Knowledge Extraction from Natural Language Requirements into a Semantic Relation Graph

Abstract

Knowledge extraction and representation aims to identify information and to transform it into a machine-readable format. Knowledge representations support Information Retrieval tasks such as searching for single statements, documents, or metadata. Requirements specifications of complex systems such as automotive software systems are usually divided into different subsystem specifications. Nevertheless, there are semantic relations between individual documents of the separated subsystems, which have to be considered in further processes (e.g. dependencies). If requirements engineers or other developers are not aware of these relations, this can lead to inconsistencies or malfunctions of the overall system. Therefore, there is a strong need for tool support in order to detects semantic relations in a set of large natural language requirements specifications. In this work we present a knowledge extraction approach based on an explicit knowledge representation of the content of natural language requirements as a semantic relation graph. Our approach is fully automated and includes an NLP pipeline to transform unrestricted natural language requirements into a graph. We split the natural language into different parts and relate them to each other based on their semantic relation. In addition to semantic relations, other relationships can also be included in the graph. We envision to use a semantic search algorithm like spreading activation to allow users to search different semantic relations in the graph.

Artificial intelligence-based public healthcare systems: G2G knowledge-based exchange to enhance the decision-making process

“…This could be due to the logic operators or binary elements that are usually used to present the AI-based COVID-19 task solution and which are designed for knowledge-based mechanisms. Another explanation for the impact of Int.AI.PDR on AI.D is the simplicity in designing and implementing the logic in an automated system (Zhu et al, 2014); moreover, knowledge representation using natural language is easy to understand and develop (Schlutter & Vogelsang, 2018). Regarding the impact of G2G.KE on AI.D, problem representation presents how the solver cognitively engaged in processing information relating to the target problem when using the AI-based COVID-19 task solution.…”

“…Since its central and core parts consist of interrelated statements that are not entirely identical but have a similar representation of the natural language, this mechanism has a significant advantage (Zhu et al, 2014). Knowledge representation using natural language is easy to understand and develop (Schlutter & Vogelsang, 2018). The AI-based COVID-19 task solution can be used in an automated decision system in which a general solution to a problem is fed using human logic (Harris & Davenport, 2005).…”

Automatic Transformation of Natural to Unified Modeling Language: A Systematic Review

“…Besides, we also found the use of OpenNLP and SharpNLP. Stanford's POS tagger, Brill, TreeTagger, dependency parser, [11], [17], [35] use case [37], [57] test case [35], [57] processed srs [58], [59] activity diagram [7], [60] meta model [57], [61] Other Outputs: traceability [44]; graph [62]; sbvr [63]; er-diagram [64]; processed named entity [65]; b-spec [66]; owl class [67]; proposal [68]; collaboration diagram [13]; test cases [57]; natural language [69]; feature diagram [70]; gui prototype [71]…”

“…Studies nlp [5], [8], [11], [12], [14], [15], [17]- [19], [25]- [31], [33], [36], [40], [42]- [47], [49], [51]- [53], [56]- [58], [60], [62], [63], [65], [68], [71] rule [1], [15], [17], [21], [23], [24], [29]- [31], [33], [44], [49] pos tagger [1], [5], [6], [15], [19], [25]- [28], [40]- [42], [53], [64] parse [1], [13], [21], [28], [41],…”

Improving Trace Link Recovery Using Semantic Relation Graphs and Spreading Activation

“…The graph structure resembled a tree structure with multiple roots (identifier vertices) and vertices arranged in levels for verbs, arguments, and noun phrases. This structure supported that short phrases have a greater distance (i.e., are less relevant) to a certain specification than complex phrases or whole (verb) statements [22]. We used the given structure of semantic role labeling (SRL), which associates arguments with a semantic role to their predicate within a sentence.…”