Resources, Tools and Algorithms for the Semantic Web

As the vast majority of knowledge existing on the web is still embedded into written texts, it is obvious that any attempt to intelligently solve the problems of exploiting the web without heavily relying on Natural Language Processing (NLP) technologies is doomed to failure. It is generally known that the key issue in NLP is ambiguity. Ambiguity is present at all levels of traditional structuring of a language system (phonology, morphology, lexicon, syntax, semantics) and not dealing with it at the proper level exponentially increases the complexity of the problem solving. Currently, the state-of-the-art taggers (combining various models, strategies and processing tiers) ensure no less than 97-98% accuracy in the process of morpholexical full disambiguation. For such taggers a 2-best tagging is practically 100% accurate. One further step is the word sense disambiguation (WSD) process. In the Fregean compositional semantics, the meaning of a complex expression is supposed to be derivable from the meanings of its parts, and the way in which those parts are combined. Depending on the representation formalisms for the word-meaning representation, various calculi may be considered for computing the meaning of a complex expression from the atomic representations of the word senses. Obviously, one should be able, beforehand, to decide for each word in a text which of its possible meanings is the contextually right one. Therefore, it is a generally accepted idea that the WSD task is highly instrumental (if not indispensable) in semantic processing of natural language documents. SENSEVAL series of evaluation competitions on WSD is a very good source for learning how WSD evolved in the last 6-7 years and where is it nowadays. However, most of the successful commercial applications in language processing (text and/or speech) dispense of any explicit concern on semantics, with the usual motivations stemming from the computational high costs required by dealing with semantics in case of large volumes of data. With recent advances in corpus linguistics and statistical-based methods in NLP, revealing useful semantic features of linguistic data is becoming cheaper and cheaper and the accuracy of this process is steadily improving. Lately, there seems to be a growing acceptance of the idea that multilingual lexical ontologies might be the key towards aligning different views on the semantic atomic units to be used in characterizing the general meaning of various and multilingual documents. Depending on the granularity at which semantic distinctions are necessary, the accuracy of the basic semantic processing (such as word morphosyntactic disambiguation and word sense discrimination) can be very high with relatively low complexity computing. The current understanding of the Semantic Web concept depicts mainly two things: sharing and retrieval of information. Sharing information is about encoding information in such a way that any independent software application can query the data store and receive both information and information type and structure. Information retrieval is thought of as being an effective way of accessing any piece of data given that the software agent ‘knows’ already the structure it must crawl in order to find it. The authoritative statement about the Semantic Web belongs to Tim Berners-Lee ([15]): “The goal of the Semantic Web initiative is to create a universal medium for the exchange of data where data can be shared and processed by automated tools as well as by people. The Semantic Web is designed to smoothly interconnect personal information management, enterprise application integration, and the global sharing of commercial, scientific and cultural data. We are talking about data here, not human documents.”