Submitted Successfully!
To reward your contribution, here is a gift for you: A free trial for our video production service.
Thank you for your contribution! You can also upload a video entry or images related to this topic.
Version Summary Created by Modification Content Size Created at Operation
1
Maximilian Lowin
 -- 2468 2024-02-05 10:43:26 |
2 references update and layout
Fanny Huang
 Meta information modification 2468 2024-02-07 06:31:43 |

Video Upload Options

Do you have a full video?
Send video materials Upload full video

Confirm

Are you sure to Delete?
Yes No
Cite
If you have any further questions, please contact Encyclopedia Editorial Office.
Lowin, M. Association Rule Mining in Facility Management. Encyclopedia. Available online: https://encyclopedia.pub/entry/54752 (accessed on 18 May 2024).
Lowin M. Association Rule Mining in Facility Management. Encyclopedia. Available at: https://encyclopedia.pub/entry/54752. Accessed May 18, 2024.
Lowin, Maximilian. "Association Rule Mining in Facility Management" Encyclopedia, https://encyclopedia.pub/entry/54752 (accessed May 18, 2024).
Lowin, M. (2024, February 05). Association Rule Mining in Facility Management. In Encyclopedia. https://encyclopedia.pub/entry/54752
Lowin, Maximilian. "Association Rule Mining in Facility Management." Encyclopedia. Web. 05 February, 2024.
Association Rule Mining in Facility Management
Edit
This entry is adapted from the peer-reviewed paper 10.3390/make6010013

Maintenance represents a substantial share of work in various industries. Due to its significant financial impact, industry and research focus on improving the effectiveness of maintenance. Predictive Maintenance (PM) is one way to reduce costs and downtimes by planning maintenance work based on an asset’s actual condition rather than relying on fixed time-based maintenance cycles. Association rule mining (ARM) is a suitable method for PM tasks when data are unlabeled and less structured, as is the case in the facility management domain.

predictive maintenance facility management association rule mining (ARM)

1. Introduction

Maintenance represents a substantial share of work in various industries. Due to its significant financial impact [1], industry and research focus on improving the effectiveness of maintenance. Predictive Maintenance (PM) is one way to reduce costs and downtimes by planning maintenance work based on an asset’s actual condition rather than relying on fixed time-based maintenance cycles [2]. Multiple industries like aviation [3][4], manufacturing [5][6], and chemistry [7][8] vastly apply PM into their operational routine. Also, other industries like construction and facility management can benefit from improved maintenance methodologies. Building information modeling (BIM) and computer-aided facility management (CAFM) generate a large amount of data that can potentially be used for PM [9]. Such systems utilize digital twins to map the properties of physical entities like buildings or machines into the virtual world [10]. However, research on PM for facility management is rare. Some works in this area, such as those presented in [11] or [12], concentrate on the maintenance of specific infrastructure in buildings like heating, ventilation, and air conditioning (HVAC). Other works like [13] utilize sensor information to predict an infrastructure’s condition. However, most data are unstructured and are in text form [10][14]. This textual form is problematic since most PM algorithms require structured data, e.g., sensor measurements, to transform them into information ready for decision-making. Unlike numeric information, computers cannot interpret textual information directly. For instance, different textual formulations may have the same semantic meaning, while one word in a text can drastically change its whole meaning.
Nevertheless, neglecting textual data misses the opportunity to apply PM to areas where only textual data are available, e.g., in facility management, where facility owners rarely install sensors or are not allowed to store sensor information to protect the inhabitants’ or employees’ privacy. Thus, it is essential to utilize the already existing textual data by using knowledge extraction from textual databases [14]. Maintenance request analysis is an upcoming field in this setting [9]. Maintenance requests are textual information about specific maintenance tasks that describe defect infrastructure and add more contextual information. Facility managers often store them in CAFM software to track status updates and staff assignments. Utilizing this information for PM can potentially optimize processes and reduce breakdowns. However, there is little research on this topic [15], and most work still focuses on PM based on structured data [10].
Another problem in facility management and construction is the phenomenon of data monopoly: most companies that manage extensive facilities cannot share their data with others, making the datasets needed for PM challenging to access [9]. However, big datasets containing data from several companies are a crucial prerequisite for adopting most supervised PM models that require explicit labels. These labels rarely exist in the facility management context since they are expensive to collect. Additionally, relying on supervised labels is inadequate in this context since the nature of maintenance works is an ongoing process for PM and facility management [16]. A possible solution to overcome this limitation is using semi-supervised learning, transfer learning, and pre-trained models [9]. However, based on current understanding, there is no literature about applying such algorithms for PM in facility management. 

2. Predictive Maintenance for Facilities

Traditional maintenance management techniques are run-to-failure and preventive maintenance [2]. Run-to-failure maintenance strategies perform no maintenance work until a specific infrastructure breaks. Preventive maintenance strategies perform a time-driven schedule, e.g., yearly maintenance checks [2].
However, both approaches contradict optimized decision-making strategies that minimize costs and increase quality, safety, and productivity [10]. Therefore, the PM approach relies on an infrastructure’s physical condition to predict its future condition and potential failures [2]. To determine the physical condition, traditional PM algorithms use sensor information like the machine’s temperature, voltage, or current [17]. Typical PM approaches detect faults, i.e., low-frequency but high-impact events [3], or calculate an infrastructure’s lifetime and metrics like the mean time to failure [18]. However, the definition of PM is ambiguous, but its main aim is to improve an infrastructure’s operation using data [2]. T
Since most PM approaches rely on sensor data, applying PM for facility management is not trivial. Most corporate information is in the form of texts, especially in the construction sector [10][14]. Such texts can be document files, sheets, or semi-structured forms like Extensible Markup Language or Hypertext Markup Language that require Natural Language Processing (NLP) to extract the information from human language texts [9]. Maintenance requests are a valuable source of information to assess the condition of specific infrastructure in the context of facility management [15]. The work of [19] is among the first to utilize these requests. In the study, the authors created a prediction model that automatically assigns staff to specific requests containing unstructured text. Another field in this area is a relation analysis of such maintenance requests [10]. Wu and colleagues [10], in an extensive literature review, present that co-occurrence analysis, ARM, heuristic rules, or supervised learning are suitable approaches for relation analysis. While staff assignment is also a form of supervised machine learning, it requires explicit labels to learn a classification task. The authors of this literature review outline that most papers build on a supervised problem, while manually labeling datasets is often cumbersome and impractical.
In contrast, using unstructured text data is cheaper to collect and more accessible [10]. Only a few works use unsupervised techniques on textual data for predictive maintenance. Akhbardeh and colleagues utilize clustering on maintenance logs to find similar maintenance logs [20]. Bhardwaj and colleagues apply hand-crafted lexicographic sentiment analysis on maintenance reports to identify infrastructure health status [21]. A last unsupervised PM approach utilizing maintenance logs extracts machine components and their associated failures [22]. However, these approaches mainly help to estimate the status quo from maintenance requests but do not predict helpful information. Other works that integrate unsupervised machine learning for predictive maintenance, like anomaly detection, do not rely on textual data but on sensor data [23][24][25]. 

