Determining abbreviations in Kariyer.net domain

In this paper, studies determining abbreviations and their meanings in job texts are explained. The data used in this study consist of job texts stored in the Kariyer.net database. The applied method consists of two separate steps: first, the words and phrases in all job text documents are vectorised with the Word2Vec model. The phrases and abbreviations that are compatible with each other in the proximity of these word vectors are then checked and matched. In the second step, sentences with abbreviations and their meanings in the dataset are defined by the rules determined by Regex. Then, the appropriate abbreviations are collected and added to the dictionary. Keywords: Word embeddings, text mining, abbreviation detection.