Four-Step Research Topic Condensation

Problem

A folder of papers can easily become a broad related-work summary, but the harder task is extracting falsifiable scientific questions, defensible research gaps, and actionable topic directions. This skill treats a literature directory as the input and produces gap analysis, blank-area mapping, candidate scientific questions, interdisciplinary topics, and an interactive knowledge graph for inspecting concept relationships.

How It Works

• Common gap extraction: after full-text extraction, it counts limitation keywords and extracts critical context to surface lower-level methodological weaknesses rather than generic claims like more research is needed.
• Multi-dimensional blank mapping: it organizes gaps across knowledge, relationships, theory, and method, then supplements the evidence with future-work statements to produce comparable entry points.
• Scientific question generation: it clusters themes through an inductive path, builds a domain ontology and inference structure through a deductive path, and applies interdisciplinary lenses to derive meta-questions, layered questions, and testable hypotheses.
• Evaluation and iteration: it scores candidates on originality, interdisciplinary fit, application potential, and theoretical depth, then supports expert-debate style iteration with prioritized candidates and a scoring report.

Boundaries

• It expects a reasonably complete literature directory; each gap, blank, or hypothesis should be tied to source evidence rather than fabricated reasoning.
• Knowledge graph entities must come from the analyzed literature. External theories may be used only as imported lenses in interdisciplinary reinforcement, not as native domain constructs.
• Stage boundaries are explicit, especially before evaluation and iteration: the user must select the candidate scope instead of silently scoring everything.
• It is best suited to research topic condensation, gap analysis, and concept visualization; it does not replace experimental design, statistical validation, or open-ended ideation without literature grounding.