Automatic extraction of glossary terms from natural language requirements

Abstract

We present a method for the automatic extraction of glossary terms from unconstrained natural language requirements. The glossary terms are identified in two steps -a) compute units (which are candidates for glossary terms) b) disambiguate between the mutually exclusive units to identify terms. We introduce novel linguistic techniques to identify process nouns, abstract nouns and auxiliary verbs. The identification of units also handles co-ordinating conjunctions and adjectival modifiers. This requires solving co-ordination ambiguity and adjectival modifier ambiguity. The identification of terms among the units adapts an in-document statistical metric. We present an evaluation of our method over a real-life set of software requirements' documents and compare our results with that of a base algorithm. The intricate linguistic classification and the tackling of ambiguity result in superior performance of our approach over the base algorithm.

Requirements Engineering: Best Practice

“…SA Diagrams Specifying functions, processes, structure, and behaviour with one of the graphical notations proposed by structured analysis to achieve precision and make structure visible. Tables Specifying concepts to achieve an understanding of the terminology (Dwarakanath, Ramnani et al 2013) and or rules for how conditions affect system behaviour (Bajec and Krisper 2005). UML Diagrams Specifying functions, scenarios, processes, rules, relations, behaviour, and deployment with graphical notations from the Unified Modelling Language to increase precision and show structure.…”

Extraction of System States from Natural Language Requirements

“…The authors in [30] use a NER-like algorithm to automate glossary term extraction for requirements documents.…”

Automatic Glossary Term Extraction from Large-Scale Requirements Specifications

“…The terms extracted by their approach achieved an average relevance rating of 0.47 on a scale from 0 (bad) to 2 (good). Other examples of hybrid approaches are proposed by Park et al [15] and Dwarakanath et al [4]. Note, however, that all mentioned approaches use a linguistic (rule-based) approach to chunking rather than a statistical chunker.…”

“…Our paper is the first to report on automatic glossary term extraction on large-scale requirements specifications, represented by the CrowdRE dataset, which is 5 to 10 times larger than datasets that have been used for evaluating related approaches [2] [4]. We investigate the effects of single parts of our approach (linguistic extraction and statistical filtering) to the dataset and argue that for large-scale requirements specifications, statistical filters are inevitable although they may reduce recall.…”