Pith. sign in

REVIEW 2 major objections 1 minor 20 references

TSRuleGrowth : Extraction de r\`egles de pr\'ediction semi-ordonn\'ees \`a partir d'une s\'erie temporelle d'\'el\'ements discrets, application dans un contexte d'intelligence ambiante

T0 review · 2 major / 1 minor · reviewed 2026-05-24 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read TSRuleGrowth extracts partially-ordered prediction rules from discrete time series by introducing a new support measure.

desk verdict TSRuleGrowth claims a new support for mining partially-ordered rules from time series but the abstract gives no definition or check that it preserves the anti-monotonicity needed for correct pruning. read the letter →

arxiv 1907.10054 v1 pith:PWVEBTAZ submitted 2019-07-23 cs.AI

Benoit Vuillemin (LIRIS) , Lionel Delphin-Poulat (FTR&D) , Rozenn Nicol , Laëtitia Matignon (SMA) , Salima Hassas (MSI) This is my paper
classification cs.AI
keywords timeseriesruleminingpartiallyorderedrulesambientintelligenceconnectedobjectspredictionsequentialpatternsuserhabits
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper presents TSRuleGrowth as a new algorithm designed to discover semi-ordered rules in time series of discrete elements. It adapts established rule-mining techniques by defining a fresh notion of support that accounts for the temporal ordering in the data. The authors apply the method to real-world data collected from connected objects in an ambient intelligence environment, showing how it can surface recurring user habits as prediction rules.

What carries the argument

TSRuleGrowth algorithm, which adapts rule-mining search using a redefined support that respects temporal sequence constraints.

What would settle it

Apply TSRuleGrowth to a synthetic time series containing known planted rules and check whether the output recovers those rules at the expected support levels without spurious additions.

Watch

Extended reading notes

Core claim

TSRuleGrowth is an algorithm that mines partially-ordered rules over time series by taking principles from sequential rule mining and applying them through a newly defined support measure, with validation on connected-environment data that extracts user habits.

Load-bearing premise

A new support definition can be introduced for time series that keeps the validity of rule-mining principles intact without adding bias or invalidating the discovered rules.

Editorial extensions

If this is right

  • Partially-ordered rules can be extracted directly from streams of discrete events recorded over time.
  • User habits become expressible as prediction rules linking sequences of actions from multiple connected objects.
  • Standard rule-mining search strategies remain usable once the support calculation is adjusted for temporal data.
  • The approach yields actionable patterns for ambient intelligence systems without requiring full linear ordering of events.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same support adjustment might transfer to other ordered data types such as event logs in software or sensor streams in manufacturing.
  • If the new support avoids bias, the algorithm could be combined with existing frequent-pattern miners to handle mixed discrete-continuous series.
  • Scalability tests on larger IoT datasets would reveal whether the partial-order search remains tractable as the number of distinct elements grows.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 1 minor

Summary. The paper introduces TSRuleGrowth, an algorithm for extracting partially-ordered prediction rules from discrete-element time series. It adapts state-of-the-art rule-mining techniques (e.g., RuleGrowth) to the temporal setting through a new notion of support and applies the method to real data from connected objects in an ambient-intelligence environment in order to discover user habits.

Significance. If the novel support measure can be shown to preserve the anti-monotonicity and downward-closure properties required for correct candidate generation and pruning, the algorithm would constitute a practical extension of sequential rule mining to ordered time series and could be useful for habit extraction in IoT/ambient-intelligence applications. The use of real connected-environment data is a positive aspect, but the absence of any formal definition or validation prevents assessment of whether these properties actually hold.

major comments (2)
  1. [Abstract] Abstract: the central innovation is described only as 'a new notion of support' with no mathematical definition, no statement of how support is computed over temporally ordered discrete events, and no verification that the measure satisfies anti-monotonicity (support(r) ≤ support(r') whenever r' is a proper subsequence of r). This property is load-bearing for the correctness of any growth-based pruning procedure.
  2. [Abstract] No section or equation is supplied that defines the support measure, proves the required lattice properties, or gives the pseudocode for the adapted growth procedure; without these the claim that the algorithm 'takes principles from the state of the art' cannot be evaluated for completeness or bias.
minor comments (1)
  1. [Title] The title is written in French while the abstract is in English; a consistent language choice would improve accessibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. We address the major comments point by point below and will incorporate revisions to improve the presentation of the support measure and its properties.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central innovation is described only as 'a new notion of support' with no mathematical definition, no statement of how support is computed over temporally ordered discrete events, and no verification that the measure satisfies anti-monotonicity (support(r) ≤ support(r') whenever r' is a proper subsequence of r). This property is load-bearing for the correctness of any growth-based pruning procedure.

