Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Proceedings |
| Untertitel | 32nd IEEE International Requirements Engineering Conference, RE 2024 |
| Herausgeber/-innen | Grischa Liebel, Irit Hadar, Paola Spoletini |
| Herausgeber (Verlag) | IEEE Computer Society |
| Seiten | 229-239 |
| Seitenumfang | 11 |
| ISBN (elektronisch) | 9798350395112 |
| ISBN (Print) | 979-8-3503-9512-9 |
| Publikationsstatus | Veröffentlicht - 2024 |
| Veranstaltung | 32nd IEEE International Requirements Engineering Conference, RE 2024 - Reykjavik, Island Dauer: 24 Juni 2024 → 28 Juni 2024 Konferenznummer: 32 |
Publikationsreihe
| Name | Proceedings of the IEEE International Conference on Requirements Engineering |
|---|---|
| ISSN (Print) | 1090-705X |
| ISSN (elektronisch) | 2332-6441 |
Abstract
With the rise of artificial intelligence in industry, many companies rely on machine learning methods such as time series forecasting. By processing data from the past, such systems can provide predictions for data in the future. In practice, however, there is often skepticism about the quality of the forecasts. Explainability has been identified as a means to address this skepticism and foster trust. While there are already different methods to explain time series forecasts, it is unclear which of these explanations are actually useful for stakeholders. To investigate the need for explanations for time series forecasts, we conducted a study at a mid-sized German company in the energy domain. Throughout the study, 23 participants were shown five examples of different explanation types. For each type of explanation, we tested if it actually helped our participants to better understand the forecasts. We found that visual explanations including decision trees and feature importance charts were able to improve domain experts' understanding of time series forecasts. Textual explanations tended to lead to confusion rather than empowerment. While the exact findings and preferable types of explanations may vary between companies, our concrete results can provide a starting point for in-depth analyses in other environments.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Ingenieurwesen (insg.)
- Betriebswirtschaft, Management und Rechnungswesen (insg.)
- Strategie und Management
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Proceedings : 32nd IEEE International Requirements Engineering Conference, RE 2024. Hrsg. / Grischa Liebel; Irit Hadar; Paola Spoletini. IEEE Computer Society, 2024. S. 229-239 (Proceedings of the IEEE International Conference on Requirements Engineering).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Explainability Requirements for Time Series Forecasts
T2 - 32nd IEEE International Requirements Engineering Conference, RE 2024
AU - Droste, Jakob
AU - Fuchs, Ronja
AU - Deters, Hannah
AU - Klunder, Jil
AU - Schneider, Kurt
N1 - Conference code: 32
PY - 2024
Y1 - 2024
N2 - With the rise of artificial intelligence in industry, many companies rely on machine learning methods such as time series forecasting. By processing data from the past, such systems can provide predictions for data in the future. In practice, however, there is often skepticism about the quality of the forecasts. Explainability has been identified as a means to address this skepticism and foster trust. While there are already different methods to explain time series forecasts, it is unclear which of these explanations are actually useful for stakeholders. To investigate the need for explanations for time series forecasts, we conducted a study at a mid-sized German company in the energy domain. Throughout the study, 23 participants were shown five examples of different explanation types. For each type of explanation, we tested if it actually helped our participants to better understand the forecasts. We found that visual explanations including decision trees and feature importance charts were able to improve domain experts' understanding of time series forecasts. Textual explanations tended to lead to confusion rather than empowerment. While the exact findings and preferable types of explanations may vary between companies, our concrete results can provide a starting point for in-depth analyses in other environments.
AB - With the rise of artificial intelligence in industry, many companies rely on machine learning methods such as time series forecasting. By processing data from the past, such systems can provide predictions for data in the future. In practice, however, there is often skepticism about the quality of the forecasts. Explainability has been identified as a means to address this skepticism and foster trust. While there are already different methods to explain time series forecasts, it is unclear which of these explanations are actually useful for stakeholders. To investigate the need for explanations for time series forecasts, we conducted a study at a mid-sized German company in the energy domain. Throughout the study, 23 participants were shown five examples of different explanation types. For each type of explanation, we tested if it actually helped our participants to better understand the forecasts. We found that visual explanations including decision trees and feature importance charts were able to improve domain experts' understanding of time series forecasts. Textual explanations tended to lead to confusion rather than empowerment. While the exact findings and preferable types of explanations may vary between companies, our concrete results can provide a starting point for in-depth analyses in other environments.
KW - Empirical Research
KW - Explainability
KW - Re-quirements Elicitation
KW - Time Series Forecasting
UR - http://www.scopus.com/inward/record.url?scp=85202751217&partnerID=8YFLogxK
U2 - 10.1109/RE59067.2024.00030
DO - 10.1109/RE59067.2024.00030
M3 - Conference contribution
AN - SCOPUS:85202751217
SN - 979-8-3503-9512-9
T3 - Proceedings of the IEEE International Conference on Requirements Engineering
SP - 229
EP - 239
BT - Proceedings
A2 - Liebel, Grischa
A2 - Hadar, Irit
A2 - Spoletini, Paola
PB - IEEE Computer Society
Y2 - 24 June 2024 through 28 June 2024
ER -