CoChat
Sign Up
Cited by 28Electric Vehicles and InfrastructureSmart Grid Energy ManagementEnergy Load and Power Forecasting

Seasonality Effect Analysis and Recognition of Charging Behaviors of Electric Vehicles: A Data Science Approach

Open access
Dominguez-Jimenez, Juan A., Campillo, Javier E., Montoya, Oscar Danilo, Delahoz, Enrique, Hernández, Jesus C.·2020·Sustainability
DOI 10.3390/su12187769
Open this paper in CoChat — free accountOpen PDF

Abstract

Electric vehicles (EVs) presence in the power grid can bring about pivotal concerns regarding their energy requirements. EVs charging behaviors can be affected by several aspects including socio-economics, psychological, seasonal among others. This work proposes a case study to analyze seasonal effects on charging patterns, using a public real-world based dataset that contains information from the aggregated load of the total charging stations of Boulder, Colorado. Our approach targets to forecast and recognize EVs demand considering seasonal factors. Principal component analysis (PCA) was used to provide a visual representation of the variables and their contribution and the correlation among them. Then, twelve classification models were trained and tested to discriminate among seasons the charging load of electric vehicles. Later, a benchmark stage is presented for regression as well as for classification results. For regression models, examined through Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE), the random Forest provides better prediction than quasi-Poisson model widely. However, it was observed that for large variations in electric vehicles’ charging load, quasi-Poisson fits better than random forest. For the classification models, evaluated through Accuracy and the Area under the Curve, the Lasso and elastic-net regularized generalized linear (GLMNET) model provided the best global performance with accuracy up to 100% when evaluated on the test dataset. The results of this work offer great insights for enhancing demand response strategies that involve PEV charging regarding charging habits across seasons.

AI snapshot

Written from the paper’s full text.

TL;DR

This paper analyzes seasonal effects on electric vehicle charging behavior using aggregated public charging-station data from Boulder, Colorado, combining PCA, feature importance, statistical tests, classification, and regression. It finds that seasonality significantly shapes charging load and charging time, and that machine-learning models can forecast and discriminate seasonal charging patterns with high accuracy.

Key findings

  • ANOVA indicated significant seasonal differences in energy consumption (KWH) and charging time (CTM) with p < 0.05; winter had the lowest EV load but the highest charging time, autumn reached maximum demand, and spring/summer charging habits were similar. Chi-square tests found EV load was not related to day of week (χ = 58.317, df = 48, p = 0.1462) but was related to weekday/weekend (χ = 20.704, df = 6, p = 0.007); charging time was not affected by weekday/weekend (χ = 6.3759, df = 5, p > 0.05).
  • Among 12 classifiers, XGBOOST had the highest mean global accuracy (99.71%) and mean AUC (99.96%) in Table 2; for season-pair discrimination, GLMNET performed best, with test mean accuracy 98.78% (95% CI 93.39–99.97%), training mean accuracy 98.24%, balanced accuracy near 98.56%, and a No Information Rate of 37.8%.
  • For regression, random forest generally outperformed quasi-Poisson: in scenario I, RF had RMSE = 11.65 and MAPE = 0.08% versus QPM RMSE = 25.66 and MAPE = 1.3822%; in scenario II, RF had RMSE = 8.82 and MAPE = 0.28% versus QPM RMSE = 1.80 and MAPE = 0.44%; in scenario III, RF had RMSE = 6.68 and MAPE = 0.6140% versus QPM RMSE = 12.35 and MAPE = 1.0534%. QPM fit better during large variations in charging load.
  • The conclusion states RF reached MAPE up to 0.08% and an RMSE of 2.27, and that GLMNET was the best classification method according to mean accuracy and ROC.

Methods

  • Case study using a public real-world aggregated dataset from 44 city-operated public charging stations in Boulder, Colorado, from 1 January 2018 to 28 February 2019; the authors added mean temperature and temporal variables including day, weekday, weekend, and season.
  • Analysis was performed in RStudio (Version 1.1.442); methods included PCA for visualization (first two principal components represented 72.2% of the information), Boruta feature importance, ANOVA, and Chi-square tests.
  • Twelve classification models were trained and tested with 10-fold cross-validation and an 80/20 train/test split; two regression models (quasi-Poisson and random forest) were evaluated across three seasonal forecasting scenarios using accuracy/AUC and MAPE/RMSE.