    Authors: We agree that the abstract provides insufficient detail on the new support measure. In the revised manuscript we will expand the abstract to include a concise mathematical definition of support over temporally ordered discrete events together with an explicit statement that the measure satisfies anti-monotonicity. The full formal definition and proof will be added to the body of the paper so that the pruning procedure's correctness can be directly assessed. revision: yes

  2. Referee: [Abstract] No section or equation is supplied that defines the support measure, proves the required lattice properties, or gives the pseudocode for the adapted growth procedure; without these the claim that the algorithm 'takes principles from the state of the art' cannot be evaluated for completeness or bias.

    Authors: We acknowledge that the current version of the manuscript does not supply the requested formal definition, lattice-property proofs, or pseudocode. We will add a dedicated subsection containing (i) the precise definition of the temporal support measure, (ii) proofs of anti-monotonicity and downward closure, and (iii) the adapted growth pseudocode. These additions will allow readers to evaluate how TSRuleGrowth extends RuleGrowth and will remove any ambiguity about completeness or bias. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No derivation chain or equations present; no circularity detectable

full rationale

The abstract and high-level description introduce TSRuleGrowth and a new support notion but contain no equations, derivations, or load-bearing steps. Without any quoted mathematical reductions, self-definitions, fitted predictions, or self-citation chains that collapse to inputs, no circularity of the enumerated kinds can be exhibited. The paper is treated as self-contained against external benchmarks with no evidence of internal equivalence by construction.

Assumptions & free parameters 0 free parameters · 0 assumptions · 0 invented entities

Only the abstract is available; no free parameters, axioms, or invented entities are described.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TSRuleGrowth : Extraction de r\`egles de pr\'ediction semi-ordonn\'ees \`a partir d'une s\'erie temporelle d'\'el\'ements discrets, application dans un contexte d'intelligence ambiante." pith.science (2026). https://pith.science/paper/PWVEBTAZ

@misc{pith2026190710054,
  author       = {Pith},
  title        = {Pith review of: TSRuleGrowth : Extraction de r\`egles de pr\'ediction semi-ordonn\'ees \`a partir d'une s\'erie temporelle d'\'el\'ements discrets, application dans un contexte d'intelligence ambiante},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PWVEBTAZ}},
  note         = {Machine review of arXiv:1907.10054}
}
read the original abstract

This paper presents a new algorithm: TSRuleGrowth, looking for partially-ordered rules over a time series. This algorithm takes principles from the state of the art of rule mining and applies them to time series via a new notion of support. We apply this algorithm to real data from a connected environment, which extract user habits through different connected objects.

Figures

Figures reproduced from arXiv: 1907.10054 by the authors.

Figure 5
Figure 5. figure 5. Pour cette règle [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages

  1. [1]

    Ces éléments représentent des variations basiques de don- nées boursières

    propose un système de fouille de règles basiques, où un élément en prédit un autre, sur une séquence d’éléments. Ces éléments représentent des variations basiques de don- nées boursières. Il peut aussi rechercher des règles plus complexes, où la condition est une séquence. Ce système permet donc de trouver des règles sur une série temporelle. Cependant, i...

    work page

  2. [2]

    Mais, recherchant des règles complètement ordonnées, il ne peut pas être appliqué dans notre cas

    peut être considéré comme une amélioration de [5], car il recherche des règles où la prédiction n’est pas limitée à un élément. Mais, recherchant des règles complètement ordonnées, il ne peut pas être appliqué dans notre cas

    work page

  3. [3]

    Le support d’un élément, d’un ensemble d’éléments ou d’une règle est le nombre de fenêtres dans lesquelles cet élément, ensemble ou règle apparaît

    introduit une notion de support pour série temporelle, via une fenêtre glissante à durée fixe. Le support d’un élément, d’un ensemble d’éléments ou d’une règle est le nombre de fenêtres dans lesquelles cet élément, ensemble ou règle apparaît. L’algorithme trouve des règles semi- ordonnées, en cherchant des ensembles fréquents d’élé- ments, puis en les comb...

    work page

  4. [4]

    Comment définir le support d’une règle dans une série temporelle qui évite le problème de la section 3.1 ?

    work page

  5. [5]

    Comment construire un algorithme de fouille de règles sur cette nouvelle mesure de support ? De plus, cet algorithme doit aborder les points suivants :

    work page

  6. [6]

