Ontology learning

The problem that ontology learning deals with is the knowledge acquisition bottleneck, that is to say the difficulty to actually model the knowledge relevant to the domain of interest. Ontologies are the vehicle by which we can model and share the knowledge among various applications in a specific domain. So many research developed several ontology learning approaches and systems. In this paper, we present a survey for the different approaches in ontology learning from semi-structured and unstructured date

Ontologies are often viewed as the answer to the need for interoperable semantics in modern information systems. The explosion of textual information on the Read/Write Web coupled with the increasing demand for ontologies to power the Semantic Web have made (semi-)automatic ontology learning from text a very promising research area. This together with the advanced state in related areas, such as natural language processing, have fueled research into ontology learning over the past decade. This survey looks at how far we have come since the turn of the millennium and discusses the remaining challenges that will define the research directions in this area in the near future.

Quran is the holy book of Muslims that contains the commandment of words of Allah. Quran provides instructions and guidance to humankind in achieving happiness in life in the world and the hereafter. As a holy book, Quran contains rich knowledge and scientific facts. However, humans have difficulty in understanding the Quran content. It is caused by the fact that the meaning of the searched message content depends on the interpretation. Ontology able to store the knowledge representation of Holy Quran. This paper studies recent ontology on Holy Quran research. We investigate the current trends and technology being applied. This investigation cover on several aspects, such as outcomes of previous studies, language which used on ontology development, coverage area of Quran ontology, datasets, tools to perform ontology development ontology population techniques, approaches used to integrate the knowledge of Quran and other resources into ontology, ontology testing techniques, and limitations on previous research. This review has identified four major issues involved in Quran ontology, i.e. availability of Quran ontology in various translation, ontology resources, automated process of Meronymy relationship extraction, and Instances Classification. The review of existing studies will allow future researchers to have a broad and useful background knowledge on primary and essential aspects of this research field.

Stemming is a key step in most text mining and information retrieval applications. Information extraction, semantic annotation, as well as ontology learning are but a few examples where using a stemmer is a must. While the use of light stemmers in Arabic texts has proven highly effective for the task of information retrieval, this class of stemmers falls short of providing the accuracy required by many text mining applications. This can be attributed to the fact that light stemmers employ a set of rules that they apply indiscriminately and that they do not address stemming of broken plurals at all even though this class of plurals is very commonly used in Arabic texts. The goal of this work is to overcome these limitations. The evaluation of the work shows that it significantly improves stemming accuracy. It also shows that by improving stemming accuracy, tasks such as automatic annotation, and keyphrase extraction can also be significantly improved.

The Semantic Web was designed to represent the enormous data that is existing on the World Wide Web in a machine readable format. The research shows the long period of time that was spent on the Emails for communication and information exchange. Adding the semantics to the existing Email systems could not only provide for the valuable usage of time and resources, but also refreshes the meaning of Email communication. The presented research work examines the ontology extraction process from the Email
systems adopting scalable pattern rules that is based on the extracted techniques. The proposed architecture is designed to handle the unstructured Emails and the ontologies that are extracted from the Email which is divided into four main components as follows: the Ontology Learning Component, the Management Component, the Semantic Email Component and the Client Side Plugin.

Ontologies constitute an approach for knowledge representation that can be shared establishing a shared vocabulary for different applications and are also the backbone of the Semantic Web. Thus a fast and efficient ontology development is a requirement for the success of many knowledge based systems and for the Semantic Web itself. However, ontology development is a difficult and time consuming task. Ontology learning is an approach for the problem of knowledge acquisition bottleneck that aims at reducing the cost of ontology construction through the development of automatic methods for the extraction of knowledge about a specific domain and its representation in an ontology like structure. This paper provides a discussion on existing ontology learning techniques and the state of the art of the field.

The reasoning tasks that can be performed with semantic web service descriptions depend on the quality of the domain ontologies used to create these descriptions. However, building such domain ontologies is a time consuming and difficult task.