3. Association Rule Mining

ARM is a suitable method for PM tasks [7]. In their seminal work, Agrawal and colleagues [26] introduce ARM by utilizing shopping basket data and mine rules that present what items customers have purchased together. These rules are in the form of A → B, where A is the antecedent; B is the consequent, and A and B are a set of items/articles (n ≥ 1) in the basket. To determine relevant rules and their quality, ARM algorithms use different metrics, i.e., especially support and confidence, which the literature defines as follows [7]:
s u p p o r t A = A m ,
s u p p o r t A → B = A ∪ B m ,
c o n f i d e n c e A → B = s u p p o r t A → B s u p p o r t A .
Support is the probability of finding a transaction containing a respective item set [7]. The expression |𝐴| represents the number of transactions containing A; |𝐴∪𝐵| represents the number of transactions containing both A and B, and m represents the number of transactions. In addition, confidence is the conditional probability of finding the item set of the consequence given the occurrence of the antecedent [7]. These definitions are consistent with a wide range of works in the field of ARM [26][27][28][29]. Another popular metric is the lift, which calculates whether the two item sets A and B of rule A → B are dependent or independent of each other [27]:
l i f t A → B = s u p p o r t A → B s u p p o r t A * s u p p o r t B .
A lift value of one means that there is no relation between both item sets; a degree greater than one indicates a positive dependence and makes them interesting for further mining [27].
The most prominent algorithm for finding association rules is Apriori, which was introduced by [30]. It was one of the first algorithms for ARM that worked efficiently by maintaining minimum support and confidence in the association rules [30]. In their paper, the authors also present two other algorithms, AprioriTid and Apriori Hybrid. In [31], the authors give an overview of different ARM algorithms like the aforementioned Apriori derivatives and newer algorithms like frequent pattern-growth (FP-growth) and evaluate their performance regarding data support, speed, and accuracy. However, all ARM algorithms have in common that they rely on support and confidence.
Indeed, practice vastly uses ARM algorithms. One significant advantage is their ability to deal with large amounts of unstructured data and that they provide very interpretable rules suitable to enhance decision-making [28]. For instance, [32] applies the FP-growth algorithm for PM sensor data of Internet of Things hardware to extract association rules. ARM can also be used on text data, even if one cannot apply it on raw text but on different representations like a bag of words or term frequency [28]. Bag of words and term frequency are typical text representations using NLP. The concept of the bag of words considers the frequency of each word in a text while ignoring their position in the text [33]. The term frequency originates from [34] and describes how often a specific term (e.g., a word) occurs in a document. If one considers multiple documents, one can calculate the document frequency, i.e., the number of documents a term occurs [33]. Word embeddings are feature representations of a word (e.g., in the form of a vector) where each dimension aims to capture its syntactic and semantic meaning [35]. Typical tasks using ARM on text are summarization, topic and event detection, forecasting, and collaborative social systems [28]. For instance, the work of [36] applies Apriori to customer reviews. The work of [37] follows the Apriori algorithm, combining text analysis with knowledge bases to form semantic rules. In [38], the authors mine association rules from medical records by extracting medical features like symptoms, diseases, and medicine. The work of [27] employs ARM to find associations between words in the Azerbaijani language using a bag-of-words approach. Finally, the authors of [7] demonstrate the suitability of ARM for industrial use cases. They first generate association rules and then apply linear programming to select components to repair to improve their infrastructure’s overall robustness.
However, there is one major problem when applying ARM on text using simple representations like bag-of-words or term frequency–inverse document frequency (TF–IDF). Unstructured free texts use heterogeneous language with little consistency, i.e., they contain typos, acronyms, abbreviations, and jargon [39][40]. Applying ARM on large text can result in data dispersion and binary representations that might lead to sparse matrices [28]. Since facility management is a domain where one can expect such an inconsistency, there is a need to extend the typical ARM definitions. Transfer learning and large language models like BERT or GPT can overcome such limitations by capturing contextual information [9]. Bidirectional Encoder Representations from Transformers (BERT) is a language representation model that is already pre-trained, meaning that one only needs to fine-tune a model for a specific use case, making it handy to apply for various problems [41]. A transformer is an ML model that typically uses an encoder component to transform text into a computer-understandable format (i.e., encodes an input sequence of symbol representations like words into a sequence of continuous representations) [42] and adds the concept of attention. The attention mechanism allows a neural network to focus on specific, relevant parts of the input sequence [40]. The Generative Pre-trained Transformer (GPT) is a transformer model that can handle text and images as inputs and generates text as output by adding an additional decoder component to recast the continuous representation into the original symbol representation [43]. The model became famous for its application in the software ChatGPT. 

