Twitter Data Mining for the Diagnosis of Leaks in Drinking Water Distribution Networks
Abstract
:1. Introduction
- (1)
- Visual Inspection: Videos can provide a more detailed and dynamic view of the leak, helping experts to identify the exact location and extent of the leak. This information can help in assessing the severity of the leak and prioritizing repair work.
- (2)
- Audio Analysis: Videos can also capture the sound of the leak, which can help in diagnosing the type of leak and its possible cause. For example, the sound of water gushing out of a pipe may indicate a large, pressurized leak, while a faint dripping sound may indicate a smaller, slower leak.
- (3)
- Social Context: Videos can also provide social context to the leak, such as the presence of bystanders or the reaction of the public to the leak. This information can help in assessing the urgency of the repair work and in planning effective communication strategies.
- (4)
- Pattern recognition: By using machine learning algorithms, it is possible to train models on video data to recognize patterns of leak behavior.
2. Related Works
3. Materials and Methods
- 1
- Fault detection: its goal is to identify the presence of a possible fault in the system.
- 2
- Fault isolation: the purpose of this task is to pinpoint the exact cause of the fault, as well as its localization.
- 3
- Fault identification: This task completes the fault diagnosis by characterizing the type of fault, its size, and its profile.
- 1
- Signal-based methods: The mechanism of these methods is based on the search for fault symptoms in the signals of the systems under-diagnosis, such as pressure, flow, temperature, vibrations, etc. In a healthy state, features of the acquired signals such as frequency, amplitude, phase, damping, and ripple have nominal values, which in a faulty state differ from the nominal ones. These methods can be classified according to the feature typologies: time domain techniques, frequency domain techniques and time/frequency domain techniques.
- 2
- Model-based methods: These methods can be roughly classified into (1) methods without estimation error feedback and (2) methods with estimation error feedback. In the first class of methods, a mathematical model is fed with the input information of the system under diagnosis. The response of the model is compared with the response of the system. If there is a discrepancy between the responses, it is probably because there exists a fault or a set of faults. For obtaining good results with this type of method, it is necessary that the model be well calibrated in healthy conditions. Otherwise, false alarms will appear. A drawback of this class is that a disturbance can be mistaken for a fault. Another drawback of methods without estimation error feedback is that they only serve to detect faults but not to locate or isolate them. The methods with estimation feedback errors are more advantageous. The error between the model response and system response is injected into the model as an additional input and then multiplied by a gain that causes this error to approach zero over time. These methods are: (1) usefulness in detecting, locating, and isolating faults; (2) usefulness for applications in real time; (3) being robust against disturbances. These methods are also known as observer-based methods since a model with estimation error feedback is called a state observer.
- (3)
- Data-based methods: Methods based on models are appropriate to use when the dimension of the process (under-diagnosis) is low and it can be modeled with low-order models. However, to diagnose faults in more complex systems, these methods are no longer recommended. An alternative to diagnosing faults without using models is using data-based methods, which use information acquired from the process (under diagnosis). It can be said that these methods are the recent alternative to active supervision of systems too complex to have an explicit analytic model or signal symptoms of faulty behavior. Data-based methods use artificial intelligence and statistics algorithms to learn from data for discovering hidden patterns in the information (system variables).
3.1. Twitter Data Mining for Fault Diagnosis
3.2. Twitter Data Mining for Diagnosing Leaks
3.2.1. Collection of Tweets
3.2.2. Tweets Cleaning
3.2.3. Information Identification
3.2.4. Transformation of Information
3.2.5. Information Display
Algorithm 1 Twitter Search and Filtering Algorithm |
|
3.2.6. Use of Information
- Leak density is a metric used to describe the frequency or concentration of leaks in a water distribution network. It is typically measured as the number of leaks per unit area (e.g., leaks per square kilometer). The calculation of leak density involves dividing the total number of leaks in a given area by the size of that area. The leak density can be used by water distribution companies to prioritize maintenance or repair work in the areas with the highest leak density. By addressing leaks in these areas, water utilities can reduce water losses, improve the efficiency of the distribution system, and potentially save money on water treatment and pumping costs. Additionally, the leak density can be used to identify potential areas of pipe corrosion or damage, or to detect patterns in the occurrence of leaks, which can help inform decisions on pipe replacement or rehabilitation.
