We give a thorough exposition of our natural language database interface that produces result sets ranked according to the degree to which database records fulfill our intuitions about vague expressions in natural language such as `a small rainy city near San Francisco'. We present our toolset supporting the rapid prototyping of such database interfaces by means of a meta-level description of a data model and a lexicon for a query language. This talk is based on my M.Phil. thesis about Fuzzy Semantics which I submitted to the University of Cambridge Computer Lab. The thesis introduces a model of fuzzy semantics and provides some empirical evidence in support of this model. We also applied this model to produce a working natural language interface to a database which produces a ranking of all records in a database that match our intuition of, for instance, a `small rainy city near San Francisco'. In the proposed talk we will demonstrate the general background of this work, and will go into detail about one particular aspect which might be of broader interest to a hacker community: The linguistic data modelling language that we used to provide a meta-level description of a data model, together with a description of how linguistic expressions are to be interpreted with respect to the defined data model. The attachment of this submission is my thesis plus an "APPENDIX A". This APPENDIX A will constitute the focus of attention for my proposed talk.
Secdocs is a project aimed to index high-quality IT security and hacking documents. These are fetched from multiple data sources: events, conferences and generally from interwebs.
Serving 8166 documents and 531.0 GB of hacking knowledge, indexed from 2419 authors from 163 security conferences.