4. Extensions of Association Rule Mining

Most ARM research consists of applying ARM algorithms to various domains or focuses on improving the performance of different mining techniques. However, two crucial extensions to ARM that are necessary when applying it to facility management are emphasized: (1) the integration of similarity; and (2) temporal information.
The main idea of integrating similarity into ARM is that a specific term or phrase can be very similar or even identical to another term while not being classified as such by ARM due to its separate representation. For instance, the work of [44] gives the example of Ceylon as the former name of Sri Lanka. Both terms reflect the same nation. However, they have different textual representations. In their work, the authors of [44] use Apriori to detect semantically identical but temporally different concepts and utilize Jaccard’s coefficient as a similarity measure.
Another strain of the literature integrates the concept of similarity by mining rules from similar texts. The work of [14] applies ARM to text previously clustered by similarity. However, one can also integrate the concept of similarity directly after rule creation. In [45], the authors present a way to apply cosine similarity to rules generated by Apriori to unify similar rules. In [46], the authors present a similar approach where they create rules by running Apriori and extend their rules based on similarity by re-calculating the support of extended rules. They also use cosine similarity as a general-purpose metric but emphasize that a domain-specific similarity metric could be helpful. However, their approach is only suitable for generating rules with a unitary length of the antecedent [46]. 
Another essential extension when using ARM for facility management is integrating temporal information. In a literature review about temporal ARM performed in [29], the authors construct a taxonomy of how to incorporate time into ARM. They suggest integrating time as an implied component or an integral component. When integrating time as an implied component, the time variable provides information about the order, including temporal constraints and sequences [29]. Furthermore, the authors describe time as an integral component in that the time variable becomes an attribute within the learning process and that time indicates potentially periodical or time interval-based patterns. Both approaches are helpful for ARM and PM. For instance, the work of [47] applies temporal ARM for train maintenance. It uses time information as a criterion that ARM can predict target events like repairs early enough to allow for logistic and maintenance actions. On the contrary, it utilizes time information as a limit that the potential items for ARM are also recent and relevant. This procedure allows for splitting the prediction time into warning and monitoring times [47]. The work of [48] adds a recency weight to transactions to limit items being temporally close to each other. In addition, it uses a time decay function to avoid making wrong decisions with out-of-date rules and adopt the corresponding definition of the support function.

