MétaCan
Menu
Back to cohort
Record W1596584299 · doi:10.21427/9w8z-hc83

Sweetening the Dataset : Using Active Learning to Label Unlabelled Datasets

2021· article· en· W1596584299 on OpenAlexfundno aff
Rong Hu, Brian Mac Namee, Sarah Jane Delany

Bibliographic record

VenueArrow@dit (Dublin Institute of Technology) · 2021
Typearticle
Languageen
FieldComputer Science
TopicMachine Learning and Algorithms
Canadian institutionsnot available
FundersScience Foundation IrelandCanada Millennium Scholarship Foundation
KeywordsComputer scienceMachine learningArtificial intelligenceActive learning (machine learning)Set (abstract data type)Focus (optics)Supervised learningExpert systemTraining setArtificial neural network

Abstract

fetched live from OpenAlex

Supervised machine learning approaches assume the existence of a large collection of manually labelled examples of the problem under consideration. However, in many cases such a collection does not exist and creating one is time consuming and expensive. This can be a barrier to the use of supervised learning in certain situations, particularly when the doubt as to whether the system will work or not makes the cost of creating a dataset unjustifable. Active learning is a machine learning technique that has been used widely to create classification systems in the absence of large numbers of labelled examples, but that can also be used to create such collections. This paper will describe a system that uses active learning to label large collections of unlabelled data. We will show that the system can create an accurately labelled dataset aproximately 10 times the size of the set of examples manually labelled by an expert. The experiments described are based on recipe data from the 1st Computer Cooking Contest to be held at ECCBR'08 and focus on identifying those recipes in the set that are desserts.

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.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.015
metaresearch head score (Gemma)0.034
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.015
Threshold uncertainty score0.078

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0150.034
Meta-epidemiology (narrow)0.0020.001
Meta-epidemiology (broad)0.0020.002
Bibliometrics0.0060.005
Science and technology studies0.0020.002
Scholarly communication0.0060.007
Open science0.0070.007
Research integrity0.0060.007
Insufficient payload (model declined to judge)0.0030.003

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 head0.024
GPT teacher head0.297
Teacher spread0.273 · 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

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designNot applicable
Domainnot available
GenreMethods

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations12
Published2021
Admission routes1
Has abstractyes

Explore more

Same venueArrow@dit (Dublin Institute of Technology)Same topicMachine Learning and AlgorithmsFrench-language works237,207