INAGQA: Cross-Domain Performance Drop for German Financial Question Answering

INAGQA is a novel QA system achieving 0.91 F1 score on German financial queries, validated on 2,100 expert-annotated questions. The dataset likely contains performance metrics and experimental results comparing INAGQA to systems like BERT-KGQA (F1: 0.83) and template-based systems (F1: 0.79). Authored by Jamal Al Qundus and last updated in May 2026, this 5.5 KB Excel file contributes to information systems research with language-sensitive design principles.

Use Cases

• Benchmarking German-language QA system performance based on reported F1 scores and accuracy metrics.
• Studying headword-centric parsing and hybrid disambiguation algorithms for handling linguistic variability.
• Analyzing the impact of domain-optimized shallow parsing on reducing relation-linking errors for compound nouns.
• Evaluating knowledge integration techniques for user-curated data in semantic QA systems.

Strengths

• Performance validated on 2,100 expert-annotated German financial questions.
• Reports specific metrics: 0.91 F1 score, 35% reduction in relation-linking errors, 98% accuracy on temporal/quantitative variants.
• System demonstrates practical application with a 2.1s average response time in a financial analyst case study.

Limitations

• Dataset is very small at 5.5 KB, indicating limited scope.
• Row count and column-level documentation are unknown; field semantics must be inferred after download.
• Description metadata is limited; actual data quality and structure require manual inspection.

Provenance

Source

figshare, authored by Jamal Al Qundus.

Collection Method

Likely contains experimental results and performance metrics from the INAGQA QA system research.

Freshness

Last updated 2026-05-04 17:51:22; freshness should be verified.