    Comment limiter la durée des règles trouvées ?

    work page

  7. [7]

    Comment limiter la recherche à certains éléments dans la condition ou de la prédiction ?

    work page

  8. [8]

    coupée en deux

    Comment éviter qu’une règle soit trouvée deux fois ? RuleGrowth répond aux points 4 et 5, mais ne prend que des transactions en entrée. TRuleGrowth, utilise une fe- nêtre glissante qui peut être utilisée pour répondre au pro- blème 3 avec quelques modifications. Notre système utilise les principes de TRuleGrowth, mais les applique aux séries temporelles, p...

    work page

Show all 20 references
  1. [9]

    Ces règles, puisque la fenêtre d’observation est plus grande, peuvent révéler les habitudes de l’utilisateur

    Python 3.7.3, CPU : Intel(R) Xeon(R) Gold 5118 @ 2.30GHz, RAM : 128GiB, Ubuntu 18.04.2 LTS, Multiprocessing ajouté au code 7 0:00:09 0:01:26 0:14:24 2:24:00 24:00:00 1 10 100 1000 10000 100000 1 2 5 10 15 20 25 30 Temps d'exécution Nombre de règles Taille de window (en seconde...

  2. [10]

    Ahn and J.-Y

    K.-I. Ahn and J.-Y . Kim. Efficient Mining of Frequent Itemsets and a Measure of Interest for Association Rule Mining. J. Inf. Knowl. Manag. , 03(03) :245– 257, Sept. 2004

    2004

  3. [11]

    J. C. Augusto and P. McCullagh. Ambient intelli- gence : Concepts and applications. Comput. Sci. Inf. Syst., 4(1) :1–27, 2007

    2007

  4. [12]

    P. J. Azevedo and A. M. Jorge. Comparing Rule Mea- sures for Predictive Association Rules. In Machine Learning : ECML 2007 , Lecture Notes in Computer Science, pages 510–517. Springer Berlin Heidelberg, 2007

    2007

  5. [13]

    Cumin, G

    J. Cumin, G. Lefebvre, F. Ramparany, and J. L. Crow- ley. A Dataset of Routine Daily Activities in an Ins- trumented Home. In 11th International Conference on Ubiquitous Computing and Ambient Intelligence (UCAmI), Nov. 2017

    2017

  6. [14]

    Das, K.-I

    G. Das, K.-I. Lin, H. Mannila, G. Renganathan, and P. Smyth. Rule Discovery from Time Series. In Proceedings of the Fourth International Conference on Knowledge Discovery and Data Mining, KDD’98, pages 16–22. AAAI Press, 1998

    1998

  7. [15]

    Deogun and L

    J. Deogun and L. Jiang. Prediction Mining – An Ap- proach to Mining Association Rules for Prediction. 8 In Rough Sets, Fuzzy Sets, Data Mining, and Granu- lar Computing, Lecture Notes in Computer Science, pages 98–108. Springer Berlin Heidelberg, 2005

    2005

  8. [16]

    J. K. Febrer-Hernández, R. Hernández-León, C. Feregrino-Uribe, and J. Hernández-Palancar. SPaC-NF : A classifier based on sequential patterns with high netconf. Intell. Data Anal. , 20(5) :1101– 1113, Sept. 2016

    2016

  9. [17]

    Fournier-Viger, T

    P. Fournier-Viger, T. Gueniche, S. Zida, and V . S. Tseng. ERMiner : Sequential Rule Mining Using Equivalence Classes. In Advances in Intelligent Data Analysis XIII, pages 108–119. Springer International Publishing, 2014

    2014

  10. [18]

    Fournier-Viger, C.-W

    P. Fournier-Viger, C.-W. Wu, V . S. Tseng, L. Cao, and R. Nkambou. Mining Partially-Ordered Sequen- tial Rules Common to Multiple Sequences. IEEE Transactions on Knowledge and Data Engineering , 27(8) :2203–2216, Aug. 2015

    2015

  11. [19]

    Mannila, H

    H. Mannila, H. Toivonen, and A. Inkeri Verkamo. Discovery of Frequent Episodes in Event Sequences. Data Min Knowl Discov, 1(3) :259–289, Jan. 1997

    1997

  12. [20]

    Schlüter and S

    T. Schlüter and S. Conrad. About the analysis of time series with temporal association rule mining. In 2011 IEEE Symposium on Computational Intelligence and Data Mining (CIDM), pages 325–332, Apr. 2011. 9

    2011

Pith tools

Reviewed May 24, 2026 · model on record in the stance chip above.