Limitations

  • The paper states the main limitation is that the dataset provides only aggregated information, so the approach cannot support short-term decisions such as when to connect or disconnect a specific user to the grid; it is instead suited to medium- and long-term planning. Future work is proposed with non-aggregated data and socio-demographic variables.
  • The paper notes that the resulting dataset is not balanced during training and testing, and that no feature selection techniques were used.
  • Evident from the text: the narrative is inconsistent about the best classifier—Table 2 shows XGBOOST with the highest mean global accuracy/AUC, while the abstract and conclusion emphasize GLMNET as best—and the conclusion’s RF RMSE of 2.27 does not match the scenario-specific RF RMSE values reported in the results.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/su12187769

Cites

44 of 44 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

8 of 20 cited sources checked · 4 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    EV charging loads are sensitive to seasons.

    “EV charging loads are sensitive to seasons”
    Quote in the paperSubstantial claim3 of 3 sources checked
    • cites [15] Zhao, Y., Che, Y., Wang, D., Liu, H., Shi, K., and Yu, D. (2018). An optimal domestic electric vehicle charging strategy for reducing network transmission loss while taking seasonal factors into consideration. Appl. Sci., 8. source
    • cites [16] Boston, 2016, Plug-in Vehicle Behaviors: An analysis of charging and driving behavior of Ford plug-in electric vehicles in the real world, World Electr. Veh. J. source

      Section 3.1 (Charge Events per Week; page unavailable) reports that PHEVs charge slightly less in winter and BEVs slightly more, with a California exception. This supports seasonal variation in charging frequency, but the paper does not establish seasonal changes in electrical charging loads. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites [17] Ul-Haq, A., Azhar, M., Mahmoud, Y., Perwaiz, A., and Al-Ammar, E.A. (2017). Probabilistic modeling of electric vehicle charging pattern associated with residential load for voltage unbalance assessment. Energies, 10. source
  • Premisecitation

    Cold temperatures cause drastic changes in human habits, especially during sub-zero temperatures, as in Canada, Sweden, Norway and Finland, where people prefer to stay indoors.

    “This behavior is explained from several domains, for example, cold temperatures cause drastic changes in human being's habits, especially during sub-zero temperatures as the Canada, Sweden, Norway and Finland cases [51] were people prefer to stay indoors.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [51] Sovacool, 2018, Reviewing Nordic transport challenges and climate policy priorities: Expert perceptions of decarbonisation in Denmark, Finland, Iceland, Norway, Sweden, Energy source

      Section 3.2 says winter weather can make travel in Finland take longer, and section 3.5 discusses the need for 4x4 vehicles in Icelandic snow. Neither passage addresses sub-zero temperatures causing people in Canada, Sweden, Norway, or Finland to prefer staying indoors. Searches for staying indoors and cold-weather changes in habits found no passage making that point. (LLM, conf=0.75)

      Experimental — accuracy not yet measured
  • Findingown data

    In the second scenario, QPM (RMSE = 1.80, MAPE = 0.44%) tends to underestimate the real load demand of EVs users.

    “QPM (RMSE = 1.80,MAPE = 0.44%) tends to underestimate the real load demand of EVs users.”
    Quote in the paperNot a substantial claim
  • Findingown data

    The five models with the highest performance values are XGBOOST, TBAG, NB, DT, and BLR.

    “The five models that have the highest performance values are XGBOOST, TBAG, NB, DT, and BLR.”
    Quote in the paperNot a substantial claim
  • Findingown data

    For classification models, GLMNET provided the best global performance with accuracy up to 100% on the test dataset, evaluated through Accuracy and AUC.

