As web search engines continue to improve, good results to such a query have improved. This type of question and answer often makes use of a semantic technology called natural language processing, one of many related technologies that comprises of the semantic software technology landscape. Quite simply, the vision of semantic search is the availability of software algorithms that would improve retrieval for the average person by interpreting their native inquiry and returning relevant results semantically (Lynda, 2010). As early as the 1980s significant research appeared in information science literature about the development of expert systems for the improvement of search results. (Peffers, 2007). Many universities, ICT companies, and major corporations have published research and came up with various algorithm techniques for machine-aided searching over three decades (and earlier when much of this work was classified as artificial intelligence). By the late 1990s and early 2000s, these technologies began to be described as semantic search components. (Akinsola, 2013). In 2001 Tim Berners-Lee published an article in Scientific American proposing a semantic web evolving out of the expanding worldwide web.
Web search is a key technology of the Web, which is essentially based on a combination of textual keyword search with an importance ranking of the documents depending on the link structure of the Web (Fazzinga and Lukasiewicz, 2010). For this reason, it has many limitations, and there are a lot of research activities towards more intelligent Web search, called semantic search on the Web, or also Semantic Web search, which is currently one of the popular research topics in both the Semantic Web and Web search (Baeza-Yates and Raghavan, 2010).
There is no unique definition of the notion of semantic search on the Web. However, the most common use is the one as an improved form of search on the Web, where meaning and structure are extracted from both the user’s Web search queries and different forms of Web content, and exploited during the Web search process (Herman, 2016). Such semantic search is often achieved by using Semantic Web technology for interpreting Web search queries and resources relative to one or more underlying ontologies, describing some background domain knowledge by connecting the Web resources to semantic annotations, or by extracting semantic knowledge from Web resources. Such a search usually also aims at allowing for more complex Web search queries whose evaluation involves reasoning over the Web (Fazzinga and Lukasiewicz, 2010). Another common use of the notion of semantic search on the Web is the one as search in the large datasets of the Semantic Web as a future substitute of the current Web. This second use of the notion of semantic web is closely related to the first one, since the extraction of semantic knowledge from Web resources, actually corresponds to producing a knowledge base, which may be encoded using Semantic Web technology. That is, the latter semantic search on the Web can essentially be considered as a sub problem of the former one. Another closely related use is the one as natural language search on the Web, where search queries are formulated in (written or even spoken) natural language (Mathieu et al., 2011).
Many approaches try to translate such queries into formal queries in a structured query language, which are generally available in the above semantic search in the context of the Semantic Web. The response to such natural language queries might be Web resources as usual, or they may likewise be structured or natural language results, towards more informative results, an illustration is by demonstrating the structured information extracted from the resulting Web pages, and by additionally connecting the search result with Wikipedia articles. This is another meaning of semantic search, which is actually a very simple form of question answering. Frequently, the notion of semantic search also covers some other semantic ideas and concepts (Fazzinga and Lukasiewicz, 2010). An example is the proposal of related searches, such as the completion and correction of Web search queries, which are well-known from standard Web search engines. Another example is full-text similarity search, where blocks of text ranging from phrases to full documents, rather than few keywords, are submitted. Closely related is ontological similarity search (Janowicz, 2008), in light of the closeness of ontological entities.
In this study, a domain ontology (the human anatomy ontology) is utilize to facilitate the concept of ontology whose design is becoming increasingly recognized as central to medical informatics and significantly more so to bioinformatics.
Ontology is an explicit and formal description of the conceptualization of domain of discourse (Gruber, 1993 and Guarino, 1998). Ontology represents domain knowledge with less emphasis on the linguistic realizations (words) of knowledge objects (Reiter and Buitelaar, 2008). In its most fundamental form, ontology comprises of a set of classes and a set of relations that describe the properties of each class. Ontology formally characterize relevant knowledge in a domain of discourse and can be used to interpret data in this domain, to reason over knowledge that can be extracted or inferred from this data and to incorporate extracted knowledge with other data or knowledge extracted elsewhere. It is used to capture the knowledge of any particular domain to avoid ambiguity of terms. The significant standpoint of ontology is that it will provide a globally unique identifier for all concepts, and it is used to capture knowledge in any given domain. Ontology describes the concepts in a given domain and the relationships that hold between them (Madurai and Sai-Baba, 2012). It helps to share common comprehension of the structure of information among the users, to enable reuse, analyze the domain knowledge. It enables to merge already existing knowledge by expanding it further (Michael and Steven, 2011).
WordNet provides a rich knowledge base in which concepts, termed synonymy sets or synsets, are connected by semantic relations. WordNet is depended on psycholinguistic theories to deﬁne word meaning and model word meaning relationships, and also meaning to meaning associations (Lin and Sandkuhl, 2008). WordNet tries to focus on the word meanings rather than word forms, however infection morphology is likewise considered. WordNet comprises of three databases, one for nouns, one for verbs and a third for adjectives and adverbs, but disregards prepositions and determinants. WordNet defines lexical entries (words with their linguistic meaning and possibly their morpho-syntactic features) with less emphasis on the domain knowledge associated with these (Rosse and Mejino, 2003). WordNet consists of a set of synonyms (synsets). A synset denotes a concept or a sense of a group of terms, but does not include senses for prefixed and multiword FMA terms. WordNet also provides textual descriptions of the concepts containing deﬁnitions and examples. WordNet does not include information about the etymology or the pronunciation of words and it contains only limited information about usage. WordNet intends to cover most of everyday English words and does not include much domain-specific terminology, for example the Foundation Model of Anotomy (Shvaiko, 2010).
