Pith. Sign in

RECORD

Enriching Artificial Intelligence Explanations with Knowledge Fragments

PaperPDF from arXiv

data files

Paper breakdownThe authors' claims, conditions and measurements, as JSON Signed receiptProof this review has not been altered
X Bluesky LinkedIn Reddit Hacker News
Jože M. Rožanec, Elena Trajkova, Inna Novalija, Patrik Zajec, Klemen Kenda, Blaž Fortuna, Dunja Mladenić

Artificial Intelligence · Submitted 12 Apr 2022

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim Furthermore, the insights into models' rationale can be enriched with domain knowledge.

how it is stated · asserted 0.84 read the record →

Comment (0) Read on arXiv ↗
Cite ▾
arXiv ID2204.05579 (version 1)
Pith IDL4IXKURM, Pith's permanent record of this paper and its review
CitationBla\v{z} Fortuna, Dunja Mladeni\'c, Elena Trajkova, Inna Novalija, Jo\v{z}e M. Ro\v{z}anec, Klemen Kenda, Patrik Zajec. "Enriching Artificial Intelligence Explanations with Knowledge Fragments." arXiv:2204.05579 (2022). https://pith.science/paper/L4IXKURM
Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 4,110 input tokens.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.06 · human participants 0.07 · controlled comparison 0.27 · randomized assignment 0.02 · states quantitative result 0.03 · compares with prior results 0.81 · mentions code or data availability 0.03.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; .

Artificial Intelligence models are increasingly used in manufacturing to inform decision-making. Responsible decision-making requires accurate forecasts and an understanding of the models' behavior. Furthermore, the insights into models' rationale can be enriched with domain knowledge. This research builds explanations considering feature rankings for a particular forecast, enriching them with media news entries, datasets' metadata, and entries from the Google Knowledge Graph. We compare two approaches (embeddings-based and semantic-based) on a real-world use case regarding demand forecasting.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#sentencerolethe authors' own workstates the main conclusionstrength
s0 Artificial Intelligence models are increasingly used in manufacturing to inform decision-making. background 1.00 0.05 0.02 asserted 1.00
s1 Responsible decision-making requires accurate forecasts and an understanding of the models' behavior. background 0.98 0.04 0.03 asserted 1.00
s2 Furthermore, the insights into models' rationale can be enriched with domain knowledge. background 0.93 0.08 0.10 asserted 0.84
s3 This research builds explanations considering feature rankings for a particular forecast, enriching them with media news entries, datasets' metadata, and entries from the Google Knowledge Graph. method 0.57 0.94 0.13 asserted 0.98
s4 We compare two approaches (embeddings-based and semantic-based) on a real-world use case regarding demand forecasting. method 0.73 0.93 0.07 asserted 0.97

What kind of paper the classifier reports this is: method or tool 0.87 · evidence from observation or experiment 0.83.

What was examined, and what was not

sources read
  • arXiv listing abstract v1
not examined
  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers
levels run
  • P abstract
not run
  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper
Original abstract

Artificial Intelligence models are increasingly used in manufacturing to inform decision-making. Responsible decision-making requires accurate forecasts and an understanding of the models' behavior. Furthermore, the insights into models' rationale can be enriched with domain knowledge. This research builds explanations considering feature rankings for a particular forecast, enriching them with media news entries, datasets' metadata, and entries from the Google Knowledge Graph. We compare two approaches (embeddings-based and semantic-based) on a real-world use case regarding demand forecasting.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

What this paper does with the arXiv papers it cites

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

What it cites · 20 sentences without a confident label

Cites as background 1 paper

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.