Mihaela Dinsoreanu

Universitatea Tehnica Cluj-Napoca | UT Cluj · Department of Computer Science

Clustering algorithms are essential in data analysis, but evaluating their performance is challenging when the true labels are not available, especially for non-convex clusters. Traditional performance evaluation metrics struggle to identify clustering quality, often assigning higher scores for linearly separated clusters than the true clusters. We...

The widespread adoption of smart home technologies has resulted in the generation of vast amounts of data related to home appliance usage. This research aims to harness the power of data analytics and pattern identification techniques to extract valuable insights from this data. We present an exploration of temporal dependency analysis applied to h...

Spike sorting is the process of grouping spikes of distinct neurons into their respective clusters. Most frequently, this grouping is performed by relying on the similarity of features extracted from spike shapes. In spite of recent developments, current methods have yet to achieve satisfactory performance and many investigators favour sorting manu...

Unlabelled: Space Breakdown Method (SBM) is a clustering algorithm that was developed specifically for low-dimensional neuronal spike sorting. Cluster overlap and imbalance are common characteristics of neuronal data that produce difficulties for clustering methods. SBM is able to identify overlapping clusters through its design of cluster centre...

Word embeddings are known to encapsulate semantic similarity and have become the preferred representation solution for NLP models. However, they fail to identify the type of semantic relationship, which – in some applications – might be crucial. This paper adapts an existing solution for enhancing word embedding representations such as to better se...

As virtual home assistants are becoming more popular, there is an emerging need for supporting languages other than English. While more wide-spread or popular languages such as Spanish, French or Hindi are already integrated into existing home assistants like Google Home or Alexa, integration of other less-known languages such as Romanian is still...

The interpretability of Graph Convolutional Neural Networks is significantly more challenging than for image based convolutional networks, because graphs do not exhibit clear spatial relations between their nodes (like images do). In this paper we propose an approach for estimating the discriminative power of graph nodes from the model learned by a...

Overlapping clusters and different density clusters are recurrent phenomena of neuronal datasets, because of how neurons fire. We propose a clustering method that is able to identify clusters of arbitrary shapes, having different densities, and potentially overlapped. The Space Breakdown Method (SBM) divides the space into chunks of equal sizes. Ba...

Identifying opinion leaders and assessing their influence is highly relevant in several domains. This chapter presents a comprehensive analysis of state-of-the-art opinion leader detection strategies in social networks and the associated challenges. We classify the approaches into four categories, depending on the types of available and observable...

The Internet of Things is about data, different devices from different places and the connectivity between them. Our goal is to find a way to interact with the devices and their data, according to the customer's requirements, in the IoT context, all this supported by the use of Web services. We present a broker based architecture for service select...

The amount of electronic medical documents is growing rapidly every day. While they carry much information, it becomes more and more difficult to manually process it. Our work represents small steps towards automatic knowledge extraction from medical documents using deep learning and similarity based methods. Our goal here is to identify in an unsu...

Opinion mining has become an important field of text mining with high impact in numerous real-world problems. The limitations most solutions have in case of supervised learning refer to domain dependence: a solution is specifically designed for a particular domain. Our method overcomes such limitations by considering the generic characteristics hid...

In this paper, we address the problem of word auto-completion for free text (e.g. messages, emails, articles, poems, etc.) written in different languages. We focus on improving the user experience by developing a user-oriented model that is able to learn different writing styles, while still providing initial predictions without any user written do...

Purpose – The purpose of this paper is to address the challenge of opinion mining in text documents to perform further analysis such as community detection and consistency control. More specifically, we aim to identify and extract opinions from natural language documents and to represent them in a structured manner to identify communities of opinio...

Considering the wide spectrum of both practical and research applicability, opinion mining has attracted increased attention in recent years. This article focuses on breaking the domain-dependency barrier which occurs in supervised opinion mining strategies by using a semi-supervised approach, which ensures domain independence. Our work devises a g...

In this paper, we propose a method for identifying opinion holders and opinion targets from Romanian texts, using an unsupervised approach. The method discriminates the opinion bearing sentences first. Then, it applies a set of rules for holder and target identification. We propose a series of identification rules in addition to an initial set exis...

The high popularity of modern web is partly due to the increase in the number of content sharing applications. The social tools provided by the content sharing applications allow online users to interact, to express their opinions and to read opinions from other users. However, spammers provide comments which are written intentionally to mislead us...

The paper proposes a solution for document and aspect levels sentiment analysis for unstructured documents written in the Romanian language. The opinion extraction relies on two different approaches for polarity identification. At the aspect level we propose a rule-based approach. For the document level we consider supervised learning techniques, b...

Opinion mining has become an important field of text mining. The limitations in case of supervised learning refer to domain dependence: a solution is highly dependent (if not specifically designed or at least specifically tuned) on a given data set (or at least specific domain). Our method is an attempt to overcome such limitations by considering t...

Sentiment classification is not a new topic but data sources having different characteristics require customized methods to exploit the hidden existing semantic while minimizing the noise and irrelevant information. Twitter represents a huge pool of data having specific features. We propose therefore an unsupervised, domain-independent approach, fo...

Considering the wide spectrum of both practical and research applicability, opinion mining has attracted increased attention in recent years. This article focuses on breaking the domain-dependency issues which occur in supervised opinion mining by using an unsupervised approach. Our work devises a methodology by considering a set of grammar rules f...

In this paper we address the problem of identifying contradictions by opinion mining across documents. Our approach involves opinion extraction and storage by processing natural language documents such as reviews, news etc. and aims the identification of contradictory opinions related to the same target expressed by the same holder or by different...

This paper offers a solution to the problem of detecting contradictions among opinions on the same topic. The opinions are extracted from a large number of unstructured documents and stored in a structured format. Due to the increase in data available for analysis, we focus on providing a storage/retrieval and analysis solution suitable for managin...

This paper presents a system for identifying communities in networks built based on opinions and social data. We show how we can build graphs from opinions and social interactions and how we identify the community structure of these graphs. We handle both types of data: one-dimensional and multidimensional. As community detection method, we use the...

Financial markets have always been one of the most common application areas for a multitude of data mining techniques. Over the years a large number of autonomous prediction systems have been designed. In this paper an approach is proposed that offers a higher degree of control over the prediction process and over the exact market aspects that are...

The paper proposes an improved approach to the problem of sentiment polarity identification. Its main focus is on identifying and extracting the relevant information from natural language texts in order to obtain a set of best predictive features to be used for the classification task. Our approach of determining the polarity of a text consists of...

In this paper we approach the context adaptation problem by defining a self-healing model that uses a policy-driven reinforcement learning mechanism to take run-time decisions. The self-healing property is enforced by monitoring the system’s execution environment for evaluating the degree of fulfilling the context policies in the current context si...

The objective of our work is to identify the most relevant content given unstructured, text-based context. In this respect, we propose a unified model that includes a generic context model and the similarity metrics in order to provide context-sensitive content. The context model relies on the underlying thematic structure of the context by means o...

Nowadays, the volume of information increases exponentially, forcing the corporations to keep their business information distributed under several heterogeneous sources such as relational databases, spread sheets, XML documents and Web pages, and stored under different structures and formats. Integrating heterogeneous sources is recently acknowledg...

Considering the huge amount of available data, the necessity of high-quality recommendation systems is obvious. The complex and diverse definitions of quality make the problem even harder. Although there are several approaches to address the problem of relevant recommendation according to a given context, they still cannot cover all the involved as...

The clustering and sorting behavior of ants, as well as the foraging behavior of birds in nature represented sources of inspiration for designing clustering methods applicable in computer science. This paper investigates how biologically-inspired clustering methods can be adapted to cluster Semantic Web services aiming at the efficiency of the disc...

We propose a model capable of providing context-sensitive content based on the similarity between an analysed context and the recommended content. It relies on the underlying thematic structure of the context by means of lexical and semantic analysis. For the context, we analyse both the static characteristics and dynamic evolution. The model has a...

Inspired from biology, in this paper we propose a hybrid firefly method for selecting the optimal solution in semantic Web service composition. In our approach, the search space of the selection method is represented by an Enhanced Planning Graph structure which encodes all the Web service composition solutions for a given user request. As selectio...

Modern society is characterized by abundance of data, yet lack of (relevant) information. A major challenge consists in selecting valuable information according to specific criteria. Moreover, ranking it and defining relevance according to the context is decisive. We propose a framework to retrieve content according to context. Our approach relies...

This paper presents a method for Web service clustering based on Particle Swarm Optimization aiming at the efficiency of the discovery process. The proposed method clusters services based on the similarity between their semantic descriptions. To evaluate the semantic similarity we have defined a set of metrics which compute the degree of match betw...

This paper presents a Tabu search-based method for selecting the optimal or a near-optimal solution in semantic Web service composition. The proposed method is applied on an Enhanced Planning Graph structure which encodes all the composition solutions that satisfy a user request. The criteria for selecting the optimal solution include the QoS attri...

This paper presents a framework for automatic service composition which combines a composition graph model with an Ant Colony Optimization metaheuristic to find the optimal composition solution. The composition graph model encodes all the possible composition solutions that satisfy a user request. The graph will be further used as the search space...

This paper analyses the impact of biological intelligence on the problem of selecting the optimal solution in Web service composition. Thus, we propose two selection methods, one inspired by the behaviour of bees searching for food and another one inspired by the behaviour of cuckoos searching for the nests where to lay eggs. The methods use a comp...

In this paper we present a bee-inspired method for selecting the optimal composition solution. The proposed method uses a composition graph model and a matrix of semantic links to search for the optimal composition solution. For improving the performance of the traditional bee colony optimization algorithm a 1-OPT heuristic is defined. This makes t...

In this paper we approach the context adaptation problem by defining a self-healing model that uses a policy-driven reinforcement learning mechanism to take run-time decisions. The self-healing property is enforced by monitoring the system's execution environment for evaluating the degree of fulfilling the context policies in the current context si...

This paper presents ArhiNet, an integrated system for generating, processing and querying semantically enhanced archival eContent related to the middle age history of Transylvania, targeting cultural heritage preservation. The main workflows of the system aim to perform semantically enhanced eContent generation and knowledge acquisition, as well as...

This paper presents a technique for semantic Web service composition inspired by the behavior of ants. The proposed technique combines a service composition graph model with the ant colony optimization met heuristic to select the optimal composition solution. In our approach, we have considered as selection criteria the QoS attributes of the servic...

This paper presents an immune-inspired technique for optimizing a server energy consumption. The proposed technique is similar with an artificial immune system associated to a server, aiming to detect non-optimal server energy consumption states and to take the appropriate actions that would bring the server into an optimal state. The optimization...

In this paper we approach the context management problem by defining a self-healing algorithm that uses a policy-driven reinforcement learning mechanism to take run-time decisions. The situation calculus and information system theories are used to define and formalize self-healing concepts such as context situation entropy and equivalent context si...

This paper presents a new technique for semantic Web service composition inspired by the swarm behavior. The proposed technique combines a service composition graph model with a hybrid particle swarm optimization algorithm to identify the optimal composition solution. The following criteria are considered to select the optimal composition solution:...

This paper presents a method inspired by ants behavior for creating clusters of semantic Web services. The clustering method groups services according to their semantic similarity. Services are grouped in the same cluster if they provide similar functionality and their input and output parameters are annotated with similar ontology concepts. We pro...

This paper introduces a self-configuring middleware that manages the processes of context information acquisition and representation from smart closed environments, targeting the development of context aware applications. The envi-ronment context information is modeled using three sets: context resources, context actors and context policies. The co...

This paper presents a self-adapting algorithm that can automatically detect the changes in a system execution context and decide how the system should react. The self-adapting algorithm is characterized by a closed feedback loop with four phases: monitoring, analyzing, planning and execution. The monitoring phase uses the RAP (Resources, Actions, P...

This paper presents an ant-inspired method for clustering semantic Web services. The method considers the degree of semantic similarity between services as the main clustering criterion. To measure the semantic similarity between two services we propose a matching method and a set of metrics. The proposed metrics evaluate the degree of match betwee...

This paper presents a new approach for the automatic composition of semantic Web services based on the AI planning graph technique. In the context of Web service composition we have extended the planning graph with the new concepts of service cluster and semantic similarity link and have adapted and enhanced an immune-inspired algorithm that ranks...

This paper approaches the use of both context and semantic information in the information retrieval process with the goal of developing context-based semantically enhanced information retrieval systems. To achieve our objective we have identified, defined and formalized three distinct types of context information relevant for an information retriev...

This paper presents a new Web service composition method which combines the AI planning graph technique with an immune-inspired algorithm to find the optimal composition solution. Simultaneously with the planning graph onstruction, a matrix of semantic links is built to storethe semantic links established between the services on different layers of...

Software reuse is probably the most elusive promise of OOP but achieving effective reuse requires a significant up-front effort, strict discipline and careful planning while its results are difficult to measure. The lack of clear measurements often hinder a company to measure the degree of success in applying systematic reuse and estimate the organ...

This paper proposes a generic policy based self-management model that can be used to automatically detect and repair the problems appeared during the context adaptation processes. To successfully capture and evaluate the dynamic rules that govern the context aware adaptation processes we have defined an generic context policy representation model a...

This paper presents an immune-inspired algorithm applied in the context of Web service composition to select the optimal composition solution. Our approach models Web service composition as a multi-layered process which creates a planning-graph structure along with a matrix of semantic links. We have enhanced the classical planning graph with the n...

In this paper we propose a hierarchical data model for representing ontologies. The model is suited for efficiently handling hierarchical relationships, frequently found in the ontology structure. Our hierarchical data model defines a generic representation using hierarchies that can be persisted in a general purpose records structure. We introduce...

This paper proposes a generic self-healing algorithm that automatically detects, diagnoses and repairs the problems appeared during the context aware systems' adaptation processes. The set, situation calculus and information system theories are used to define and formalize the concepts of context situation entropy and equivalent context situations...

This paper addresses two fundamental research problems in the domain of context aware autonomic systems: the development of a generic context model that can be used to represent general purpose contexts in a system interpretable way and the autonomic context model management. The proposed context model uses two equivalent and synchronized ways of r...

This paper proposes a solution for preserving the cultural heritage by performing knowledge acquisition from historical documents. We developed a system that gathers knowledge by processing the content of historical documents to enable knowledge retrieval as response to ontologically-guided queries. Knowledge acquisition, one of the main workflows...

This paper addresses the problem of creating, processing and querying semantically enhanced eContent from archives and digital libraries. We present an analysis of the archival domain, resulting in the creation of an archival domain model and of a domain ontology core. Our system adds semantic mark-up to the historical documents content, thus enabl...