Anaphora resolution is the process of determining the antecedent of an anaphor (AR). AR is a challenging and complex issue in the field of Natural Language Processing (NLP). Very few works have been done for anaphora resolution in the Afaan Oromo language and this is due to lack of resources. Most researchers have applied a rule-based approach for anaphora resolution in Afaan Oromo. This paper proposed the pronominal anaphora resolution model for the Afaan Oromo language developed using a conditional random field (CRF), which deals with both Intra-sentential and inter-sentential kind of anaphora. CRF++ 0.58 tool has been used to develop the model. Afaan Oromo texts are collected from different sources such as Afaan Oromo holly bible, BBC Afaan Oromo news, Afaan Oromo grade 9 and 11 student textbooks to evaluate the performance of this model. Totally 1330 sentences with 12571 tokens were collected for both independent and hidden anaphors. From this collected and prepared dataset, 80 % of the dataset was used for training and the remaining 20% is for testing data. The precision of 78.87 percent, recall of 91.80 percent, and F-measure of 84.85 percent for resolution of independent anaphors were obtained using the CRF approach for Afaan Oromo pronominal anaphora resolution. The precision of 80.41 percent, recall of 95.12 percent, and F-measure of 87.15 percent were obtained for the resolution of hidden anaphors.

Natural Language Processing, Anaphora resolution, Afaan Oromo, Machine Learning, Conditional Random Fields.