Rule-based coreference resolution module for Polish ⋆ Maciej Ogrodniczuk and Mateusz Kopeć Institute of Computer Science, Polish Academy of Sciences Abstract. This paper presents the results of the first attempt of co- reference resolution for Polish running on true mention boundaries and using a few rich rules, corresponding to syntactic constraints (elimina- tion of nested nominal groups), syntactic filters (elimination of syntac- tic incompatible heads), semantic filters (wordnet-derived compatibility) and selection (weighted scoring). The results are compared to human annotation and presented in four sets: with two common baselines: all- singletons/head-match, and two slightly more complex settings with four and five rules. Keywords: coreference resolution of Polish, Polish anaphora resolution 1 Introduction Although few anaphora resolution attempts were already made for Polish (see e.g. [7], [8], [9]) they were either purely theoretical or pronoun-limited. This paper presents the first coreference resolution module for Polish, intended to provide starting ground for further experiments and generate the reference baseline to be compared with future more advanced rule-based and statistical coreference resolvers. The scope of the resolution is limited to identity-of-reference direct nominal coreference. The module design follows Haghighi and Klein’s approach [4] by building on the richness of important characteristics rather than multitude of weak features. For languages such as Polish which still lacks advanced discourse processing tools, this approach seems very promising also because of practical reasons. Additional intention of this attempt is gathering experience for the next phases of recently started project Computer-based methods for coreference res- olution in Polish texts which tasks also include creation of the corpus of Polish manually annotated with various types of coreference. 2 System Description The implemented module uses standard best-first entity-based model based on syntactic constraints (elimination of nested nominal groups), syntactic filters ⋆ The work reported here was carried out within the Computer-based methods for coreference resolution in Polish texts project financed by the Polish National Science Centre (contract number 6505/B/T02/2011/40).