I worked in the PEARL project (Wellcome-funded project) at Kings College of London from October 2011 to April 2012. The objective of this project was to work on creating a national standardised solution to the management of cohort studies. It aimed to create a grid-based framework concerned with addressing issues associated with data linkage and extraction and longitudinal follow-up of patients across their healthcare institutions and their respective Electronic Health Records (EHRs).
As part of the Pearl project I focused on a framework for data integration; in particular I was interested on semantic record linkage. This part of the project attempted to link records from different heterogeneous sources automatically. The linkage was a semantic linkage which enriched the records with semantic information which can be used during the exploitation phase. An evaluation was carried out in terms of standard measures from the Information Retrieval Community. In particular, for my case of study, I was interested in records related to Asthma as I wanted to have a system which pull out automatically patients who have presented any asthma episode. Finally, I believe my research work, in the Pearl project could benefit from the use of ALSPAC as ALSPAC is the most comprehensive repository in Europe of longitudinal studies.
My outputs in the Pearl project are as follows:
I have developed and tested a method which automatically find similar medical records in databases / datasets. The method consists of two phases namely homogenization and linking. Homogenization performs filling-in of missing data using the obtained predicted values. The linking step is performed using similarity measures between the medical records. The method has been tested using in a first instance, immunisation datasets which are part of the GPRD database (General Practice Research Database). My paper produced in the Pearl project is titled “Methodology for Record Linkage: Medical Domain Case of Study”. IJKSR Journal 6(4), pp. 18-35, 2015. Link to the more details of this paper: http://mariavargas-vera.com/papers/refereed-journals-book-chapters/
Other papers that I have produced in the PEARL project are listed as follows:
Data Integration Framework: A Children and Parents Cohort Case Study. IJKSR Journal 7(1), pp. 99-112, 2016.
An Agent Based Real-time Clinical Trial Recruitment System.Workshop on Agents applied in Health Care A2HC in conjunction with the Eleventh International Conference on Autonomous Agents and Multiagent Systems (AAMAS-2012)., June 4-8; Valencia Spain, 2012.
Funded By: Wellcome-Trust – (2011-2013)
I worked in the Virtual Computing Mphil/PhD Project, which was a Computing Department funded project at the Open University UK; (2008-2010). This project uses Second Life(Linden labs) as a platform for a E-Learning environment. The main users of such system are external MPhil/PhD students which joined the Open University as a long distance learning institution.
- The Virtual MPhil: a part-time online research degree for Computing (phase 2) was ready on October 2009. The main objective of this “program of studies” as described by the Team of the MPhil programme is shown as follows: “This programme was designed to provide a supportive, professional-oriented structure within which advanced IT knowledge and related research skills and methods can be acquired. The new degree was available from October 2009 and is offered for part- time study at a distance, supported by a range of technologies for online study, supervision and assessment. The programme is particularly relevant to students who intend to go on to study for a PhD, as well as to IT professionals seeking to enhance knowledge and skills, particularly when research and scholarship are a significant part of their work. In being part-time and at a distance, the programme is especially suited to the needs of the working professional. From an academic perspective, the ultimate aim of the project is the creation of a vibrant online research community for Computing, supported by a rich and flexible Virtual Learning Environment (VLE): rich to encompass the many processes of research; flexible to meet different user needs, preferences and attitudes to technology.” The technologies offered in the Virtua MPhil comprises DeepThink (our OU immersive 3-D campus), Moodle, Elluminate, MyStuff (EPortfolios), Skype and Ning. Synchronous and asynchronous internet and web technologies are the backbone for an integrated learning environment to support the part time MPhil Computing program at the Open University.
- OU 3D-Virtual Campus Pictures
- Virtual MPhil program of studies in Computing
- During my participation of the project Team_MPhil. After I finished my contract Team
- Second life campus SLurl
- Moddle site (permission required)
- OU 3D-Virtual Campus video
- Deep|Think article in the OU Magazine OpenHouse, Isuue number 425, 2009. The glamorous avatar Cielo – “Maria’s avatar” – in the center; printed on the 24th November 2009 read more
- The 3-D Virtual Campus (Deep|Think) is currently in use with the first generation of part-time students who started on October 2009. Preliminary evaluation has been carried out, however, more evaluations are planned to take place in the future.
- The OU Virtual Campus to support the Computing MPhil/PhD program at the OU (Phase 1) was ready on 2008!. The moodle “Computing Mphil Course” and others synchronous and asynchronous educational had been added to the Eduational Environment. We decided to create an unusual design for the OU Virtual Campus like the one shown in the pictures (see link above). This was motivated by the idea “open spaces to our students”. In addition, we also wanted to inject a sense of freedom in the design!. We expect that this Virtual Campus will provide with Learning Spaces where students can meet and learn collaboratively (i.e. share/exchange ideas on their MPhil/PhD studies) besides to enjoy a peaceful/beautiful scenery.
Funded By: Computing Department, Open University, UK – (2008-2010)
In this project we performed named entity recognition and extraction of relations. Named entity recognition has been studied largely in the Information Extraction community as it is one of the first steps in the construction of an Information Extraction System. However, to extract only names without contextual information is not sufficient if we want to be able to describe facts encountered in documents. Therefore, there is a need in the Semantic Web community for extracting relations between entities. This task can be accomplished using relational learning algorithms embedded in an Information Extraction framework based on Relational Learning. In particular, in the context of “OntoLearning” we have extended two relational learning frameworks RAPIER and FOIL. Our proposed extended frameworks are equipped with DSSim (short for Dempster-Shafer Similarity) our similarity service. Both extended frameworks were tested using an electronic newsletter consisting of news articles describing activities or events happening in an academic institution. In short, during the project we worked towards the Automatic Construction of Ontologies from Text. We used axiomatic approaches to extract non-taxonomic relations from text and Information Extraction techniques for instances classification.
Funded By: Computing Department, Open University, UK – (2008-2010)
I participated in the ENIRAF project from September 2007 to November 2007. Enhanced Information Retrieval and Filtering for Analytical Systems. Founded EU project Framework 6; grant: €52,778 (http://eniraf.mis.ae.poznan.pl/). Financed by European Commission under contract no. MTKD-CT-2004-509766. Focus of the research in this project is on ontologies, web services and the Semantic Web for information filtering. Link to the project enIRaF
In this project our main objective was to analyze the main challenges of the Semantic Web Services and come out with a solution to these challenges. In addition, also in the ENIRAF project we worked on learning ontologies from text. In particular, we concentrated in the extraction of relations from text. The latest relies on the use of Natural Language Processing and an axiomatic approach for the extraction of relations.
Funded By: European Commission, Brussel, Belgium – (2007-2007)
Knowledge Web is a FP6 Network of Excellence that aims to support the transition of Ontology technology from Academia to Industry. The current consortium is integrated by 18 participants including leading partners in Semantic Web, Multimedia, Human Language Technology, Workflow and Agents. European Commission funded project (Framework 6); Grant: €6.7M (total); €375k (OU share).
My work during October 2006 -November 2006 was mainly concerned with methodologies for evaluating MAGPIE (semantic browser).
Funded By: European Commission, Brussel, Belgium – (2006-2006)
AKT was a multi-million pound (EPSRC funded project Grant: £7.5M (total); £1.3M (OU share)) , six-year collaboration between internationally recognized research groups at the Universities of Southampton, Aberdeen, Edinburgh, Sheffield and the Open University. AKT aimed to tackle fundamental problems associated with the management of knowledge & Semantic Web. My work as main Researcher in the AKT project was from October 2000 to November 2006. The Review Panel rated the project as “outstanding” scoring 34 out of a maximum possible 35 on the review criteria used to assess the results of projects by the EPSRC. First Review of the AKT project.
Dr. Maria Vargas-Vera’s AKT work.
- CONQUIRO A Cluster-Based Information Retrieval Meta-search Engine. CONQUIRO provides a suite of clustering methods: hierarchical and non-hierarchical clustering methods. CONQUIRO represents a solution to the problem of information management.
- AQUA Ontology driven Question Answering System. AQUA combines Natural Language processing (NLP), Ontologies, Logic, and Information Retrieval technologies in a uniform framework. AQUA makes intensive use of an ontology in several parts of the question answering system (read more information) . Typical questions using AQUA as closed-domain Question Answering System can be found at AQUA-QUESTIONS.
- SVE Student Essay Viewer. A system that allows tutors and students to visualize argumentation in student essays.
- EVENT (Event Recognition on News Stories). A system which allows to recognize events happening in a news stories and populates a hand-crafted ontology.
- ONTOSOPHIE is our new system for Semi-Automatic Population of Ontologies from Text with confidence values associated to both the extraction rules and extracted values (extracted slots).
- MnM Semi-Automatic Population of Ontologies from Text using Amilcare. MnM is the most well known system (in the Semantic Web Community) for ontology-driven semantic mark-up. However, MnM allows also to populate a given ontology. The latest is the main feature of the MnM system. A published Conference paper on MnM work had been widely cited in several research communities and boosted research on Information Extraction technologies. The MnM system is available as open source at http://kmi.open.ac.uk/projects/akt/MnM/
The name “MnM” is an abbreviation of “Maria and Mattia”. Therefore, I am proud to say that I have a system in my honour!. Furthermore, the MnM paper (result of the research AKT project) has been used in the context of European projects like Dotkom Project.
- INFOEXTRACTOR Semi-Automatic Population of Ontologies from Text using Marmot, Crystal and Badger
InfoExtractor talk presented at KCAP-2001, Victoria, Canada; October 2001.
Funded By: EPSRC, UK – (2000-2006)
Extraction of Fuzzy Rules from images (National Autonomous University of Mexico); CONACyT-REDII funded project: Grant: $21,600 USD; (holder)
This project was concerned with classification of facial expressions. One of the main challenges was to build a database of photographs showing different expressions. In this task we relied in students of an Acting Schools who were training to become actors. We developed a system which recognizes expressions from photographs in similar settings that P. Ekman except that we do it automatically. Another difference with P. Ekman is that we use Artificial Intelligence techniques in the recognition. Our system uses fuzzy rules to classify the universal facial expressions. Currently, it can only recognize the 6 universal expressions as categorized by P. Ekman like happiness, anger, surprise, fear, disgust and anger. However, a challenge problem was how to deal with overlapping expressions. In order to solve the latest, we thought that Fuzzy Logic was a good technique to be used.
Funded By: CONACyT, Mexico – (1998-2000)
Fuzzy Logic and Objects
Fril++: An Object-Oriented Fuzzy Logic Programming Language, (Bristol University, England UK) EPSRC funded project.
In this project we developed Fril++ an object-oriented extension to Fril. This new programming language supports the modeling of fuzzy applications by integrating a fuzzy logic programming language with the object-oriented paradigm. Fril++ includes the concept of fuzzy objects as an explicit construction in the language. Fuzzy objects are naturally related with multiple inheritance, and so we allow both of these in the language. The multiple inheritance gives us more flexibility, and in any case, the associated ambiguities are not so much of a problem in logic programming as they are in other paradigms. We also exploit the fact that information is expressed by the hierarchy itself. A Fril++ program can explicitly use the hierarchy by means of special constructions integrated into the language. Fril++ provides support for user decisions when there are several answers for a message by having a library of defuzzification methods. Finally we concentrated on testing Fril++ by using it for real applications such as an object-oriented data browser.
Funded By: Funded By EPSRC, UK – (1995-1997)
Construction of Logic Programs using the concept of Program Schemata
Automatic Programming — Program Composition (Edinburgh University, Scotland UK). PSPA-UNAM funded project; Grant: £24,000.
In this project we investigated the problem of how to build complex logic programs by repeatedly combine pairs of simpler logic programs. However, merely taking naïve combinations rapidly produces programs that are far too inefficient to be practical. Furthermore, program combination is a difficult task therefore, it is desirable that the combination process can be automated as far as possible and this is particularly important when the system is designed for use by people who are inexperienced in a logic programing language.
We considered the combination of Prolog programs, and designed and implemented an interactive system which largely solved the above problems. The system is based on program transformation complemented with knowledge of the program development. In particular, it is not possible to have a single method that will efficiently combine all of the many varieties of program that the system might encounter. Therefore, we have developed a program classification scheme, and an associated wide range of combination methods to produce efficient programs.
Funded By: UNAM, Mexico – (1991-1994)
Object-Oriented Databases & Logic Programming
A Copy Structure-Based Prolog interpreter for a Unix System (National Autonomous University of Mexico) Conacyt funded project.
In this project we designed and wrote a Prolog Interpreter as part of a project to incorporate Prolog as a query language to object-oriented database systems. This was the first contact that my department had with Prolog, and so, in order to design the interpreter we had to work out the exact definition of the Prolog grammar. In particular the grammar of Prolog infix operators is ambiguous (unlike languages such as C or Pascal), and we found that the standard tool YACC made unsuitable choices when disambiguating this grammar. To solve this we decided to use YACC and LEX to write a Prolog with a minimal and unambiguous set of operators, and then completed this to a full Prolog by use of a meta-interpreter written in the reduced Prolog. Also, the novel features of Prolog such as backtracking and cuts, implied the necessity of a very careful design of the complex stack-handling, garbage collection and other internal memory management tasks. The interpreter was written in standard Edinburgh syntax, offered a standard set of extra-logical features, tail-recursion optimization, and also non-standard improvements such as a debugger that used the user names for variables rather than the internal names. It was implemented in C, and the software was easy to transfer to different UNIX systems.
Funded By: CONACyT, Mexico – (1984-1987)
Great minds discuss ideas; average minds discuss events; small minds discuss people. Eleanor Roosevelt