- 年份:2019 年
- 編號:95
- Topic分類:4
- Topic分數:0.2807269101
- Publish:The Journal of Academic Librarianship
- 作者:Kevin W.Walker、 Zhehan Jiang
Keywords:Acquisitions, Assessment, DDA, Machine, learning, Boosting
Abstract:Application of adaptive boostingAdaBoostin demand-driven acquisitionDDApredictionA machine-learning approach Demand-driven acquisitionDDAprograms are playing an increasingly important role in academic librariesHoweverthe literature surrounding this topic illustrates the wide-rangingand frequently unpredictableresults of DDA implementationAs uncertainty aboundslibrarians continue to seek out deeper understandings of those processes driving the use and purchase of DDA materialsImplicit in this search is a desire to understand how local environmental factors and user preferences dictate broader collection use and purchasing patternsA small number of these studies have sought deeper insights through predictive modelingthough success has been limitedFollowing this line of inquirythis study explores how machine learning might enable more effective collection development and management strategies through the predictive modeling of complex collection use and purchasing patternsThis research describes a replicable implementation of an adaptive boostingAdaBoostmodel that predicts the likelihood of DDA titles being triggered for purchaseThe predictive capacity of this model is compared against a more traditional logistic regression modelThis studys results show that the AdaBoost model possesses much higher predictive capacity than a regression-based model informed by the same set of predictorsThe AdaBoost algorithmonce trained with local DDA dataprovides accurate predictions in 82of cases Library AutomationThrough automatic prediction which book is most likely to be purchasedthe process of automation is automatedCollection classificationpredict the predictive models to predict which bookmates are most likely to be triggered to purchasewhich will help better classify and manage collectionsThis research explores how to use machine learning to conduct more effective book development and management strategiesand through predicting complicated book use and purchasing modelsIn particularthis study describes a copy of the possibility of the DDA title triggering to purchase
© All Rights LibAiRsystem.

