Summary of the paper
| Title | POS Multi-tagging Based on Combined Models |
|---|---|
| Authors | Yan Zhao and Gertjan van Noord |
| Abstract | In the POS tagging task, there are two kinds of statistical models: one isgenerative model, such as the HMM, the others are discriminative models, suchas the Maximum Entropy Model (MEM). POS multi-tagging decoding method includesthe N-best paths method and forward-backward method. In this paper, we use theforward-backward decoding method based on a combined model of HMM and MEM. IfP(t) is the forward-backward probability of each possible tag t, we firstcalculate P(t) according HMM and MEM separately. For all tags options in acertain position in a sentence, we normalize P(t) in HMM and MEM separately.Probability of the combined model is the sum of normalized forward-backwardprobabilities P norm(t) in HMM and MEM. For each word w, we select the best tagin which the probability of combined model is the highest. In the experiments,we use combined model and get higher accuracy than any single model on POStagging tasks of three languages, which are Chinese, English and Dutch. Theresult indicates that our combined model is effective. |
| Language | Corpus (creation, annotation, etc.) |
| Topics | Part of speech tagging, Language modelling, Corpus (creation, annotation, etc.) |
Full paper  |
POS Multi-tagging Based on Combined Models |
| Bibtex | @InProceedings{ZHAO10.470, author = {Yan Zhao and Gertjan van Noord}, title = {POS Multi-tagging Based on Combined Models}, booktitle = {Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10)}, year = {2010}, month = {may}, date = {19-21}, address = {Valletta, Malta}, editor = {Nicoletta Calzolari (Conference Chair), Khalid Choukri, Bente Maegaard, Joseph Mariani, Jan Odjik, Stelios Piperidis, Mike Rosner, Daniel Tapias}, publisher = {European Language Resources Association (ELRA)}, isbn = {2-9517408-6-7}, language = {english} } |