introduction: the aim of this investigation was to study the developmental potential of human fragmented embryos to blastocyst stage and to evaluate thedeveloped blastocysts in the term of size, cellularity, the type of inner cell mass(icm), trophoetoderm (te) and the number of dead cells. material and methods: four to eight cell fragmented human embryos wereobtained from the embryology lab of ...

The following material presents a study about recovering punctuation marks, and capitalization information from European Portuguese broadcast news speech transcriptions. Different approaches were tested for capitalization, both generative and discriminative, using: finite state transducers automatically built from language models; and maximum entropy models. Several resources were used, includi...

We investigate the capitalization features of queries submitted to Web search engines and the relation between capitalization information, either as received from users or as hypothesized based on Web statistics, and search relevance. We observe that users tend to lowercase words in their queries significantly more often than as predicted from Web data. More importantly, we determine that docum...

We present a probabilistic bilingual capitalization model for capitalizing machine translation outputs using conditional random fields. Experiments carried out on three language pairs and a variety of experiment conditions show that our model significantly outperforms a strong monolingual capitalization model baseline, especially when working with small datasets and/or European language pairs.

The rapid proliferation of microblogs such as Twitter has resulted in a vast quantity of written text becoming available that contains interesting information for NLP tasks. However, the noise level in tweets is so high that standard NLP tools perform poorly. In this paper, we present a statistical truecaser for tweets using a 3-gram language model built with truecased newswire texts and tweets...