References

  1. Widodo, A.; Yang, B.-S. Support Vector Machine in Machine Condition Monitoring and Fault Diagnosis. Mech. Syst. Signal Process. 2007, 21, 2560–2574.
  2. Mobley, R.K. An Introduction to Predictive Maintenance; Elsevier: Amsterdam, The Netherlands, 2002; ISBN 978-0-7506-7531-4.
  3. Alestra, S.; Bordry, C.; Brand, C.; Burnaev, E.; Erofeev, P.; Papanov, A.; Silveira-Freixo, C. Rare Event Anticipation and Degradation Trending for Aircraft Predictive Maintenance. In Proceedings of the 11th World Congress on Computational Mechanics, WCCM, Barcelona, Spain, 20–25 July 2014; Volume 5, p. 6571.
  4. Zhu, H.; Gao, J.; Li, D.; Tang, D. A Web-Based Product Service System for Aerospace Maintenance, Repair and Overhaul Services. Comput. Ind. 2012, 63, 338–348.
  5. Adu-Amankwa, K.; Attia, A.K.; Janardhanan, M.N.; Patel, I. A Predictive Maintenance Cost Model for CNC SMEs in the Era of Industry 4.0. Int. J. Adv. Manuf. Technol. 2019, 104, 3567–3587.
  6. Hrnjica, B.; Softic, S. Explainable AI in Manufacturing: A Predictive Maintenance Case Study. In Advances in Production Management Systems. Towards Smart and Digital Manufacturing; Lalic, B., Majstorovic, V., Marjanovic, U., von Cieminski, G., Romero, D., Eds.; Springer International Publishing: Cham, Switzerland, 2020; pp. 66–73.
  7. Antomarioni, S.; Pisacane, O.; Potena, D.; Bevilacqua, M.; Ciarapica, F.E.; Diamantini, C. A Predictive Association Rule-Based Maintenance Policy to Minimize the Probability of Breakages: Application to an Oil Refinery. Int. J. Adv. Manuf. Technol. 2019, 105, 3661–3675.
  8. Nadj, M.; Jegadeesan, H.; Maedche, A.; Hoffmann, D.; Erdmann, P. A Situation Awareness Driven Design for Predictive Maintenance Systems: The Case of Oil and Gas Pipeline Operations. In Proceedings of the 24th European Conference on Information Systems, Istanbul, Turkey, 12–15 June 2016; pp. 1–10.
  9. Ding, Y.; Ma, J.; Luo, X. Applications of Natural Language Processing in Construction. Autom. Constr. 2022, 136, 104169.
  10. Wu, C.; Li, X.; Guo, Y.; Wang, J.; Ren, Z.; Wang, M.; Yang, Z. Natural Language Processing for Smart Construction: Current Status and Future Directions. Autom. Constr. 2022, 134, 104059.
  11. Ghofrani, A.; Nazemi, S.D.; Jafari, M.A. HVAC Load Synchronization in Smart Building Communities. Sustain. Cities Soc. 2019, 51, 101741.
  12. West, S.R.; Guo, Y.; Wang, X.R.; Wall, J. Automated Fault Detection and Diagnosis of HVAC Subsystems Using Statistical Machine Learning. In Proceedings of the Building Simulation, Sydney, Australia, 14–16 November 2011.
  13. Cheng, J.C.; Chen, W.; Tan, Y.; Wang, M. A BIM-Based Decision Support System Framework for Predictive Maintenance Management of Building Facilities. In Proceedings of the 16th International Conference on Computing in Civil and Building Engineering (ICCCBE2016), Osaka, Japan, 6–8 July 2016; pp. 711–718.
  14. Ur-Rahman, N.; Harding, J.A. Textual Data Mining for Industrial Knowledge Management and Text Classification: A Business Oriented Approach. Expert Syst. Appl. 2012, 39, 4729–4739.
  15. Bortolini, R.; Forcada, N. Analysis of Building Maintenance Requests Using a Text Mining Approach: Building Services Evaluation. Build. Res. Inf. 2020, 48, 207–217.
  16. Folino, F.; Folino, G.; Guarascio, M.; Pontieri, L. Semi-Supervised Discovery of DNN-Based Outcome Predictors from Scarcely-Labeled Process Logs. Bus. Inf. Syst. Eng. 2022, 64, 729–749.
  17. Hashemian, H.M. State-of-the-Art Predictive Maintenance Techniques. IEEE Trans. Instrum. Meas. 2010, 60, 226–236.