    “For the classification models, evaluated through Accuracy and the Area under the Curve, the Lasso and elastic-net regularized generalized linear (GLMNET) model provided the best global performance with accuracy up to 100% when evaluated on the test”
    Quote in the paperSubstantial claim
  • May combine two claims. 18
  • Classified as probably not a substantial claim. 16

  • May combine two claims.

    Interpretationcitation

    The autumn high load is strongly related to ambient temperature beginning to decrease, reducing available battery range, and owners charging more often to ensure sufficient range.

    “This fact is strongly related to the fact as the ambient temperature begins to decrease, the available battery range gets reduced, and owners may be charging their vehicles more often to ensure they have sufficient range to reach their final destination [16].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [16] Boston, 2016, Plug-in Vehicle Behaviors: An analysis of charging and driving behavior of Ford plug-in electric vehicles in the real world, World Electr. Veh. J. source

      The “Summer Winter” section says cold temperatures reduce available BEV range and suggests owners may charge more often to ensure they can reach their destination. It discusses winter charging, not an autumn high load, and does not establish a strong relationship between that load and falling temperatures. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    EV charging behavior depends on several factors such as socio-demographics, time of the day, location of CS, distance travelled, tax incentives.

    “EVs charging behavior depends on several factors such as socio-demographics, time of the day, location of CS, distance travelled, tax incentives [46].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [46] Vassileva, 2017, Adoption barriers for electric vehicles: Experiences from early adopters in Sweden, Energy source

      Section 3.3, “Driving and charging patterns” (page unavailable), reports when and where surveyed EV owners charge and how far they drive. The paper discusses socio-demographics and tax incentives in relation to EV ownership and adoption, but does not establish them as factors determining charging behavior. (LLM, conf=0.86)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim.

    Premisecitation

    Ul-Haq et al.'s results state that charging habits change across seasons.

    “Their results state that charging habits changes across seasons.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [17] Ul-Haq, A., Azhar, M., Mahmoud, Y., Perwaiz, A., and Al-Ammar, E.A. (2017). Probabilistic modeling of electric vehicle charging pattern associated with residential load for voltage unbalance assessment. Energies, 10. source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Boston and Werthman's findings show that electricity prices, local charging infrastructure, and seasonal weather changes may affect driving behavior and then charging habits.

    “Their findings shows that electricity prices, local charging infrastructure, and seasonal weather changes may affect driving behavior and then charging habits [16].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [16] Boston, 2016, Plug-in Vehicle Behaviors: An analysis of charging and driving behavior of Ford plug-in electric vehicles in the real world, World Electr. Veh. J. source

  • May combine two claims.

    Findingown data

    Only the first two principal components were selected because they represent 72.2% of the information.

    “Only the two first principal components (PCs) were selected as they represent the 72.2% of the information.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Most predictors have a significant importance, as only the variable DAY contributed less than the shadow features.

    “Most of the predictors have a significant important, as only the variable DAY contributed less than the shadow features.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In fall, the charging time is more moderate than winter but higher than Spring and Summer.

    “the charging time is more moderate than winter but higher than Spring and Summer.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    The number of charging stations increased by twice and a bit more to date.

    “however, to date, the number of stations increased by twice and a bit more.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    GLMNET may fail to recognize Summer when the season pair is SU-WIN.

    “this could fall in recognizing Summer when the pair is SU-WIN.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    During autumn, the CTM increases but not as much as in winter.

    “During autumn, the CTM increases but not as much as in winter”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The variable which provides less information is CTM followed by ACGHC.

    “Thus, the variable which provides less information is CTM followed by ACGHC.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingown data

    EVs load demand is rapidly growing.

    “From the above listed information, EVs load demand is rapidly growing and Figure 2 illustrates its behavior on the last year.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    QPM often tends to underestimate and overestimate EVs load.

    “QPM often tends to underestimate and overestimate EVs load”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    More charging sessions are associated with more energy consumption and number of ports used.

    “This behavior keeps for the rest of the charging sessions groups, more charging sessions, more energy consumption, number of ports used.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    Our findings are similar to time series based models, including CART and ARIMA.

    “Our findings are similar to time series based models, including CART and ARIMA.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Dudek

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    The resulting dataset is not balanced during training and test.