WordNet is sometimes called ontology, a relentless claim that its creators do not make. The hypernym/hyponym relationships among the noun synsets can be interpreted as specialization relations among conceptual categories. In other words, WordNet can be interpreted and used as a lexical ontology in the computer science sense. However, such ontology ought to be corrected before being used since it contains many fundamental semantic irregularities, for example, (i) the presence of common specializations for exclusive categories and (ii) redundancies in the specialization hierarchy. Furthermore, transforming WordNet into a lexical ontology usable for knowledge representation such as the Foundational Model of Anatomy, should normally also involve (i) distinguishing the specialization relations into subtypeOf and instanceOf relations, and (ii) associating intuitive unique identifiers to each category. Most projects claiming to re-use WordNet for knowledge-based applications (typically, knowledge-oriented information retrieval) simply re-utilize it directly.
WordNet has also been converted to a formal specification, by means of a hybrid bottom-up top-down methodology to automatically extract association relations from WordNet, and interpret these associations in terms of a set of conceptual relations, formally defined in the DOLCE foundational ontology (Gangemi, et al. 2003). In most works that claim to have integrated WordNet into ontologies, the content of WordNet has not simply been corrected when it seemed necessary; instead, WordNet has been heavily re-interpreted and updated whenever suitable. This was the case when, for example, the top-level ontology of WordNet was re-structured. WordNet can be treated as a partially ordered synonym resources. This study also introduces a model (PrefSyn algorithm) that resolves this problem and makes these terms to have their senses in WordNet together with their term description. WordNet can thus be seen as a combination of dictionary and thesaurus, while it is accessible by human users through a web browser.
Ontology such as Foundational Model of Anatomy (FMA) is a frame-based ontology that represents declarative knowledge about the structural organization of the human body (Rosse and Mejino, 2003). The Foundational Model of Anatomy is an evolving ontology for biomedical informatics; it is concerned with the representation of entities and relationships necessary for the symbolic modeling of the structure of the human body in a computable form that is also understandable by humans. It describes the domain of human anatomy in much detail by way of class descriptions for anatomical objects and their properties. Additionally, the FMA lists terms in several languages for many classes, which makes it a lexically enriched ontology already. Specifically, the FMA is an abstraction that explicitly represents a coherent body of declarative knowledge about human anatomy as domain ontology. This ontology is executed in an online based frame-work and then, stored in a database. It is intended to also serve as a reusable and generalized resource of deep anatomical knowledge, which can be separated, to meet the needs of any knowledge-based application that requires structural information (Brinkley, 2006). The human anatomical taxonomy (vocabularies) derived from the domain of ontology, the FMA, have their term description as well as senses as represented in the WordNet. WordNet which is a lexical database for English language that groups English word into set of synonyms called synsets, provide short definitions and usage examples, and records a number of relations among these synonym set or their members (Zhang and Olivier, 2007). These anatomical terms can be simple consisting of just a word or complex, consisting of multi-words (in which some of these terminologies are not found in the WordNet). This research work presents an approach towards the enhancement of WordNet search in particular for the human anatomy domain as represented by the foundational model of anatomy (FMA). This approach which involves the use of a PrefSyn model to enable the lookup of such terms via a simple interface which will help in identifying these prefixed terms to improve information retrieval.
WordNet provides a rich knowledge base in which concepts, termed synonymy sets or synsets, are linked by semantic relations. Terms in FMA can be simple, consisting of just one word. However, a main barrier to exploiting this term in WordNet is due to the prefixes added to human anatomical terms which describes or shows the position of the organ or tissue on the human body, and such term have no senses as well as term description in the WordNet. Due to this ambiguity, not all terms for the human anatomy can be found in the WordNet.
Furthermore, these anatomical terms can be complex or multi-words, which is a problem in finding the right level of generalization; that is, finding the concept which belongs to a given set of concepts; but it could be the case that the class which would optimally capture the generalization is not lexical (that is, synset), but abstract. It can also be the case that WordNet simply is not the kind of taxonomy required; this can be due to several reasons: incompleteness, incorrect structuring, therefore, WordNet structure should be arranged differently for particular Natural Language Processing tasks. Therefore, this project work sought to provide solutions to the above problems.
The aim of this study is to develop a PrefSyn model to enhance the WordNet search for human anatomy ontology. To achieve this aim, the following objectives are realized.
i. To develop an extraction algorithm for FMA terms.
ii. To design a prefsyn model for FMA terms in wordnet.
iii. To implement the proposed system with nltk as API using python.
- To test the enhanced WordNet for FMA terms to be able to lookup prefixed anatomical terms.
The significance of this study is to be able to lookup prefixed human anatomical terms derived from the FMA, on WordNet through an interface (that combines both the FMA and WordNet). This combines anatomical information in symbolic form, senses and terms description to information modelers and other developers of applications for education, clinical medicine, electronic health record, biomedical research and all other areas of health care delivery and management. This is intended to allow users search for simple, multi-words and complex anatomical terms.
It is designed to be understandable by the users, thus this will serve as a biomedical informatics resource for enhancing the human anatomy ontology.
The scope of the project covers both the human anatomy as domain ontology, and the WordNet containing semantic lexicon, as two resources that will be integrated through an interface using a PrefSyn model, to enable lookup of FMA terms with their senses and term descriptions. The limitations to the research work include the aspect of multiword (for example, the left antebrachial) terms of the FMA which pose a challenge in implementing the interface as this will only focuses on single prefixed anatomical terms.
In chapter one, an introduction to the proposed system is given, as well as statement of problems of previous system, aim and objectives of the study, the significance of the study, the scope and limitation to which this study covers, while chapter two gives a detailed literature review of previous related studies as regard to the proposed system. In chapter three, a methodological approach is used in implementing the proposed system. Chapter four gives the results and a quantitative discussion of the implemented system, and the last chapter provides a brief summary of this research work, concluding remark, and recommendations for future work.
This Project Material Cost Only ₦4000.00