MétaCan
Menu
Back to cohort
Record W2295690276 · doi:10.1109/icmla.2015.120

Predicting Churn of Expert Respondents in Social Networks Using Data Mining Techniques: A Case Study of Stack Overflow

2015· article· en· W2295690276 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

Venuenot available
Typearticle
Languageen
FieldBusiness, Management and Accounting
TopicCustomer churn and segmentation
Canadian institutionsUniversity of Saskatchewan
Fundersnot available
KeywordsComputer scienceRandom forestLogistic regressionAsset (computer security)IncentiveData miningMachine learningPrecision and recallSupport vector machineRecallArtificial neural networkData scienceArtificial intelligenceSocial network (sociolinguistics)World Wide WebComputer securitySocial media

Abstract

fetched live from OpenAlex

In Q&A social networks, the few respondents that answer most of the questions are an asset to that network. Being able to predict the churn of these expert respondents will enable the owners of such network put things in place in order to keep them. In this paper, we predicted the churn of expert respondents in Stack Overflow. We identified experts based on the InDegree of the respondents and the value of the incentives earned by these experts from the questions they have answered in the past. Using four data mining techniques: logistic regression, neural networks, support vector machines and random forests, we predicted user churn and evaluated our results with four evaluation metrics: percentage correctly classified, area under receiver operating characteristic curve, precision and recall. Of the four data mining algorithms, random forests performed best with PCC of 76%, ROC area of 0.82, precision of 0.76 and recall of 0.77.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.954
Threshold uncertainty score0.995

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.186
GPT teacher head0.368
Teacher spread0.182 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Quick stats

Citations13
Published2015
Admission routes1
Has abstractyes

Explore more

Same topicCustomer churn and segmentationFrench-language works237,207