  18. Virk, S.M.; Muhammad, A.; Martinez-Enriquez, A. Fault Prediction Using Artificial Neural Network and Fuzzy Logic. In Proceedings of the 2008 Seventh Mexican International Conference on Artificial Intelligence, Atizapan de Zaragoza, Mexico, 27 October 2008; pp. 149–154.
  19. Mo, Y.; Zhao, D.; Du, J.; Syal, M.; Aziz, A.; Li, H. Automated Staff Assignment for Building Maintenance Using Natural Language Processing. Autom. Constr. 2020, 113, 103150.
  20. Akhbardeh, F.; Desell, T.; Zampieri, M. NLP Tools for Predictive Maintenance Records in MaintNet. In Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing: System Demonstrations; Wong, D., Kiela, D., Eds.; Association for Computational Linguistics: Suzhou, China, 2020; pp. 26–32.
  21. Bhardwaj, A.S.; Veeramani, D.; Zhou, S. Identifying Equipment Health Status from Maintenance Records Using Lexicon Based Unsupervised Sentiment Analysis Adjusted for Negation (LUSAA-N). Comput. Ind. Eng. 2023, 186, 109693.
  22. Devaney, M.; Ram, A.; Qiu, H.; Lee, J. Preventing Failures by Mining Maintenance Logs with Case-Based Reasoning. In Proceedings of the 59th Meeting of the Society for Machinery Failure Prevention Technology (MFPT-59), Virginia Beach, VA, USA, 18–21 April 2005.
  23. Carrasco, J.; López, D.; Aguilera-Martos, I.; García-Gil, D.; Markova, I.; García-Barzana, M.; Arias-Rodil, M.; Luengo, J.; Herrera, F. Anomaly Detection in Predictive Maintenance: A New Evaluation Framework for Temporal Unsupervised Anomaly Detection Algorithms. Neurocomputing 2021, 462, 440–452.
  24. da Silva Arantes, J.; da Silva Arantes, M.; Fröhlich, H.B.; Siret, L.; Bonnard, R. A Novel Unsupervised Method for Anomaly Detection in Time Series Based on Statistical Features for Industrial Predictive Maintenance. Int. J. Data Sci. Anal. 2021, 12, 383–404.
  25. Graß, A.; Beecks, C.; Soto, J.A.C. Unsupervised Anomaly Detection in Production Lines. In Machine Learning for Cyber Physical Systems; Beyerer, J., Kühnert, C., Niggemann, O., Eds.; Springer: Berlin/Heidelberg, Germany, 2019; pp. 18–25.
  26. Agrawal, R.; Imieliński, T.; Swami, A. Mining Association Rules Between Sets of Items in Large Databases. In Proceedings of the 1993 ACM SIGMOD International Conference on Management of Data, Washington, DC, USA, 26–28 May 1993; pp. 207–216.
  27. Adamov, A.Z. Mining Term Association Rules from Unstructured Text in Azerbaijani Language. In Proceedings of the 2018 IEEE 12th International Conference on Application of Information and Communication Technologies (AICT), Almaty, Kazakhstan, 17–19 October 2018; pp. 1–4.
  28. Diaz-Garcia, J.A.; Ruiz, M.D.; Martin-Bautista, M.J. A Survey on the Use of Association Rules Mining Techniques in Textual Social Media. Artif. Intell. Rev. 2023, 56, 1175–1200.
  29. Segura-Delgado, A.; Gacto, M.J.; Alcalá, R.; Alcalá-Fdez, J. Temporal Association Rule Mining: An Overview Considering the Time Variable as an Integral or Implied Component. WIREs Data Min. Knowl. Discov. 2020, 10, e1367.
  30. Agrawal, R.; Srikant, R. Fast Elgorithms for Mining Association Rules. In Proceedings of the 20th International Conference on Very Large Data Bases (VLDB), Santiago, Chile, 12–15 September 1994; Volume 1215, pp. 487–499.
  31. Kumbhare, T.A.; Chobe, S.V. An Overview of Association Rule Mining Algorithms. Int. J. Comput. Sci. Inf. Technol. 2014, 5, 927–930.
  32. Kireev, V.S.; Guseva, A.I.; Bochkaryov, P.V.; Kuznetsov, I.A.; Filippov, S.A. Association Rules Mining for Predictive Analytics in IoT Cloud System. In Biologically Inspired Cognitive Architectures 2018; Samsonovich, A.V., Ed.; Springer International Publishing: Cham, Switzerland, 2019; pp. 107–112.
