VaryLATEX: Learning Paper Variants That Meet Constraints

Abstract

How to submit a research paper, a technical report, a grant pro posal, or a curriculum vitae that respect imposed constraints such as formatting instructions and page limits? It is a challenging task, especially when coping with time pressure. In this work, we present VaryLATEX, a solution based on variability, constraint program ming, and machine learning techniques for documents written in LATEX to meet constraints and deliver on time. Users simply have to annotate LATEX source files with variability information, e.g., (de)activating portions of text, tuning figures’ sizes, or tweaking line spacing. Then, a fully automated procedure learns constraints among Boolean and numerical values for avoiding non-acceptable paper variants, and finally, users can further configure their papers (e.g., aesthetic considerations) or pick a (random) paper variant that meets constraints, e.g., page limits. We describe our implementation and report the results of two experiences with VaryLATEX.