    “The resulting dataset is not balanced during training and test.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    EVs' energy consumption can be forecasted and discriminated from seasons using machine learning tools with high accuracy.

    “This paper shows that EVs' energy consumption can be forecasted and discriminated from seasons using machine learning tools with high accuracy.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The Boulder, Colorado EV charging dataset is representative of EV charging behavior more generally.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The classifiers were trained and tested over the test data with a 10-folds cross-validation algorithm.

    “These classifiers were trained and tested over the test data with a 10-folds cross-validation algorithm.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationown data

    Our results are comparable with several studies.

    “Several studies have been carried out in this topic, and our results are comparable with them.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The models were evaluated using a repeated 10-fold cross-validation algorithm.

    “The models were evaluated using a repeated 10-fold cross-validation algorithm.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Both RF and QPM are models able to forecast EVs load.

    “From these results, it is clear that both RF and QPM are models able to forecast EVs load.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The opposite growing directions of CTM and TEMP may suggest that they have an inverse relationship.

    “The opposite growing directions of CTM and TEMP may suggest that they have an inverse relationship.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    This work provided results with high accuracy, which allows medium and long-term planning of power systems.

    “Nevertheless, this work provided results with high accuracy, which allows medium and long-term planning of power systems.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The Boulder, Colorado EV charging dataset is representative of EV charging behavior more generally.

  • May combine two claims.

    Interpretationcitation

    The lowest MAPE value reported by Dudek was 0.92, which is higher than our obtained 0.08.

    “The lowest MAPE value that the author reported was 0.92, which is higher in contrast with the obtained by us (0.08).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Dudek

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation

    Random forest is an additive method based on several decision trees which has been widely proved to be a powerful approach to forecast electrical load.

    “RF is an additive method based on several decision trees which has been widely proved to be a powerful approach to forecast electrical load [40].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [40] Dudek, G. (2015). Short-term load forecasting using random forests. Intelligent Systems’ 2014, Springer. source

  • May combine two claims.

    Interpretationown data

    We forecasted EVs load for a more extended period than Lu et al. and obtained better MAPE in the worst-case 1.3822% and comparable RMSE.

    “we forecasted EVs load for a more extended period, and we obtained better results regarding MAPE in the worst-case 1.3822%, and from RMSE we obtained comparable results.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lu et al.

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data

    RF provided better global performance reaching MAPE up to 0.08%, and an RMSE of 2.27.

    “However, RF provided better global performance reaching MAPE up to 0.08%, and an RMSE of 2.27.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The similarity of Spring and Summer charging habits may be explained by people tending to stay outside the house to take advantage of agreeable weather and longer days.

    “This may be explained by during these seasons, people tend to stay outside the house to take advantage of agreeable weather and longer days (more sunlight hours).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Since the QPM dispersion parameter is higher than one (presence of overdispersion), QPM can be used under such consideration.

    “Since this parameter is higher than one (presence of overdispersion), QPM can be used under such consideration.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The dataset was divided into 80% for model training and 20% for testing with unseen data.

    “we divided it into 80% for model training and 20% for testing the model with unseen data.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationreasoning

    The winter behavior may aid to increase battery's lifetime, as users do not use their vehicles in the same way as during other seasons, where the ambient temperature is the warmest.

    “This behavior may aid to increase battery's lifetime, as users do not use their vehicles in the same way as during other seasons, where the ambient temperature is the warmest.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The used dataset provides only aggregated information, so the proposed approach is not able to take short-term decisions such as when to connect or disconnect a specific user to the grid.

    “The main limitation of this work relies on the fact that the used dataset provides only aggregated information and thus the proposed approach is not able to take short-term decisions such as when to connect or disconnect a specific user to the grid.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Two regression models and 12 classification models were tuned with 10-fold cross-validation to find optimal parameters to minimize performance metrics.

    “Two regression models, including QPM and RF and several classification models (12) were tuned with the 10-fold cross-validation algorithm to find the optimal parameters to minimize performance metrics.”
    Quote in the paperNot a substantial claim
33 of 33

  • The paper relies on this without support.