As the utilization and development of Web services grows rapidly, the problem of analysing of existent Web services naturally arises. The analysis provided by the recent research work mostly extends to the statistical and explanatory examination of collections of Web services while deeper analysis of semantically annotated versions of services and the practical usage of the results of analysis are yet to receive substantial attention. The vast major- ity of existing web services lack any kind of formally expressed semantics, making semantic annotation a crucial preliminary step for analysis. In the absence of appropriate reference domain ontologies, annotation of existing web services is dependent on ontology development and ontology learning techniques. There is a great need for the development of automated ontology learning systems that allow semantic annotation of large collections of web services lacking any auxiliary textual materials. There is also a lack of effec- tive evaluation frameworks for assessing the quality of the provided semantic annotation on a large scale. The availability of a large quantity of semanti- cally annotated web services should allow the development of service selection and composition methods, taking advantage of domain diversity and the real- world characteristics of Web services. Moreover, user-generated contents such as tags and ratings in social networks are a rich source of information that can be exercised to perform more efficient service selection. Similarly, any method that targets the employment of user profiles in a social network needs to address a plausible solution for privacy issues.
The main contribution of this dissertation is the development of techniques and frameworks supporting the construction, exploitation, and analysis of semantic web services. Specifically, we developed a semi-supervised method for ontology learning from Web service interface descriptions (WSDLs). The generated ontology is later used for semantic annotation of the examined web services. We also introduced a problem ontology as a specific case of task ontology, decomposing the given user query into a set of web services satisfying user requirements. The feasibility of problem ontology is evaluated as an action-planning component for a multi-robot system. Further, we developed an evaluation approach, suitable for the effective evaluation of large-scale, heterogeneous, real-world web service annotations. The approach consists of a set of procedures and metrics from network theory applied to the network structure of web services. The network is constructed by linking web services via matching input and output parameters of their operations. These web service networks are used to discover information exchange patterns among communities of services. The determined patterns can be employed for more effective service selection and composition strategies. Finally, in the context of a social network, we introduce a framework for privacy trust-aware user profile utilization. The efficiency of this framework is evaluated in the context of a user item-based recommendation system

In this paper we present a new version of OntoGen system for semi-automatic data-driven ontology construction. The system is based on a novel ontology learning framework which formalizes and extends the role of machine learning and text mining algorithms used in the previous version. List of new features includes extended number of supported ontology formats (RDFS and OWL), supervised methods

—An ontology is a formal and explicit specification of a shared conceptualization. Manual construction of domain ontology does not adequately satisfy requirements of new applications, because they need a more dynamic ontology and the possibility to manage a considerable quantity of concepts that humans cannot achieve alone. Researchers have discussed ontology learning as a solution to overcome issues related to the manual construction of ontology. Ontology learning is either an automatic or semi-automatic process to apply methods for building ontology from scratch, or enriching or adapting an existing ontology. This research focuses on improving the process of term extraction for identifying concepts in ontology learning. Available approaches for term extraction process are limited in various ways. These limitations include: (1) obtaining domain-specific terms from a domain expert as seed words without automatically discovering them from the corpus, and (2) unsuitable usage of corpora in discovering domain-specific terms for multiple domains. Our study uses linguistic analysis and statistical calculations to extract domain-specific simple and complex terms to overcome this first limitation. To eliminate the second limitation, we use multiple contrastive corpora that reduce the biasness in using a single contrastive corpus. Evaluations show that our system is better at extracting terms when compared with the previous research that used the same corpora.

The paper is a conceptual contribute to educational assessment on its methodological dimension using a balanced symbiosis between ontologies, concept maps and virtual environments (forum, wiki). The ontological structure considers the shared conceptual cores and aims to integrate the formal size of a knowledge domain with the daily practice of relational different communities of practice and learning ("assessment FOR learning" [1]).

