Skip to main content
. 2022 Jul 27;7:863126. doi: 10.3389/frma.2022.863126

Table 23.

Comparison of the terms predicted in 2015 for the next 5 years (2016–2020) with the actual observations on these years (Predictions are in italics. Terms correctly predicted to appear among the 10 top terms are marked in yellow, term correctly predicted at its rank is marked in green).

Predicted in 2015 Observed in 2020
Prediction 2016 Prediction 2017 Prediction 2018 Prediction 2019 Prediction 2020 Rank Observation 2016 Observation 2017 Observation 2018 Observation 2019 Observation 2020
Dataset Dataset Dataset Dataset Dataset 1 Dataset Dataset Dataset Dataset Dataset
DNN DNN DNN DNN DNN 2 Annotation Embedding Embedding Embedding Embedding
Annotation Neural network Neural network Neural network Neural network 3 DNN DNN Annotation Encoder BERT
POS SR RNN RNN RNN 4 embedding LSTM LSTM LSTM Annotation
Neural network Classifier POS Parser Parser 5 SR SR DNN Decoder Encoder
Classifier LM Parser SR SR 6 LSTM RNN SR LM LM
Parser POS Annotation LM Metric 7 POS Annotation RNN Metric Transformer
SR RNN Classifier Classifier POS 8 Classifier Neural network Encoder BERT SR
LM parser SR Metric Parsing 9 Neural network Classifier Classifier SR Metric
HMM HMM Metric POS Classifier 10 RNN LM Neural network Annotation LSTM
Rightly predicted in 10 tops 7 7 8 4 3
Rightly predicted at rank 1 1 1 1 1