    Seasonal factors have not been widely taken into account in recent studies.

    “seasonal factors have not been widely taken into account in recent studies”

  • The paper relies on this without support.

    Many factors still have not been taken into account when modelling users' load charging habits.

    “However, even though the high performance and accuracy of these strategies, many factors still have not been taken into account when modelling users' load charging habits.”

  • The paper relies on this without support.

    Most studies in this regard only consider information on the vehicle or charging station side.

    “However, most studies in this regard only considers information on the vehicle or charging station side.”

  • The paper relies on this without support.

    The charging behavior observed in Boulder is representative enough of PEV charging habits across seasons to inform demand response strategies and large-scale planning.

    “The results of this work offer great insights for enhancing demand response strategies that involve PEV charging regarding charging habits across seasons.”

  • The paper relies on this without support.

    The Boulder, Colorado EV charging dataset is representative of EV charging behavior more generally.

    “the demanda data of all the seasons of Boulder, Colorado were selected as a numerical example for simulation verification.”
5 of 5

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • GLMNET is called best by mean accuracy and AUC despite lower table values

    Major

    The conclusion names GLMNET the best according to mean accuracy and ROC, but Table 2 gives it a mean accuracy of 0.9822 and mean AUC of 0.9997, below XGBTREE and TBAG at 0.9969 and 1, respectively.

  • Conclusion attributes an unreported RMSE of 2.27 to RF

    Minor

    The conclusion pairs RF's reported 0.08% MAPE with an RMSE of 2.27. In the corresponding scenario, RF's reported RMSE is 11.65; 2.27 is instead the RMSE cited for Lu et al.'s study.

2 of 2

Cites

Flagged first, then by citations

1 of the 45 references couldn’t be matched to a known paper.

Research works printed but not linked: 3

  1. Wang, 2015, Big data analysis and parallel load forecasting of electric power user side, Proc. CSEE, 35, 527
  2. Wu, 2015, Power system short-term load forecasting based on improved random forest with grey relation projection, Autom. Electr. Power Syst., 39, 50
  3. Meyer, N., Whittal, I., Christenson, M., and Loiselle-Lapointe, A. (2012, January 6–9). The impact of driving cycle and climate on electrical consumption and range of fully electric passenger vehicles. Proceedings of the EVS26 International Battery, Hybrid, and Fuel Cell Electric Vehicle Symposium, Los Angeles, CA, USA.

Cited by

Sorted by citations

Related

Authors
Dominguez-Jimenez, J. A., Campillo, J. E., Montoya, O. D., Delahoz, E., & Hernández, J. C.
Affiliations
  • Hydrogen Research Institute, Université du Quebec à Trois-Rivieres, Quebec, QC 3351, Canada
  • Universidad Tecnológica de Bolívar (CO)
  • Laboratorio Inteligente de Energía, Universidad Tecnológica de Bolívar, Cartagena 131001, Colombia
  • Facultad de Ingeniería, Universidad Distrital Francisco José de Caldas, Bogotá D.C. 11021, Colombia
ORCID
Venue
Sustainability
Publisher
MDPI AG
Published
2020-09-20
Volume
12
Issue
18
Pages
7769
ISSN
2071-1050 (ISSN-L)
DOI
10.3390/su12187769
OpenAlex
W3087514398
Cited by
28
References
45
Citations by year
2026: 6 · 2025: 6 · 2024: 4 · 2023: 6 · 2022: 3 · 2021: 3
FWCI
1.01
Type
journal article
Language
en
Open access
Licence
Locations
  • Sustainability (publishedVersion, cc-by)
  • Repositorio UTB (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Greater South Information System
Funders
Departamento Administrativo de Ciencia, Tecnología e Innovación (COLCIENCIAS) (727, 2015.)
Topics
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Electric Vehicles and Infrastructure
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Smart Grid Energy Management
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Energy Load and Power Forecasting
Keywords
EV charging behavior, seasonal effects, charging stations, demand response, quasi-Poisson regression, random forest, glmnet, Boulder, Colorado

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-28.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-28.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review