- Leak frequency in a water distribution network can be calculated by dividing the total number of leaks that occurred in a given period by the total length of the water distribution network. The resulting value indicates the average number of leaks that occurred per unit length of the network during that period. For example, if there were 10 leaks in a water distribution network of 10 kilometers in a month, the leak frequency would be 1 leak per kilometer per month (10 leaks ÷ 10 kilometers = 1 leak/kilometer). By tracking the frequency of leaks over time, water utility companies can identify areas of the network that are prone to leaks and prioritize maintenance and repair efforts in those areas.
- Leak probability in an area refers to the likelihood or chance of a leak occurring in a specific region of a water distribution network. It is usually calculated based on historical data about leak occurrences in the area and other relevant factors such as the age and condition of the pipes, water pressure, and environmental conditions. By estimating the leak probability, water utilities and managers can prioritize maintenance and repair efforts, allocate resources more effectively, and proactively address potential leaks before they become significant issues.
Algorithm 2 Twitter-based Leak Diagnosis Algorithm |
|
4. Results
5. Discussion
- 1.
- Real-time reporting: Twitter allows for instant reporting of leaks, while reporting leaks via telephone may take longer as individuals may have to wait on hold or navigate automated systems.
- 2.
- More detailed information: Twitter allows for more detailed information to be provided, including pictures or videos of the leaks, which can aid in identifying the location and severity of the leak.
- 3.
- Cost-effective: Twitter reporting is a cost-effective method of identifying leaks, as it does not require the same amount of resources as traditional telephone reporting systems.
- 1.
- Limited coverage: Not everyone uses Twitter or social media, so the sample of reports may not be representative of the entire population. This can lead to underreporting or biased reporting of leaks.
- 2.
- Data privacy: Twitter reports may contain personal information or sensitive data that may need to be protected. Water utility companies may need to ensure that the data collected from social media platforms are compliant with data protection regulations.
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
Abbreviations
UN | United Nations |
WDN | Water Distribution Networks |
GIS | Geographic Information System |
API | Application Programming Interface |
NRC | National Research Council Canada |
JMA | Japan Meteorological Agency |
MELSN | Mixed Eulerian/Lagrangian Sensor Network |
ALEN | Arbitrary Lagrangian-Eulerian Network |
JCPOA | Joint Comprehensive Plan of Action |
NLTK | Natural Language Toolkit |
VADER | Valence Aware Dictionary and Sentiment Reasoner |
NER | Named Entity Recognition Model |
CNN | Convolutional Neural Network |
MELSN | Mixed Eulerian/Lagrangian Sensor Network |
ALEN | Arbitrary Lagrangian-Eulerian Network |
AI | Artificial Intelligence |
References
- Ling, T. A Global Study about Water Crisis. In Proceedings of the 2021 International Conference on Social Development and Media Communication (SDMC 2021), Sanya, China, 26–28 November 2021; Atlantis Press: Paris, France, 2022; pp. 809–814. [Google Scholar]
- Briseño, H.; Sánchez, A. Decentralization, consolidation, and crisis of urban water management in Mexico. Tecnol. y Cienc. Del Agua 2018, 9, 25–47. [Google Scholar] [CrossRef]
- Khalifa, D.S.; El Atty, A.; Donia, N.S.; Moussa, A.; Mohamed, A. Analysis and Assessment of Water Losses in Domestic Water Distribution Networks. J. Environ. Sci. 2022, 51, 1–23. [Google Scholar] [CrossRef]
- Verde, C.; Torres, L. Modeling and Monitoring of Pipelines and Networks: Advanced Tools for Automatic Monitoring and Supervision of Pipelines; Springer: Berlin/Heidelberg, Germany, 2017; Volume 7. [Google Scholar]
- Carpentier, P.; Cohen, G. State estimation and leak detection in water distribution networks. Civ. Eng. Syst. 1991, 8, 247–257. [Google Scholar] [CrossRef]
- Pérez, R.; Puig, V.; Pascual, J.; Quevedo, J.; Landeros, E.; Peralta, A. Methodology for leakage isolation using pressure sensitivity analysis in water distribution networks. Control Eng. Pract. 2011, 19, 1157–1167. [Google Scholar] [CrossRef] [Green Version]
- Soldevila, A.; Blesa, J.; Tornil-Sin, S.; Duviella, E.; Fernandez-Canti, R.M.; Puig, V. Leak localization in water distribution networks using a mixed model-based/data-driven approach. Control Eng. Pract. 2016, 55, 162–173. [Google Scholar] [CrossRef] [Green Version]
- Li, X.; Wen, Y.; Jiang, J.; Daim, T.; Huang, L. Identifying potential breakthrough research: A machine learning method using scientific papers and Twitter data. Technol. Forecast. Soc. Chang. 2022, 184, 122042. [Google Scholar] [CrossRef]
- Sakaki, T.; Okazaki, M.; Matsuo, Y. Earthquake shakes twitter users: Real-time event detection by social sensors. In Proceedings of the 19th International Conference on World Wide Web, Raleigh, NC, USA, 26–30 April 2010; pp. 851–860. [Google Scholar]
- Jordan, S.E.; Hovet, S.E.; Fung, I.C.H.; Liang, H.; Fu, K.W.; Tse, Z.T.H. Using Twitter for public health surveillance from monitoring and prediction to public response. Data 2018, 4, 6. [Google Scholar] [CrossRef] [Green Version]
- Bonifazi, G.; Breve, B.; Cirillo, S.; Corradini, E.; Virgili, L. Investigating the COVID-19 vaccine discussions on Twitter through a multilayer network-based approach. Inf. Process. Manag. 2022, 59, 103095. [Google Scholar] [CrossRef]
- Pascual-Ferrá, P.; Alperstein, N.; Barnett, D.J. Social Network Analysis of COVID-19 Public Discourse on Twitter: Implications for Risk Communication. Disaster Med. Public Health Prep. 2022, 16, 561–569. [Google Scholar] [CrossRef] [PubMed]
- Pilařová, L.; Kvasničková Stanislavská, L.; Pilař, L.; Balcarová, T.; Pitrová, J. Cultured Meat on the Social Network Twitter: Clean, Future and Sustainable Meats. Foods 2022, 11, 2695. [Google Scholar] [CrossRef] [PubMed]
- Rahman, S.; Jahan, N.; Sadia, F.; Mahmud, I. Social crisis detection using Twitter based text mining-a machine learning approach. Bull. Electr. Eng. Inform. 2023, 12, 1069–1077. [Google Scholar] [CrossRef]
- Qorib, M.; Oladunni, T.; Denis, M.; Ososanya, E.; Cotae, P. COVID-19 vaccine hesitancy: Text mining, sentiment analysis and machine learning on COVID-19 vaccination twitter dataset. Expert Syst. Appl. 2023, 212, 118715. [Google Scholar] [CrossRef]
- Choi, Y.J.C.E.J. The Early Emotional Responses and Central Issues of People in the Epicenter of the COVID-19 Pandemic: An Analysis from Twitter Text Mining. Int. J. Ment. Health Promot. 2023, 25, 21–29. [Google Scholar] [CrossRef]
- Zarrabeitia-Bilbao, E.; Rio-Belver, R.M.; Alvarez-Meaza, I.; de Alegría-Mancisidor, I.M. World Environment Day: Understanding Environmental Programs Impact on Society Using Twitter Data Mining. Soc. Indic. Res. 2022, 164, 263–284. [Google Scholar] [CrossRef]
- Alhuzali, H.; Zhang, T.; Ananiadou, S. Emotions and Topics Expressed on Twitter During the COVID-19 Pandemic in the United Kingdom: Comparative Geolocation and Text Mining Analysis. J. Med Internet Res. 2022, 24, e40323. [Google Scholar] [CrossRef]
- Behzadidoost, R.; Hasheminezhad, M.; Farshi, M.; Derhami, V.; Alamiyan-Harandi, F. A framework for text mining on Twitter: A case study on joint comprehensive plan of action (JCPOA)-between 2015 and 2019. Qual. Quant. 2022, 56, 3053–3084. [Google Scholar] [CrossRef]
- Arumugam, S.S. Development of argument based opinion mining model with sentimental data analysis from twitter content. Concurr. Comput. Pract. Exp. 2022, 34, e6956. [Google Scholar] [CrossRef]
- Jiang, J.Y.; Zhou, Y.; Chen, X.; Jhou, Y.R.; Zhao, L.; Liu, S.; Yang, P.C.; Ahmar, J.; Wang, W. COVID-19 Surveiller: Toward a robust and effective pandemic surveillance system based on social media mining. Philos. Trans. R. Soc. A 2022, 380, 20210125. [Google Scholar] [CrossRef]
- Vukmirovic, M.; Raspopovic Milic, M.; Jovic, J. Twitter Data Mining to Map Pedestrian Experience of Open Spaces. Appl. Sci. 2022, 12, 4143. [Google Scholar] [CrossRef]
- Khetarpaul, S.; Sharma, D.; Jose, J.I.; Saragur, M. Real-Time Detection and Visualization of Traffic Conditions by Mining Twitter Data. In Proceedings of the Australasian Database Conference, Sydney, Australia, 3–4 September 2022; Springer: Berlin/Heidelberg, Germany, 2022; pp. 141–152. [Google Scholar]
- de Bruijn, J.A.; de Moel, H.; Jongman, B.; de Ruiter, M.C.; Wagemaker, J.; Aerts, J.C. A global database of historic and real-time flood events based on social media. Sci. Data 2019, 6, 1–12. [Google Scholar] [CrossRef] [Green Version]
- De Bruijn, J.A.; de Moel, H.; Jongman, B.; Wagemaker, J.; Aerts, J.C. TAGGS: Grouping tweets to improve global geoparsing for disaster response. J. Geovisualization Spat. Anal. 2018, 2, 1–14. [Google Scholar] [CrossRef] [Green Version]
- Sarker, A.; O’connor, K.; Ginn, R.; Scotch, M.; Smith, K.; Malone, D.; Gonzalez, G. Social media mining for toxicovigilance: Automatic monitoring of prescription medication abuse from Twitter. Drug Saf. 2016, 39, 231–240. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Gerber, M.S. Predicting crime using Twitter and kernel density estimation. Decis. Support Syst. 2014, 61, 115–125. [Google Scholar] [CrossRef]
- Isermann, R. Fault-Diagnosis Applications: Model-Based Condition Monitoring: Actuators, Drives, Machinery, Plants, Sensors, and Fault-Tolerant Systems; Springer Science & Business Media: Berlin, Germany, 2011. [Google Scholar]
- Gonzalez-Jimenez, D.; Del-Olmo, J.; Poza, J.; Garramiola, F.; Madina, P. Data-driven fault diagnosis for electric drives: A review. Sensors 2021, 21, 4024. [Google Scholar] [CrossRef]
- Tinka, A.; Rafiee, M.; Bayen, A.M. Floating sensor networks for river studies. IEEE Syst. J. 2012, 7, 36–49. [Google Scholar] [CrossRef]
- Canepa, E.; Odat, E.; Dehwah, A.; Mousa, M.; Jiang, J.; Claudel, C. A sensor network architecture for urban traffic state estimation with mixed eulerian/lagrangian sensing based on distributed computing. In Proceedings of the International Conference on Architecture of Computing Systems, Lubeck, Germany, 25–28 February 2014; Springer: Berlin/Heidelberg, Germany, 2014; pp. 147–158. [Google Scholar]
- Hirt, C.; Amsden, A.; Cook, J. An arbitrary Lagrangian-Eulerian computing method for all flow speeds. J. Comput. Phys. 1974, 14, 227–253. [Google Scholar] [CrossRef]
- Goodchild, M.F. Citizens as sensors: The world of volunteered geography. GeoJournal 2007, 69, 211–221. [Google Scholar] [CrossRef] [Green Version]
- Yoon, S.; Elhadad, N.; Bakken, S. A practical approach for content mining of tweets. Am. J. Prev. Med. 2013, 45, 122–129. [Google Scholar] [CrossRef] [Green Version]
- Ralston, M.R.; O’Neill, S.; Wigmore, S.J.; Harrison, E.M. An exploration of the use of social media by surgical colleges. Int. J. Surg. 2014, 12, 1420–1427. [Google Scholar] [CrossRef] [Green Version]
- Kayed, M.; Dakrory, S.; Ali, A.A. Postal address extraction from the web: A comprehensive survey. Artif. Intell. Rev. 2021, 55, 1085–1120. [Google Scholar] [CrossRef]
Reference | Objective | Methodology | Application Field |
---|---|---|---|
[11] | To propose a generic multilayer network model for representing social networks. The model is designed to facilitate the analysis of topics in social media and can be adapted to different domains, including discussion analysis, topic analysis, and user information dissemination. | The model consists of two layers: one that describes the type of user interactions and another that connects users based on a topic extracted from their posts. The paper presents a case study on people’s perception of COVID-19 vaccines by analyzing a dataset of tweets, which demonstrates the effectiveness of the proposed approach. | Health |
[12] | To demonstrate the use of social network analysis to understand public discourse on Twitter around the novel coronavirus disease 2019 (COVID-19) pandemic. The authors examined different network properties that might affect the successful dissemination and adoption of public health messages from public health officials and health agencies. | The authors focused on conversations on Twitter during three key communication events from late January to early June of 2020. They employed Netlytic, a Web-based software that collects publicly available data from social media sites such as Twitter. | Health |
[13] | To identify key characteristics of cultured meat based on communication analysis of social media. | The study used Twitter as a platform and analyzed 36,356 tweets posted by 4128 individual users. The analysis found that the main communicated characteristics of cultured meat were “clean meat”, “future meat”, and “sustainable meat”. The study highlights the importance of communication in the development of the cultured meat market and the need for producers to engage with customers through appropriate communication. | Sustainability |
[14] | To find out the recent major social crisis. | First, tweeplers, a trend detector online tool, is used to identify common keywords. Then, with the help of API provided by Twitter, a search is conducted on the recent common keywords on Twitter. | Social |
[15] | To analyze citizens’ COVID-19 vaccine hesitancy. | The authors examine three sentiment computation methods (Azure Machine Learning, VADER, and TextBlob) to analyze COVID-19 vaccine hesitancy. In addition, five learning algorithms (Random Forest, Logistics Regression, Decision Tree, LinearSVC, and Naïve Bayes) with different combinations of three vectorization methods (Doc2Vec, CountVectorizer, and TF-IDF) were deployed. | Health |
[16] | To explores citizens’ emotional responses in the context of the COVID-19 pandemic. | Text Network Analysis was used to identify the tweets’ central issue. In addition to tweets being classified as either positive or negative, the negative sentiment was higher. The emotional responses in tweets were analyzed using the Bing and NRC (National Research Council Canada) dictionaries. | Social |
[17] | To understand the United Nations World Environment Day (WED) program’s impact on the digital public debate. | Tweet collection is performed by using the open-source software T-Hoarder. Subsequently, the open-source software OpenRefine and Orange Data Mining are used to prepare data for further analysis. | Social |
[18] | This study aimed to understand public emotions on topics related to the COVID-19 pandemic in the United Kingdom. | Three advanced deep learning–based models are leveraged: SenticNet 6 for sentiment analysis, SpanEmo for emotion recognition, and combined topic modeling. | Social |
[19] | To make a comparative text mining in 195.649 collected Persian and English tweets about JCPOA (Joint Comprehensive Plan of Action, commonly known as the Iran nuclear deal or Iran deal). | The authors present a rule-based expert system that uses the well-known concept of fingerprint in the judicial sciences to classify tweets. For detecting the unhashtagged tweets, each tweet in question checks itself with the generated fingerprint. | Social |
[20] | This article presents an argument-based opinion-mining model with sentimental data analysis for extracting specific arguments. | The proposed model uses natural language processing techniques, extracting argument words to define decisions. Naive Bayes classification is used for categorizing the results widely under agreed or disagreed. | Social |
[21] | To monitor the COVID-19 disease. | This system is based on a hybridization of tweet mining and deep learning techniques. | Health |
[22] | The authors focused on the potential of using Twitter data to extract relevant topics for the public space and investigate whether the sentiment for these topics can relate to urban design and the improvement of pedestrian space. | The research methodology included five phases: data collection (through Twitter API), data pre-processing (the Python library Natural Language Toolkit (NLTK) was used), data classification (conducted using the sentiment analysis tool Valence Aware Dictionary and sEntiment Reasoner-VADER), data visualization, and data analysis. | Urban Design |
[23] | To detect traffic-related events and road conditions in real time. | Classification algorithms and a custom-trained named entity recognition model (NER) were used to classify and extract contextual information and visualize it on a map to get an overall picture of the traffic conditions in a city. | Traffic |
[24] | To detect and locate floods in real time on a global scale. | Twitter’s real-time streaming (API) is used for data collection. In addition, the TAGGS [25] algorithm is used to extract locations from tweets. | Natural disasters |
[26] | To identify tweets that contain signs of drug abuse and evaluate the usefulness of Twitter in investigating patterns of abuse over time. | Data collection and quantitative and qualitative analyses were performed manually. | Health/Social disasters |
[27] | A methodology is proposed to predict crime in Chicago by combining official historical records with spatiotemporal information embedded in tweets broadcast within the area of interest. | Authors use Twitter-specific linguistic analysis and statistical topic modeling to identify discussion topics across an area of interest automatically. | Criminology |
[9] | The authors proposed an earthquake notification system. | This system is based on a Kalman Filter and a Particle Filter. In addition, the authors designed a tweet classifier based on a support vector machine to detect a seismic event. | Natural disasters |
Municipality | Phone Reports (PR) | Twitter Reports (TR) | Surface (km) | Density PR/km | Density TR/km | Probability from PR | Probability from TR |
---|---|---|---|---|---|---|---|
ALVARO OBREGON | 841 | 43 | 96.17 | 8.744930852 | 0.447124883 | 1.10 | 1.25 |
AZCAPOTZALCO | 317 | 7 | 33.66 | 9.417706477 | 0.207961973 | 1.19 | 0.58 |
BENITO JUAREZ | 410 | 22 | 26.63 | 15.39616973 | 0.826135937 | 1.94 | 2.31 |
COYOACAN | 666 | 57 | 54.40 | 12.24264706 | 1.047794118 | 1.54 | 2.92 |
CUAJIMALPA | 250 | 5 | 80.95 | 3.088326127 | 0.061766523 | 0.39 | 0.17 |
CUAUHTEMOC | 410 | 39 | 32.40 | 12.65432099 | 1.203703704 | 1.59 | 3.36 |
GUSTAVO. A. MADERO | 1862 | 14 | 94.07 | 19.7937706 | 0.148825343 | 2.49 | 0.42 |
IZTACALCO | 207 | 6 | 23.30 | 8.884120172 | 0.25751073 | 1.12 | 0.72 |
IZTAPALAPA | 1022 | 48 | 117.00 | 8.735042735 | 0.41025641 | 1.10 | 1.15 |
MAGDALENA CONTRERAS | 310 | 9 | 74.58 | 4.156610351 | 0.120675784 | 0.52 | 0.34 |
MIGUEL HIDALGO | 352 | 16 | 46.99 | 7.490955522 | 0.340497978 | 0.94 | 0.95 |
MILPA ALPA | 62 | 0 | 228.41 | 0.271441706 | 0 | 0.03 | 0.00 |
TLAHUAC | 137 | 10 | 85.34 | 1.605343333 | 0.117178345 | 0.20 | 0.33 |
TLALPAN | 849 | 65 | 312.00 | 2.721153846 | 0.208333333 | 0.34 | 0.58 |
VENUSTIANO CARRANZA | 237 | 9 | 33.40 | 7.095808383 | 0.269461078 | 0.89 | 0.75 |
XOCHIMILCO | 576 | 8 | 122.00 | 4.721311475 | 0.06557377 | 0.59 | 0.18 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Jiménez-Cabas, J.; Torres, L.; Lozoya-Santos, J.d.J. Twitter Data Mining for the Diagnosis of Leaks in Drinking Water Distribution Networks. Sustainability 2023, 15, 5113. https://doi.org/10.3390/su15065113
Jiménez-Cabas J, Torres L, Lozoya-Santos JdJ. Twitter Data Mining for the Diagnosis of Leaks in Drinking Water Distribution Networks. Sustainability. 2023; 15(6):5113. https://doi.org/10.3390/su15065113
Chicago/Turabian StyleJiménez-Cabas, Javier, Lizeth Torres, and Jorge de J. Lozoya-Santos. 2023. "Twitter Data Mining for the Diagnosis of Leaks in Drinking Water Distribution Networks" Sustainability 15, no. 6: 5113. https://doi.org/10.3390/su15065113