Abstract Quality management techniques such as the quality function deployment model can help hospitals assess and improve the quality of their services by integrating the voice of customers. The different quality parameters of this model are usually determined and assessed by experts; nonetheless, obtaining such experts is not always easy or inexpensive. Moreover, in this method, patient opinions are not usually considered directly, although they are the real users of the services and those who can best assess those services. Nevertheless, these opinions are easily accessible today, owing to the development of medical social networks where patients directly convey their opinions about the different services and features of a hospital. Therefore, it is feasible to replace expert knowledge with the information provided by these opinions. Based on this idea, this study proposes a novel fuzzy recommendation model based on the quality function deployment method to rank hospitals depending on patient opinions and preferences regarding hospital services. This model integrates a topic modeling strategy for determining hospital requirements, customer needs, and the relationship between them as well as a sentiment analysis algorithm for assessing customer satisfaction regarding hospital services. To demonstrate the usefulness of the proposed method, several experiments were conducted using patient reviews from real hospitals, and the method was compared against other recommendation models. The results prove that this approach represents a step toward more personalized and effective health care system selection considering patient preferences and opinions.

Evaluating the quality of health care systems usually entails the examination of objective variables (waiting time, patients per doctor, etc.). Nonetheless, other subjective variables can be used thanks to new Internet tools. Many online medical services let users convey their opinions on the offered services. That information is an interesting tool to measure the quality of these services. It describes sentiments about different features using a wide range of adjectives, adverbs and nouns, many times, completely different for each feature. Therefore, it is interesting to assess every feature individually using different scales. This study presents an application of a multi-granular fuzzy linguistic model to represent the opinions about the different features of health care systems with the aim of recommending hospitals according to the user preferences. To test this approach, the opinions from real hospitals have been assessed using different user preferences. The obtained results outperform other state-of-the-art approaches.

INTRODUCTION: Most hospital assessment systems are based on the study of objective statistical variables as well as patient opinions on their experiences with respect to the services offered by each hospital. Nevertheless, studies have indicated that most of these assessment systems fail to detect patient emotions when they are assessing their stays in a hospital. This information is vital to understanding most of the patient reviews, which are very complex and convey several emotions per review. Therefore, this study aimed to address the problem of detecting multiple emotions from patient reviews. METHODS: First, a large set of patient opinions was collected from a website that allowed patients to publish their experiences when visiting hospitals. Second, each opinion was labeled with the corresponding conveyed emotions. Third, a deep learning architecture based on a bidirectional gated recurrent unit with a multichannel convolutional neural network layer was proposed to detect multiple emotions from these reviews. Finally, the hyperparameters of this architecture were fine-tuned and different pretrained word embedding models were configured to test its performance. RESULTS: The results confirmed that our proposed method outperformed other deep learning and machine learning-based algorithms and achieved an average accuracy of 95.82%. Furthermore, the experiments show that clinical-domain word embedding slightly outperforms other general-domain word embeddings, although general-domain embeddings are larger in terms of dimensions. CONCLUSIONS: The combination of the gated recurrent unit and the multichannel convolutional neural network is able to exploit both semantic and syntactic characteristics of patient opinions. The findings of this study identify research gaps related to areas such as opinion-based hospital recommendations, thereby providing future research directions.

The process to assess a hospital performance usually needs the interaction of a lot of experts and patients and is very costly and time-consuming. Nevertheless, the availability of patient opinions on the Internet offers a great opportunity to develop systems that evaluate hospitals based on user feedback. The content of these opinions is very challenging, including information about the hospital services but also stories about their own patients, their families, and personal feelings or beliefs before or after leaving a hospital. Therefore, the task of recommending hospitals according to the quality of their services becomes really complicated. This study describes an application for ranking hospitals based on the user preferences about the different offered services as well as the opinions about them. First, it semiautomatically classifies all predefined hospital aspects, calculates the sentiment orientation, and represents their associated polarity by intuitionistic fuzzy sets. Second, by means of the user preferences towards the different aspects, an aggregation operator, and a multicriteria decision-making algorithm, all hospitals are ranked. To assess this methodology, a large set of reviews about hospitals have been collected. Further, considering all patient ratings about the different hospitals, an algorithm for ranking them is proposed, which develops baselines for comparison. In addition, an interval-valued Pythagorean fuzzy approach has been also implemented to compare the obtained results. These results confirm the soundness of the proposal.

Understanding opinions about products offered by big online providers, for instance, TripAdvisor, is relatively easy because the features assessed by the user about hotels are well-known (food, room, desk, sleep quality, etc.). Nonetheless, in the health domain, many times the user provides free-text reviews which are not clearly focused on a few specific features. The present study proposes a methodology for recommending hospitals according to textual opinions which describe the quality of the offered services. First, it detects hospital aspects, which represent the hospital services, by a Latent Dirichlet Allocation-based approach following the criteria of a quality model called SERVQUAL. The polarity of those aspects is computed and modelled by intuitionistic fuzzy sets. Depending on the user preferences or his/her attitude, the aspects are aggregated to finally rank the alternative hospitals following a Multicriteria Decision Making algorithm (PROMETHEE II). The methodology has been tested using a large collection of free-text reviews on hospitals, which contain information about their offered services, obtaining interesting results.

We have witnessed a flourish of review websites where users can buy many products/services and share their opinions about them. Most of those opinions may be broken down into different sub-opinions on the different aspects describing said products/services. This fact makes more complicated the task of computing the overall polarity about the product/service studied.We are presenting a fuzzy aggregation mechanism to compute the overall sentiment conveyed in a opinion/review taking into account the individual ratings for the different aspects commented by the opinion holder. This proposal has been tested using real data from Yelp dataset obtaining promising results.

Causal sentences are a main part of the medical explanations, providing the causes of diseases or showing the effects of medical treatments. In medicine, causal association is frequently related to time restrictions. So, some drugs must be taken before or after meals, being ‘after’ and ‘before’ temporary constraints. Thus, we conjecture that medical papers include a lot of time causal sentences. Causality involves a transfer of qualities from the cause to the effect, denoted by a directed arrow. An arrow connecting the node cause with the node effect is a causal graph. Causal graphs are an imagery way to show the causal dependencies that a sentence shows using plain text. In this article, we provide several programs to extract time causal sentences from medical Internet resources and to convert the obtained sentences in their equivalent causal graphs, providing an enlightening image of the relations that a text describes, showing the cause-effect links and the temporary constraints affecting their interpretation.

The automatic keyphrases extraction is a useful task for many computational applications in the natural language processing and text mining fields. Several solutions to this problem have been reported, but the obtained results still show low rates of accuracy and performance. In this paper, a new unsupervised method for keyphrase extraction from text documents is proposed. The use of lexical-syntactic patterns is combined with a graph-based topic modeling in this approach. The topic modeling is supported by a semantic analysis process carried out from the fuzzy logic perspective. The method was evaluated with the SemEval-2010 and Inspec datasets and compared with other state-of-the-art proposals.

Within the framework of the Intelligent Data Suite (IDS) that is being developed by the company Prometeus Global Solutions, there is a Sentiment Analysis and Opinion Mining module focused on detecting `dangerous' (to the tool user company) messages on Social Media. This can be useful for sending `early warnings' to alert tool user company analysts to take preventive measures against potentially harmful messages. In this paper, a brief description of IDS features, regarding tweets filtering and classification is firstly presented. Affective Norms for English Words (ANEW) provides a set of normative emotional ratings for a large number of words in English and three emotions (valence, arousal and dominance) measures for each term. It is used as a basis for describing a fuzzy model containing five categories for representing the opinion of a microblogging text (very negative, negative, neutral, positive and very positive). The proposal is implemented and tested on the IDS framework.

Most of the well-known recommender systems are based on a proprietary database and a finite and well-known set of elements to recommend. This article presents a methodology that allows us to make recommendations on broad domains where we do not have our own database of recommendable items. This methodology proposes the definition of the characteristic aspects of the domain treated and the extraction of information based on online content tools. These tools will provide the system with subjective and objective information that will make it possible to characterize the recommended elements and thus, adapt the recommendations to the user's preferences.

Inliers and Outliers are widely studied in the world of databases. In the early 80's, L. A. Zadeh proposed an interesting approach to the concept of `prototype' which contrasts with the classical theories of prototypes from the field of cognitive psychology. This definition has not been used much in fuzzy scientific proposals. Metasearch engines are often designed and used for Information Retrieval tasks because they are cheap and easy to develop, as they get their results from multiple search engines. In this paper, a definition of Inliers and Outliers is presented and related to Zadeh's concept of fuzzy prototype: Inliers with the prototype of the borderline elements and Outliers with the prototype of the bad elements of a dataset. Then, this approach is used for dealing with the User Profiles in a real Scientific Metasearcher with the aim of proposing recommendations, sending warnings and in general enriching the management of these User Profiles. An example and experiments proposal in this real system are also presented.

This paper presents an experiment to show the importance of causal sentences in summaries. Presumably, causal sentences hold relevant information and thus summaries should contain them. We perform an experiment to refute or validate this hypothesis. We have selected 28 medical documents to extract and analyze causal and conditional sentences from medical texts. Once retrieved, classic metrics are used to determine the relevance of the causal content among all the sentences in the document and, so, to evaluate if they are important enough to make a better summary. Finally, a comparison table to explore the results is showed and some conclusions are outlined.

Decision support systems, embedded in modern telemedicine applications, are a tool to improve the skills of general practitioners and patients in decision-making in medicine. Nowadays, one of the more challenging problems in this context is how to diagnose those diseases, whose early clinical signs are often subtle, and many of their common signs and symptoms are similar to other. These “fuzzy diseases,” even they can have distinctive features, are not diagnosable through a concrete clinical test or symptom, and, thus, they are difficult to recognize, especially in their initial phases when they might be mistaken for other similar ones. Then, the diagnosis of a fuzzy disease set is based on the exclusion of symptoms and tests results, due to the similarity between them. In the present article, it is proposed the development of a Clinical Decision Support System framework to diagnose a set of fuzzy diseases, concretely applied to Fibromyalgia and associated syndromes. For this purpose, in this paper a reasoning method that uses theories about conceptual categorization from the psychology, pattern recognition, and Zadeh's prototypes has been designed. Through the use of this model, satisfactory results in the evaluation of patients were obtained.

Currently, data repositories contain a plethora of information in different formats, most of which consists of text. This situation has raised interest in the study of techniques to automate the identification of the most relevant sentences of a document with the goal of generating a text summary. This article presents a technique for extracting the most representative sentences in a document, employing a user-defined criteria. The criteria is learned by the system using an optimization technique and a training document where the user has ranked the sentences according to their relevance. The proposed method has been applied to a five-chapter thesis with good results. At the end of this paper we provide some conclusions as well as ideas for future work.

There are several types of business documents, ranging from brief accounting documents to complex legal agreements. Companies extensively use such documents to communicate, transact business and analyse their productivity. This results in the generation of a large number of documents daily, and small- and medium-sized enterprises are easily overwhelmed by this situation. Given this background, companies require software solutions which provide all of the features required by users for optimal document management, as well as optimising management processes and automating the extraction of relevant information from the documents. Open-source software provides these organizations with low-cost, high-quality software which incorporates an array of advanced features that extend beyond only storage solutions. In this study, we test several computational-intelligence open-source software tools in order to enhance the information-retrieval capabilities in small business document-management systems. We implement a prototype to test these Natural Language Processing (NLP) tools and Machine-Learning techniques in a business environment, with the aim of choosing the best alternative for each process.

This paper shows how some fuzzy logic tecniques applied to a recommender engine can be used in a Learning Object Repository. A Fuzzy Linguistic model based on three dimensions: structural, contextual, personal is proposed. The contextual and personal dimensions are modelled using domain ontologies and a automatically built fuzzy ontology, respectively. The experiment results indicate that the presented approach is useful and warrants further research in recommending and retrieval information.

Nowadays, the impact of e-Learning developments on improving educational activities is becoming more evident. The Portfolio approach has emerged as important alternative to increase the learning process. However, most of tools to support portfolio assessment are far from to come up to the expectations. In this work we propose a fuzzy ontology-based framework to support portfolio assessment. Our approach focuses on portfolio semantic representation and conceptual matching to generate a portfolio evaluation report, which helps teachers in portfolio assessment tasks. The initial experiments results indicate that the approach is useful and warrants further research.

The aim of this paper is to approach causal questions in a medical domain. Causal questions par excellence are what, how and why-questions. The `pyramid of questions' shows this. At the top, why-questions are the prototype of causal questions. Usually why-questions are related to scientific explanations. Although cover law explanation is characteristically of physical sciences, it is less common in biological or medical knowledge. In medicine, laws applied to all cases are rare. It seems that doctors express their knowledge using mechanisms instead of natural laws. In this paper we will approach causal questions with the aim of: (1) answering what-questions as identifying the cause of an effect; (2) answering how-questions as selecting an appropriate part of a mechanism that relates pairs of cause-effect (3) answering why-questions as identifying ultimate causes in the answers of how-questions. In this task, we hypothesize that why-questions are related to scientific explanations in a negative and a positive note: (i) as previously said, scientific explanations in biology are based on mechanisms instead of natural laws; (ii) scientific explanations are generally concerned with deepening, providing explanations as detailed as possible. Thus, we conjecture that answers to why-questions have to find the ultimate causes in a mechanism and link them to the prior cause summarizing the intermediate nodes in order to provide a comprehensible answer. The Mackie's INUS causality offers a theoretical support for this solution.

Software development project management has a poor reputation in terms of avoiding cost and schedule overruns. The cause of this situation is based on the feature of the software development process that is characterized by quickly growing complexity and change. Therefore, there are many uncertainties to define exactly the necessary time to complete a tasks according to the person's performance. In this scenario Soft-Computing techniques may offer new approaches with the aim of helping the participants of the project to manage their time, give priority to their activities and readjust the work to complete satisfactorily the project tasks. This work presents an automatic features extraction process with the aim of defining the elements involved in a software project. This knowledge is represented by means fuzzy sets and fuzzy prototypes. The source of data is the Personal Software Project time recording logs. A preliminary experiment illustrates the feasibility of this approach.

Causality is not only a matter of causal statements, but also of conditional sentences. In conditional statements, causality generally emerges from the entailment relationship between the antecedent and the consequence. This entailment is frequently vague and uncertain in nature. In this article, we present a method of retrieving crisp and imperfect conditional and causal sentences identified by some linguistic patterns. These sentences are pre-processed to obtain both single cause-effect structures and causal chains. The result is displayed automatically in an imperfect causal graph by means of a Java application. The causal graph shows the strenght of the causal link labelling it with a fuzzy quantifier and the intensity of the cause or effect nodes with linguistics hedges. The knowledge base used to provide automatic information based on causal relations was some medical texts, suited for the described process.

Category-based information filtering is ground on the representation of user preferences according to a set of categories of similar items. The use of type 1 fuzzy sets provides a good method to represent categories when only one static interpretation of them is considered. This representation is not enough when documents do not have the same meaning for two different users because there are some degrees of subjectivity. On the other hand, type 2 fuzzy sets have been successfully applied to manage uncertainty more effectively than type-1 fuzzy sets in several environments. This paper presents a method to manage efficiently uncertainties in the filtering process in environments where there is a constant flow of new information (news, e-mail, etc.) and multiple users are involved. The proposed solution is based on the extension of the categories-based filtering method using interval type 2 fuzzy sets for representing each category and the user preferences. Experimental results, that illustrate the feasibility of this approach, are provided.

The focus of this paper is the study of causality in both its crisp and approximate forms. Crisp causality is characterized by some properties and modalities and is related to semantic implications. This paper presents a program that extracts causal and conditional sentences with causal content from several texts. The examples extracted show that, even in scientific texts, causality can be imprecise or imperfect, as shown by the linguistic modifiers or the fuzzy quantifiers embedded in them. Quantum mechanics introduces imprecision in physics, but social sciences are the disciplines that show more circumstantial and imperfect links between cause and effect. In social sciences, there are two theoretical paradigms to understanding imperfect causality: (i) cause as an `ideal type', from Weber, (ii) cause as a `family resemblance predicate', from Anscombe, a follower of Wittgenstein's philosophy. Our work provides two short exemplifications of these paradigms using the causal or conditional sentences retrieved from texts of different genres.

The goal of this paper is to present the architecture of a metasearch engine called Castalia, still under development, which includes several underlying Q&A systems. Usually metasearch engines manage typical search engines like Google or Yahoo, but in this case the encapsulation of Q&A systems proposes new challenges that can be modeled by fuzzy logic apart from the other existing challenges such as the fuzzy modeling of temporal or causal questions.

Existing search engines and question-answering (QA) systems have made possible processing large volumes of textual information. Current work on QA has mainly focused on answering two basic types of questions: factoid and definition questions. However, the capability to synthesize an answer to a query by drawing on bodies of information which reside in various parts of the knowledge base is not among the capabilities of those systems. In this paper, a system oriented to infer query answers from a collection of propositions expressed in natural language is introduced. By means of a specific example, it is outlined how the system proceeds to face those situations. This approach is based on the use of formal constraining relations modeling copulative and comparative sentences. Combining those propositions with others contained in different knowledge bases and applying deduction rules, the desired answer could be obtained.

In this poster, a tool named BUCEFALO is presented. This tool is specially designed to improve the information retrieval tasks in web-based Personal Health Records (PHR). This tool implements semantic and multilingual query expansion techniques and information filtering algorithms in order to help users find the most valuable information about a specific clinical case. The filtering model is based on fuzzy prototypes based filtering, data quality measures, user profiles and healthcare ontologies. The first experimental results illustrate the feasibility of this tool.

The need for an explicit common terminology within Empirical Software Engineering (an ESE-Glossary of terms) was highlighted in the ISERN 2007 meeting [2]. The goal was to define a glossary of terms related to ESE based on an initial glossary published in http://lens-ese.cos.ufrj.br/wikiese. This initial glossary was built manually, based on expert knowledge. However, owing to the dynamic nature of the research works in ESE, this glossary must be dynamically updated with information extracted from the relevant documents in the research domain. Automation is, therefore, mandatory. We propose a text mining technique for the automatic extraction of the most relevant terms used in ESE documents. Our technique also provides the relationships between terms, with the degree of affinity between them. Our approach could, therefore, be useful in the improvement of the initial glossary of terms and in discovering relationships between terms.

This work focuses in the design and implementation of a digital fuzzy controller, that uses a novel tuning technique called simple tuning algorithm (STA) to achieve the desired controllerpsilas response. Improvements over exciting architectures to implement the fuzzy inference systems (FIS) in a field programmable gate array (FPGA) are presented. A methodology that minimizes the test and validation process is given. Experiments were done using a geared direct current (DC) motor; the control goal was to maintain the speed at a given value.

The information retrieval is necessary for the execution of our daily activities. The people use the computers during hours to find information although they are not safe of its existence. The Internet offers an opportunity to provide an almost infinite source with information. The personal computers, on the other hand, also store a great amount of information. The challenges and the characteristics of the organization and information retrieval are treated in this paper in both stadiums, the Web and the Desktop. A study of the investigations is made therefore in the state of the question

In this paper, a novel document filtering model in dynamic web repositories based on fuzzy deformable prototypes is presented. This model is based on fuzzy hierarchical categorization of documents. It defines an easy process to deal with the incoming documents and an efficient method to update their structure. The process is performed comparing the fuzzy prototypes of document cluster with the available information about documents contents. It exploits conceptual-based filtering criteria and category-based filtering techniques to deliver to the user an intelligent structure of the documents. Since filtering is a dynamic process, the knowledge base can update the hierarchy of existing documents. The clusters hierarchy can be easily and efficiently updated when new documents income on the repository by means of an inference method which is based on fuzzy deformable prototypes.

Natural languages (NLs) are basically systems for describing perceptions, which are intrinsically imprecise. Soft computing approach to NL-Computation considers that new tools such as generalized constraints and prototypical forms (protoforms) should be employed to deal with such semantic imprecision. If a NL proposition could be expressed in terms of a generalized constraint, it could be considered precise, at least, in some degree. In this paper, NL adjectives are considered as constraints. Therefore, generalized constraints are proposed to specify the attribute values associated with nouns by means of adjectives. Then, after recognizing a noun phrase in a sentence, it could be transformed into a generalized constraint. This way, a lot of data could be retrieved from documents in NL, and used to deduce new information. Protoforms allow describing the deep semantic structure of propositions. They could be used to represent noun phrases, involving adjectives and nouns, as symbolic expressions.

Abstract This article presents a semantic‐based Web retrieval system that is capable of retrieving the Web pages that are conceptually related to the implicit concepts of the query. The concept of “concept” is managed from a fuzzy point of view by means of semantic areas. In this context, the proposed system improves most search engines that are based on matching words. The key of the system is to use a new version of the Fuzzy Interrelations and Synonymy‐Based Concept Representation Model (FIS‐CRM) to extract and represent the concepts contained in both the Web pages and the user query. This model, which was integrated into other tools such as the Fuzzy Interrelations and Synonymy based Searcher (FISS) metasearcher and the fz‐mail system, considers the fuzzy synonymy and the fuzzy generality interrelations as a means of representing word interrelations (stored in a fuzzy synonymy dictionary and ontologies). The new version of the model, which is based on the study of the cooccurrences of synonyms, integrates a soft method for disambiguating word senses. This method also considers the context of the word to be disambiguated and the thematic ontologies and sets of synonyms stored in the dictionary.