| 103 | |
| 104 | |
| 105 | == HaBiT Evaluation Workshop 2017 == |
| 106 | |
| 107 | Feb 16-21 2017, Faculty of Informatics, Brno, Czech Republic |
| 108 | |
| 109 | |
| 110 | The goal of the Workshop was an evaluation of our new Efficient Annotation Framework for under-resourced languages. The objective aimed at tasks connected to the design and analysis of a new methodology for crowdsourcing annotation of previously uncovered language data as well as verification of the methodology by implementation of a new annotating framework. |
| 111 | |
| 112 | MU has prepared the tools with corpora in the 4 Ethiopian languages HaBiT project deals with plus Czech and Norwegian. These tools allowed for efficient preparation of new Part-of-Speech (PoS) taggers of these languages. The text were annotated (during the workshop) by native speakers for PoS tags, and the tool learned and improved itself during this process. |
| 113 | |
| 114 | [[Image(IMG_20170221_151251.jpg,width=30%]] |
| 115 | [[Image(IMG_20170221_143702.jpg,width=30%]] |
| 116 | [[Image(IMG_20170221_140651.jpg,width=30%]] |
| 117 | |
| 118 | [[Image(IMG_20170221_140132.jpg,width=30%]] |
| 119 | [[Image(IMG_20170221_140103.jpg,width=30%]] |
| 120 | [[Image(IMG_20170221_134458.jpg,width=30%]] |
| 121 | |