[ to the main ]
My research
These are the major topics of my current research and scientific work:Self organizing web portal and evolutionary navigation
A method was invented to perform self-organising evolutionary navigation for unstructured web sources. A combination of search engine elements (search spider) and evolutionary algorithm is used. An application was also developed coded in PHP+MySQL and the the method was tested and displayed an impressive performance. Please check a demo of self organizing web navigation solution. Your comments and feedback are always welcome. If interested, you can view the paper on Self-organising web portal with evolutionary links panel by me and Viktor Popov. You can find the method and algorithm described there. The Power Point presentation is also available for download: Self Organising Web Portal - presentation.The developed method can be used to present search result in a better way in search engines or to create self-organising web portals with information collected from a set of predefined sources.
The developed method can be used anywhere when it's necessary to provide a convenient navigation over the information from different web sources or to present unstructured information. This may be very useful for automated creation of self-maintained web portals and communities for groups of people interested in particular topics.
This can also be used to filter search results and provide more convenient and intelligent search which means faster and more precise retrieval of the information.
The database maintained by the search spider of the system allows additional features to be implemented to improve navigation, usability and search engine optimisation such as:
-
similar pages retrieval - it is possible to implement a button at each page
to find the pages with relevant content from the current and other websites which are
indexed
visual representation of content - it is possible to implement visual representation
mechanisms such as SOM
site maps - site maps and googlemaps can be generated automatically for users'
convenience and better search engine indexing
-
search spider and keyword analyser - the data is collected and analysed automatically so it is possible to provide navigation
over any set of web pages, the system should only be supplied with start urls which have
links to other pages
the developed evolutionary algorithm has its own advantages:
limited set size - the set of links is tiny and resembles navigational panels and menus most internet surfers are used to; this is good for usability - users attention doesn't disperse, psychologically it is easier to deal with small groups of objects than with a big heap as it's done in search engines
random factor - it is good to propose at least something when user didn't specify his interests yet or he doesn't know what to start with, so the initial state of the system is a random set of links fetched from the database
time factor - the ranks of the pages decrease with time so system gradually 'forgets' the interests of user when he doesn't support them anymore which is good as user may change his interests or become more specific, so the system is alive and ready for interaction
page similarity - special algorithm is implemented to extract the pages which are most relevant to those selected by user. Since more than one page (topic of interest) can be selected and the algorithm is iterative, it allows to find the information step-by-step, in few mouse clicks. No need to meditate on 'advanced search' form, you can just click and see what happens, which is more usable and corresponds with human's nature.
Note: Page similarity algorithm was improved since the paper was published.
now the probability of new pages to appear depends not only on one page that was selected
during previous iteration but on other pages with positive ranks in the set and both the
number of matching keywords and pages' ranks are considered. So the algorithm became more
precise and provides more relevant pages to the topics selected by user. The research and
development is being carried out to improve the page similarity method furthermore
Intelligent analysis of results in computer based assessments
In distance learning the function of measuring the students' knowledge is carried out by computer based assessment systems. However the development of these systems and distance learning area on a whole is obstructed by the fact that human analyst is still required to analyse the reliability of tests used and results obtained.The study reveals that it is possible by analysing the results of assessments to reveal the qualitative adjectives of the tests. This consequently means that it is possible to implement the self-assessment approach into automated testing systems, which would help teachers to evaluate their tests and assessments results without the help of specialists. This will also provide a platform for further automation and improvement of computer based assessment systems using the same principle.
Following the study, a set of tools was developed aimed to conduct a comprehensive analysis of results of assessments, to discover the foregoing factors and to report the conclusions in an understandable form.
[ distance learning, computer based assessment, intelligent analysis, statistics, self-analysis, assessment results, test quality ]
Here you can view the paper on Intelligent analysis tools for computer based assessments by me and Viktor Popov.
The presentation on "Intelligent analysis tools for computer based assessments" presented at Data Mining and Information Engineering 2006 conference in Prague on 13th of July can be downloaded here (just click the title of presentation).
WIT distance learning system
Please visit Distance Learning System of Wessex Institute of Technology. This system was developed by me and it is written in PHP+MySQL. Many thanks to Justing Kercher and others who made design template, helped to test and improve the system. WIT DL is being used for distance learning courses and is a support centre for students and researchers linked with WIT. It's also used as a platform for my research in distance learning area. For example, intelligent analysis tools to analyse the quality of tests and their questions were applied for the assessment module within WIT DL system.Global Knowledge Map - Towards Idealistic Knowledge Representation System
One of the most significant problems which inhibit further developments in the areas of Knowledge Representation and Artificial Intelligence is a problem of semantic alignment or knowledge mapping. The progress in its solution will be greatly beneficial for the tasks of information retrieval, ontology alignment, relevance calculation, text understanding etc. In the paper the concept of multidimensional global knowledge map, elaborated through unsupervised extraction of dependencies from large documents corpus, is proposed. In addition, the problem of direct Human - Knowledge Representation System interface is addressed and a concept of adaptive decoder proposed for the purpose of interaction with previously described unified mapping model. In combination these two approaches are suggested as basis for a development of a new generation of knowledge representation systems.Other
I have strong scientific and practical interests and ongoing research in such topics as semantic web, clustering technologies, e-commerce, usability, mobile internet, data mining, text mining, web mining, text compression, self-organising web and adaptive interfaces, artificial intelligence, innovative e-commerce etc.Please feel free to contact me (see main page for contacts) in case you have any questions or ideas for our collaboration.
[ to the main ]