase the training set to support potentially more accurate or alternative models. Evaluate image-based detection using profile photos when sufficient training data are available. Assess behavioral data over periods longer than 24 hours, balancing possible accuracy gains against the operational need for early intervention.

          

              

    
    

        
Related Research Insights

        

            See how this publication connects to RSRC's living evidence syntheses through current citations and research-topic mapping.