Home /About SECABA Lab

About SECABA Lab

Exaple GraphSECABA Lab is a multi-disciplinary laboratory composed of researchers on Library Sciences and/or Computer Sciences. SECABA develops technologies based on Soft Computing tools to improve the information access processes on both the Web and Digital Libraries (DL). The main is to facilitate the users the expression of their information needs in the information access processes and satisfy those needs with relevant information.

So, SECABA develops new user-centered approaches to evaluate quality on the Web and DL in order to identify the problems of information access services and support the development of better information access systems. SECABA mainly works in the following research lines: Information Retrieval (IR) foundations, Web Information Retrieval, Recommender Systems, Web Quality Evaluation, Semantic Web and Digital Libraries, Internet of Things and Digital technology consulting for bussinesses.

On the Quality Evaluation field SECABA leads and produces quality studies in a variety of fields as libraries (specially digital ones), air transport and hotels. To do so, we make use of new evaluation techniques both at a subjective level (from users perceptions) and objective levels (though the use of different measures and indicators). Quality is defined as the totality of characteristics of an entity that bear on its ability to satisfy stated and implied needs. Needs may change with time, which implies a permanent review of requirements for quality. It is crucial that these requirements fully reflect the customer's needs. Quality is often equated with usability ("fitness for use"), functionality, customer satisfaction, or conformity with generally accepted standards and requirements.

Our work on IR involves designing sophisticated IRSs based on fuzzy logic and evolutionary computing with two different goals: facilitating the users the expression of their information needs in the form of advanced fuzzy queries, and effectively solving those needs with relevant information. Different specific research lines are grouped into this generic objective:

  • Design of a FIRS incorporating an advanced query language able to manage multi-level, multi-granular and no-balanced linguistic information.
  • Design of several tools to assist the user in the use of the complex weighted query language of our FIRS such as:
    1. Inductive Query by Example (IQBE) techniques, based on sophisticated evolutionary approaches such as genetic algorithm-programming, simulated annealing-programming and multiobjective evoluEAs, to automatically derive fuzzy queries from a set of documents provided by the user in an off-line manner. On the other hand, we also apply these evolutionary techniques to automatically derive queries from other information retrieval models.
    2. Relevance feedback techniques to on-line refine previous linguistically weighted queries provided by the user using the same evolutionary techniques.
    3. Recommendation systems allowing the users to formulate their queries taking into account the information obtained from others users having similar information needs in the past.

    The combination of these approach in the IR framework allows us to design more powerful user profiles for information routing purposes. These profiles are so considered as persistent queries, with a more representative fuzzy structure, that are automatically derived from set of relevant and irrelevant documents representing the user's information needs by means of EAs.

  • Desing of other web tools to improve the representation, understanding, visility and access to the Web-based IRS (like search engineers, digital libraries, etc). These techiques are:
    1. The Semantic Web, which is the abstract representation of data on the World Wide Web, based on the Resource Description Framework (RDF) standards and other standards to be defined. The Semantic Web brings to the Web the idea of having data defined and linked in a way that it can be used for more effective discovery, automation, integration, and reuse across various applications.
    2. Web Services Quality Evaluation, with the growth of information on the Internet and the development of more sophisticated web-based information systems, there is now the more likely possibility of finding information and answers to real questions. But, within the morass of networked data are both valuable nuggets and an incredible amount of junk. Web services quality evaluation desings systematic approach to evaluating the Web-sites of these information services to determine their quality.