Papers › Inferring the source of official texts: can SVM beat ULMFiT?

Inferring the source of official texts: can SVM beat ULMFiT?

2 Mar 2020International Conference on Computational Processing of the Portuguese Language 2020 3archive 2025-07-28

Pedro Henrique Luz de Araujo, Teófilo Emidio de Campos, Marcelo Magalhães Silva de Sousa

Official Gazettes are a rich source of relevant information to the public. Their careful examination may lead to the detection of frauds and irregularities that may prevent mismanagement of public funds. This paper presents a dataset composed of documents from the Official Gazette of the Federal District, containing both samples with document source annotation and unlabeled ones. We train, evaluate and compare a transfer learning based model that uses ULMFiT with traditional bag-of-words models that use SVM and Naive Bayes as classifiers. We find the SVM to be competitive, its performance being marginally worse than the ULMFiT while having much faster train and inference time and being less computationally expensive. Finally, we conduct ablation analysis to assess the performance impact of the ULMFiT parts.

PaperPDFCode

Code

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Text ClassificationTransfer Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Text Classification DODF Data ULMFiT (pre-trained vocab, no gradual unfreezing) Average F1 0.8918 #1 of 5 Archive leaderboard report
Text Classification DODF Data ULMFiT (pre-trained vocab, no gradual unfreezing) Weighted F1 0.9257 #1 of 5 Archive leaderboard report
Text Classification DODF Data SVM + word counts (pre-trained vocab) Average F1 0.8782 #2 of 5 Archive leaderboard report
Text Classification DODF Data SVM + word counts (pre-trained vocab) Weighted F1 0.9049 #2 of 5 Archive leaderboard report
Text Classification DODF Data SVM + tf-idf (no pre-trained vocab) Average F1 0.8755 #3 of 5 Archive leaderboard report
Text Classification DODF Data SVM + tf-idf (no pre-trained vocab) Weighted F1 0.8917 #3 of 5 Archive leaderboard report
Text Classification DODF Data ULMFiT (no pre-trained vocab) Average F1 0.8469 #4 of 5 Archive leaderboard report
Text Classification DODF Data ULMFiT (no pre-trained vocab) Weighted F1 0.8974 #4 of 5 Archive leaderboard report
Text Classification DODF Data ULMFiT (pre-trained vocab) Average F1 0.8374 #5 of 5 Archive leaderboard report
Text Classification DODF Data ULMFiT (pre-trained vocab) Weighted F1 0.9088 #5 of 5 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AWD-LSTMActivation RegularizationDiscriminative Fine-TuningDropConnectDropoutEmbedding DropoutLSTMSVMSigmoid ActivationSlanted Triangular Learning RatesTanh ActivationTemporal Activation RegularizationULMFiTVariational DropoutWeight Tying

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections