Lacramioara Dranca
Research Subjects
Lacramioara Dranca has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Arantza Illarramendi
University of the Basque Country
6 shared publications A. Goni
3 shared publicationsA. Burgos
2 shared publicationsAlfredo Burgos
University of the Basque Country, Paseo Manuel Lardizabal 1, 20018 Donostia-San Sebastian, Spain
2 shared publications- Alfredo Goñi
University of the Basque Country
2 shared publications - Fernando Bobillo
Department of Computer Science and Systems Engineering, University of Zaragoza, C. María de Luna, 1, 50018 Zaragoza, Spain
2 shared publications
Latest Publications
- Open Access
- 29 May 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1764367
LLMs for industrial databases: an agro-food production plant use case
Introduction: Extracting value from industrial time-series databases such as InfluxDB 2.0 requires expertise in specialized query languages (InfluxQL, Flux) that domain experts typically lack, and no labeled corpora exist for translating natural language into them. Methods: We present a modular system that translates questions written in Spanish into executable InfluxQL and Flux queries over an InfluxDB 2.0 instance deployed in a real agro-food production plant operating under a self-consumption energy scheme. It comprises a semantic layer implemented as a knowledge graph encoding the InfluxDB schema and its correspondence with the plant's physical components; a hierarchical entity-linking module; a lightweight language model fine-tuned for NL-to-query generation; and a query validation and sanitization module. To obtain training data, we develop a fully automated synthetic dataset distillation pipeline that uses a large teacher model with contextual retrieval from the official InfluxDB 2.0 documentation; each candidate undergoes parsing, AST extraction, and semantic checking against the knowledge graph, and only validated samples are retained. The corpus fine-tunes compact Small Language Models through domain-level conditioning followed by task-specific instruction tuning. Results: Performance is evaluated on a manually curated suite of 56 Spanish queries, evenly distributed across seven operationally relevant query families. Compact models reliably generate syntactically valid queries, but functional evaluation against the live instance reveals a three-layer pattern-high parser validity, moderate execution success, and lower result correctness-that locates the residual gap at the semantic layer. Discussion: The contributions are a complete NL-to-InfluxDB pipeline grounded in an explicit semantic representation of a real industrial schema; a documentation-driven synthetic data generation process with automatic syntactic and semantic verification; and a parameter-efficient fine-tuning strategy enabling query generation with lightweight models suitable for resource-constrained environments.
- 29 May 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1764367
Open AccessLLMs for industrial databases: an agro-food production plant use case
Introduction: Extracting value from industrial time-series databases such as InfluxDB 2.0 requires expertise in specialized query languages (InfluxQL, Flux) that domain experts typically lack, and no labeled corpora exist for translating natural language into them. Methods: We present a modular system that translates questions written in Spanish into executable InfluxQL and Flux queries over an InfluxDB 2.0 instance deployed in a real agro-food production plant operating under a self-consumption energy scheme. It comprises a semantic layer implemented as a knowledge graph encoding the InfluxDB schema and its correspondence with the plant's physical components; a hierarchical entity-linking module; a lightweight language model fine-tuned for NL-to-query generation; and a query validation and sanitization module. To obtain training data, we develop a fully automated synthetic dataset distillation pipeline that uses a large teacher model with contextual retrieval from the official InfluxDB 2.0 documentation; each candidate undergoes parsing, AST extraction, and semantic checking against the knowledge graph, and only validated samples are retained. The corpus fine-tunes compact Small Language Models through domain-level conditioning followed by task-specific instruction tuning. Results: Performance is evaluated on a manually curated suite of 56 Spanish queries, evenly distributed across seven operationally relevant query families. Compact models reliably generate syntactically valid queries, but functional evaluation against the live instance reveals a three-layer pattern-high parser validity, moderate execution success, and lower result correctness-that locates the residual gap at the semantic layer. Discussion: The contributions are a complete NL-to-InfluxDB pipeline grounded in an explicit semantic representation of a real industrial schema; a documentation-driven synthetic data generation process with automatic syntactic and semantic verification; and a parameter-efficient fine-tuning strategy enabling query generation with lightweight models suitable for resource-constrained environments.
- 1 October 2019
- Published by Elsevier in International Journal of Approximate Reasoning
- Vol. 113, 354-371
Gait recognition using fuzzy ontologies and Kinect sensor data
- 1 October 2019
- Published by Elsevier in International Journal of Approximate Reasoning
- Vol. 113, 354-371
Gait recognition using fuzzy ontologies and Kinect sensor data
- Open Access
- 10 December 2018
- Published by Springer Nature in BMC Bioinformatics
- Vol. 19 (1) , 471
Using Kinect to classify Parkinson’s disease stages related to severity of gait impairment
Parkinson’s Disease (PD) is a chronic neurodegenerative disease associated with motor problems such as gait impairment. Different systems based on 3D cameras, accelerometers or gyroscopes have been used in related works in order to study gait disturbances in PD. Kinect Ⓡ has also been used to build these kinds of systems, but contradictory results have been reported: some works conclude that Kinect does not provide an accurate method of measuring gait kinematics variables, but others, on the contrary, report good accuracy results. In this work, we have built a Kinect-based system that can distinguish between different PD stages, and have performed a clinical study with 30 patients suffering from PD belonging to three groups: early PD patients without axial impairment, more evolved PD patients with higher gait impairment but without Freezing of Gait (FoG), and patients with advanced PD and FoG. Those patients were recorded by two Kinect devices when they were walking in a hospital corridor. The datasets obtained from the Kinect were preprocessed, 115 features identified, some methods were applied to select the relevant features (correlation based feature selection, information gain, and consistency subset evaluation), and different classification methods (decision trees, Bayesian networks, neural networks and K-nearest neighbours classifiers) were evaluated with the goal of finding the most accurate method for PD stage classification. The classifier that provided the best results is a particular case of a Bayesian Network classifier (similar to a Naïve Bayesian classifier) built from a set of 7 relevant features selected by the correlation-based on feature selection method. The accuracy obtained for that classifier using 10-fold cross validation is 93.40%. The relevant features are related to left shin angles, left humerus angles, frontal and lateral bents, left forearm angles and the number of steps during spin. In this paper, it is shown that using Kinect is adequate to build a inexpensive and comfortable system that classifies PD into three different stages related to FoG. Compared to the results of previous works, the obtained accuracy (93.40%) can be considered high. The relevant features for the classifier are: a) movement and position of the left arm, b) trunk position for slightly displaced walking sequences, and c) left shin angle, for straight walking sequences. However, we have obtained a better accuracy (96.23%) for a classifier that only uses features extracted from slightly displaced walking steps and spin walking steps. Finally, the obtained set of relevant features may lead to new rehabilitation therapies for PD patients with gait problems.
- 10 December 2018
- Published by Springer Nature in BMC Bioinformatics
- Vol. 19 (1) , 471
Open AccessUsing Kinect to classify Parkinson’s disease stages related to severity of gait impairment
Parkinson’s Disease (PD) is a chronic neurodegenerative disease associated with motor problems such as gait impairment. Different systems based on 3D cameras, accelerometers or gyroscopes have been used in related works in order to study gait disturbances in PD. Kinect Ⓡ has also been used to build these kinds of systems, but contradictory results have been reported: some works conclude that Kinect does not provide an accurate method of measuring gait kinematics variables, but others, on the contrary, report good accuracy results. In this work, we have built a Kinect-based system that can distinguish between different PD stages, and have performed a clinical study with 30 patients suffering from PD belonging to three groups: early PD patients without axial impairment, more evolved PD patients with higher gait impairment but without Freezing of Gait (FoG), and patients with advanced PD and FoG. Those patients were recorded by two Kinect devices when they were walking in a hospital corridor. The datasets obtained from the Kinect were preprocessed, 115 features identified, some methods were applied to select the relevant features (correlation based feature selection, information gain, and consistency subset evaluation), and different classification methods (decision trees, Bayesian networks, neural networks and K-nearest neighbours classifiers) were evaluated with the goal of finding the most accurate method for PD stage classification. The classifier that provided the best results is a particular case of a Bayesian Network classifier (similar to a Naïve Bayesian classifier) built from a set of 7 relevant features selected by the correlation-based on feature selection method. The accuracy obtained for that classifier using 10-fold cross validation is 93.40%. The relevant features are related to left shin angles, left humerus angles, frontal and lateral bents, left forearm angles and the number of steps during spin. In this paper, it is shown that using Kinect is adequate to build a inexpensive and comfortable system that classifies PD into three different stages related to FoG. Compared to the results of previous works, the obtained accuracy (93.40%) can be considered high. The relevant features for the classifier are: a) movement and position of the left arm, b) trunk position for slightly displaced walking sequences, and c) left shin angle, for straight walking sequences. However, we have obtained a better accuracy (96.23%) for a classifier that only uses features extracted from slightly displaced walking steps and spin walking steps. Finally, the obtained set of relevant features may lead to new rehabilitation therapies for PD patients with gait problems.
- 14 September 2017
- Published by Springer Nature
- p. 397-404
A Fuzzy Ontology-Based System for Gait Recognition Using Kinect Sensor
- 14 September 2017
- Published by Springer Nature
- p. 397-404
A Fuzzy Ontology-Based System for Gait Recognition Using Kinect Sensor
- 3 August 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability
- Vol. 230 (5) , 485-501
A model-driven approach to survivability requirement assessment for critical systems
Survivability is a crucial property for those systems – such as critical infrastructures or military Command and Control Information Systems – that provide essential services, since the latter must be operational even when the system is compromised due to attack or faults. This article proposes a model-driven method and a tool –MASDES– to assess the survivability requirements of critical systems. The method exploits the use of (1) (mis)use case technique and UML profiling for the specification of the survivability requirements and (2) Petri nets and model checking techniques for the requirement assessment. A survivability assessment model is obtained from an improved specification of misuse cases, which encompasses essential services, threats and survivability strategies. The survivability assessment model is then converted into a Petri net model for verifying survivability properties through model checking. The MASDES tool has been developed within the Eclipse workbench and relies on Papyrus tool for UML. It consists of a set of plug-ins that enable (1) to create a survivability system view using UML and profiling techniques and (2) to verify survivability properties. In particular, the tool performs model transformations in two steps. First, a model-to-model transformation generates, from the survivability view, a Petri net model and properties to be checked in a tool-independent format. Second, model-to-text transformations produce the Petri net specifications for the model checkers. A military Command and Control Information Systems has been used as a case study to apply the method and to evaluate the MASDES tool, within an iterative-incremental software development process.
- 3 August 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability
- Vol. 230 (5) , 485-501
A model-driven approach to survivability requirement assessment for critical systems
Survivability is a crucial property for those systems – such as critical infrastructures or military Command and Control Information Systems – that provide essential services, since the latter must be operational even when the system is compromised due to attack or faults. This article proposes a model-driven method and a tool –MASDES– to assess the survivability requirements of critical systems. The method exploits the use of (1) (mis)use case technique and UML profiling for the specification of the survivability requirements and (2) Petri nets and model checking techniques for the requirement assessment. A survivability assessment model is obtained from an improved specification of misuse cases, which encompasses essential services, threats and survivability strategies. The survivability assessment model is then converted into a Petri net model for verifying survivability properties through model checking. The MASDES tool has been developed within the Eclipse workbench and relies on Papyrus tool for UML. It consists of a set of plug-ins that enable (1) to create a survivability system view using UML and profiling techniques and (2) to verify survivability properties. In particular, the tool performs model transformations in two steps. First, a model-to-model transformation generates, from the survivability view, a Petri net model and properties to be checked in a tool-independent format. Second, model-to-text transformations produce the Petri net specifications for the model checkers. A military Command and Control Information Systems has been used as a case study to apply the method and to evaluate the MASDES tool, within an iterative-incremental software development process.
- 1 February 2015
- Published by Springer Nature
- p. 86-100
Modelling and Verification of Survivability Requirements for Critical Systems
- 1 February 2015
- Published by Springer Nature
- p. 86-100
Modelling and Verification of Survivability Requirements for Critical Systems
- 1 November 2009
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 96 (2) , 141-157
Architecture, cost-model and customization of real-time monitoring systems based on mobile biological sensor data-streams
- 1 November 2009
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 96 (2) , 141-157
Architecture, cost-model and customization of real-time monitoring systems based on mobile biological sensor data-streams
- 21 August 2009
- Published by IOP Publishing in Physiological Measurement
- Vol. 30 (9) , 983-998
Real-time detection of transient cardiac ischemic episodes from ECG signals
We propose a new algorithm to detect and classify transient cardiac ischemia episodes, designed with the goal of providing a real-time execution without penalizing the classifier accuracy much. The algorithm is based on a novel mixture of time-domain analysis and machine learning techniques, specifically bagging of decision trees, and it has been developed using a well-recognized and freely distributed database, namely the long-term ST database. The ST episode detection sensitivity/positive predictivity using the annotation protocol A for this database is 68.26%/74.91%. The sensitivity result increases until 93.97% for the most dangerous episodes in terms of duration and magnitude (annotated according to protocol C). The test of the algorithm over the freely distributed part of the European Society of Cardiology database has shown results of sensitivity and positive predictivity of 83.33% and 77.31%, respectively. Those results are close to the results obtained by related works that present approaches to detect ischemia episodes off-line, which is remarkable if we take into account that in our real-time approach, less information is available during the classification process.
- 21 August 2009
- Published by IOP Publishing in Physiological Measurement
- Vol. 30 (9) , 983-998
Real-time detection of transient cardiac ischemic episodes from ECG signals
We propose a new algorithm to detect and classify transient cardiac ischemia episodes, designed with the goal of providing a real-time execution without penalizing the classifier accuracy much. The algorithm is based on a novel mixture of time-domain analysis and machine learning techniques, specifically bagging of decision trees, and it has been developed using a well-recognized and freely distributed database, namely the long-term ST database. The ST episode detection sensitivity/positive predictivity using the annotation protocol A for this database is 68.26%/74.91%. The sensitivity result increases until 93.97% for the most dangerous episodes in terms of duration and magnitude (annotated according to protocol C). The test of the algorithm over the freely distributed part of the European Society of Cardiology database has shown results of sensitivity and positive predictivity of 83.33% and 77.31%, respectively. Those results are close to the results obtained by related works that present approaches to detect ischemia episodes off-line, which is remarkable if we take into account that in our real-time approach, less information is available during the classification process.
- 1 April 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 97-105
Real-Time Monitoring of Mobile Biological Sensor Data-Streams: Architecture and Cost-Model
In this paper an architecture useful to build systems that monitor data streams generated by biological sensors attached to mobile users is presented. In order to illustrate the suitability and applicability of our proposal, we introduce two particular real systems: an anywhere and at any time monitoring system of heart arrhythmias and a portable and light system that monitor sleep apneas. We also propose a cost model that takes into account three aspects related to the system efficiency: selection of the optimal granularity, that is, the selection of the size of the input data stream package that has to be acquired in order to start a new processing cycle; the possible use of compression techniques to store and send the acquired input data stream and; finally, the performance of a local analysis versus a remote one.
- 1 April 2008
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 97-105
Real-Time Monitoring of Mobile Biological Sensor Data-Streams: Architecture and Cost-Model
In this paper an architecture useful to build systems that monitor data streams generated by biological sensors attached to mobile users is presented. In order to illustrate the suitability and applicability of our proposal, we introduce two particular real systems: an anywhere and at any time monitoring system of heart arrhythmias and a portable and light system that monitor sleep apneas. We also propose a cost model that takes into account three aspects related to the system efficiency: selection of the optimal granularity, that is, the selection of the size of the input data stream package that has to be acquired in order to start a new processing cycle; the possible use of compression techniques to store and send the acquired input data stream and; finally, the performance of a local analysis versus a remote one.
- 1 January 2006
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 719-726
Using DecisionTrees for Real-Time Ischemia Detection
In this paper we investigate if techniques based on decision trees are also useful to classify ischemia. In order to do that, we have based on a previous own algorithm that is able of detecting ST-segment episodes, can be executed in real time and is lightweight enough to be implemented on mobile devices such as PDAs. Three main steps are performed by that algorithm: 1) signal preprocessing that extracts important features from the ECG signal 2) detection of suspect ST segment events and 3) rejection of irrelevant ST segment events. We have found that techniques based on decision trees are interesting for the last step in order to obtain a set of rules that reject some of the suspect ST segment events detected in the second step. The freely available part of the LTST database has been used to develop the detector and the rest of the same database has been used for validation purposes. The sensitivity of the detector over all the records of the LTST database is 89.89% for episodes annotated according to C protocol and the positive predictivity is 70.03% for episodes annotated according to A protocol. Those results improve our previously obtained ones and we think that they are comparable to other results that appear in the literature. However, it has to be noticed that we have not found works that detect ischemia in real time and show validation results over the LTST database
- 1 January 2006
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 719-726
Using DecisionTrees for Real-Time Ischemia Detection
In this paper we investigate if techniques based on decision trees are also useful to classify ischemia. In order to do that, we have based on a previous own algorithm that is able of detecting ST-segment episodes, can be executed in real time and is lightweight enough to be implemented on mobile devices such as PDAs. Three main steps are performed by that algorithm: 1) signal preprocessing that extracts important features from the ECG signal 2) detection of suspect ST segment events and 3) rejection of irrelevant ST segment events. We have found that techniques based on decision trees are interesting for the last step in order to obtain a set of rules that reject some of the suspect ST segment events detected in the second step. The freely available part of the LTST database has been used to develop the detector and the rest of the same database has been used for validation purposes. The sensitivity of the detector over all the records of the LTST database is 89.89% for episodes annotated according to C protocol and the positive predictivity is 70.03% for episodes annotated according to A protocol. Those results improve our previously obtained ones and we think that they are comparable to other results that appear in the literature. However, it has to be noticed that we have not found works that detect ischemia in real time and show validation results over the LTST database