Extracting Knowledge Tokens from Text Streams
Eugene Alferov, Vadim Ermolayev · 2013
This problem analysis paper presents our position on how could the solution be sought to the problem of extracting semantically rich fragments from a stream of plain text posts. We first present our understanding of the problem context and explain the focus of our research. Further, in the problem setting section we elaborate the workflow for knowledge extraction from in- coming information tokens. This workflow is then used as a key to structure our review of the literature on the relevant component techniques which may be ex- ploited in a combination to achieve the desired outcome. We finally outline our plan for conducting the experiments with an aim to validate the workflow and find a proper combination of the component techniques for all steps which may solve our specific research problem.