  33. Jurafsky, D.; Martin, J.H. Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition; Prentice Hall: Upper Saddle River, NJ, USA, 2008; ISBN 0-13-095069-6.
  34. Luhn, H.P. A Statistical Approach to Mechanized Encoding and Searching of Literary Information. IBM J. Res. Dev. 1957, 1, 309–317.
  35. Turian, J.; Ratinov, L.; Bengio, Y. Word Representations: A Simple and General Method for Semi-Supervised Learning. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, Uppsala, Sweden, 11–16 July 2010; pp. 384–394.
  36. Hemalatha, B.; Velmurugan, T. Direct-Indirect Association Rule Mining for Online Shopping Customer Data Using Natural Language Processing. Int. J. Recent Technol. Eng. 2019, 8, 2277–3878.
  37. Ren, S.; Li, Z.; Wang, H.; Li, Y.; Shen, K.; Cheng, S. NEARM: Natural Language Enhanced Association Rules Mining. In Proceedings of the 2018 IEEE International Conference on Data Mining Workshops (ICDMW), Singapore, 17–20 November 2018; pp. 438–445.
  38. Lakshmi, K.S.; Vadivu, G. Extracting Association Rules from Medical Health Records Using Multi-Criteria Decision Analysis. Proc. Comput. Sci. 2017, 115, 290–295.
  39. Edwards, B.; Zatorsky, M.; Nayak, R. Clustering and Classification of Maintenance Logs Using Text Data Mining. In Proceedings of the 7th Australasian Data Mining Conference, Glenelg, SA, Australia, 27–28 November 2008; Volume 87, pp. 193–199.
  40. Usuga-Cadavid, J.P.; Lamouri, S.; Grabot, B.; Fortin, A. Using Deep Learning to Value Free-Form Text Data for Predictive Maintenance. Int. J. Prod. Res. 2022, 60, 4548–4575.
  41. Devlin, J.; Chang, M.-W.; Lee, K.; Toutanova, K. BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, MN, USA, 2–7 June 2019; Association for Computational Linguistics: Minneapolis, MN, USA, 2019; pp. 4171–4186.
  42. Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A.N.; Kaiser, Ł.; Polosukhin, I. Attention Is All You Need. Adv. Neural Inf. Process. Syst. 2017, 30, 5998–6008.
  43. OpenAI GPT-4 Technical Report. arXiv 2023, arXiv:2303.08774.
  44. Kaluarachchi, A.C.; Varde, A.S.; Bedathur, S.; Weikum, G.; Peng, J.; Feldman, A. Incorporating Terminology Evolution for Query Translation in Text Retrieval with Association Rules. In Proceedings of the 9th ACM International Conference on Information and Knowledge Management, New York, NY, USA, 6–11 November 2010; Association for Computing Machinery: New York, NY, USA, 2010; pp. 1789–1792.
  45. Zeng, A.P.; Liu, D.; Chen, H.M. An Improved Apriori Algorithm Based on Similarity. Adv. Mater. Res. 2012, 532–533, 1825–1829.
  46. Keith Norambuena, B.; Villegas, C. An Extension to Association Rules Using a Similarity-Based Approach in Semantic Cector Spaces. Intell. Data Anal. 2019, 23, 587–607.
  47. Sammouri, W.; Côme, E.; Oukhellou, L.; Aknin, P. Mining Floating Train Data Sequences for Temporal Association Rules within a Predictive Maintenance Framework. In Advances in Data Mining. Applications and Theoretical Aspects; Perner, P., Ed.; Springer: Berlin/Heidelberg, Germany, 2013; pp. 112–126.
  48. Han, Y.; Yu, D.; Yin, C.; Zhao, Q. Temporal Association Rule Mining and Updating and Their Application to Blast Furnace in the Steel Industry. Comput. Intell. Neurosci. 2020, 2020, 7467213.
More
© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.
Upload a video for this entry
Information
Subjects: Computer Science, Information Systems
Contributor MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :
Maximilian Lowin
View Times: 82
Revisions: 2 times (View History)
Update Date: 07 Feb 2024
Table of Contents
    1000/1000
    Hot Most Recent
    Feedback