As more information becomes available on the Web, there has been a crescent interest in effective personalization techniques. Personal agents providing assistance based on the content of Web documents and the user interests emerged as a viable alternative to this problem. Provided that these agents rely on having knowledge about users contained into user profiles, i.e., models of user preferences and interests gathered by observation of user behavior, the capacity of acquiring and modeling user interest categories has become a critical component in personal agent design. User profiles have to summarize categories corresponding to diverse user information interests at different levels of abstraction in order to allow agents to decide on the relevance of new pieces of information. In accomplishing this goal, document clustering offers the advantage that an a priori knowledge of categories is not needed, therefore the categorization is completely unsupervised. In this paper we present a document clustering algorithm, named WebDCC (Web Document Conceptual Clustering), that carries out incremental, unsupervised concept learning over Web documents in order to acquire user profiles. Unlike most user profiling approaches, this algorithm offers comprehensible clustering solutions that can be easily interpreted and explored by both users and other agents. By extracting semantics from Web pages, this algorithm also produces intermediate results that can be finally integrated in a machine-understandable format such as an ontology. Empirical results of using this algorithm in the context of an intelligent Web search agent proved it can reach high levels of accuracy in suggesting Web pages. r

Ontologies in current computer science parlance are computer based resources that represent shared conceptualizations for a specific domain. This paper first introduces ontologies in general and subsequently, in particular, shortly outlines the DOGMA ontology learning approach. The paper also introduces the reader in the field of Knowledge Discovery in Text before, in the main part, work in progress is described and experimentally evaluated. It concerns a potential method that ultimately aims at automatically extracting concepts and conceptual relationships from texts in an unsupervised way. Preliminary outcomes are presented based on the clustering of nominal terms and prepositional phrases according to cooccurrence frequencies in the verb-object syntactic context.

Today, the notion of Semantic Web has emerged as a prominent solution to the problem of organizing the immense information provided by World Wide Web, and its focus on supporting a better co-operation between humans and machines is noteworthy. Ontology forms the major component of Semantic Web in its realization. However, manual method of ontology construction is time-consuming, costly, error-prone and inflexible to change and in addition, it requires a complete participation of knowledge engineer or domain expert. To address this issue, researchers hoped that a semi-automatic or automatic process would result in faster and better ontology construction and enrichment. Ontology learning has become recently a major area of research, whose goal is to facilitate construction of ontologies, which reduces the effort in developing ontology for a new domain. However, there are few research studies that attempt to construct ontology from semi-structured Web pages. In this paper, we present a complete framework for ontology learning that facilitates the semi-automation of constructing and enriching web site ontology from semi structured Web pages. The proposed framework employs Web Content Mining and Web Usage mining in extracting conceptual relationship from Web. The main idea behind this concept was to incorporate the web author's ideas as well as web users’ intentions in the ontology development and its evolution.

Resumo: Este trabalho apresenta a ferramenta OntoLP, desenvolvida como um plug-in para o ambiente Protégé, que faz a análise de um corpus de domínio em língua portuguesa e sugere candidatos a conceitos e hierarquias ao engenheiro de ontologia com base no conhecimento representado nos textos.

After the vision of the Semantic Web was broadcasted at the turn of the millennium, ontology became a synonym for the solution to many problems concerning the fact that computers do not understand human language: if there were an ontology and every document were marked up with it and we had agents that would understand the markup, then computers would finally be able to process our queries in a really sophisticated way. Some years later, the success of Google shows us that the vision has not come true, being hampered by the incredible amount of extra work required for the intellectual encoding of semantic mark-up -as compared to simply uploading an HTML page. To alleviate this acquisition bottleneck, the field of ontology learning has since emerged as an important sub-field of ontology engineering.

Automatic detection of antonymy is an important task in Natural Language Processing (NLP). However, currently, there is no effective measure to discriminate antonyms from synonyms because they share many common features. In this paper, we introduce APAnt, a new Average-Precision-based measure for the unsupervised identification of antonymy using Distributional Semantic Models (DSMs). APAnt makes use of Average Precision to estimate the extent and salience of the intersection among the most descriptive contexts of two target words. Evaluation shows that the proposed method is able to distinguish antonyms and synonyms with high accuracy, outperforming a baseline model implementing the co-occurrence hypothesis.

This is the second of a two-part paper to review ontology research and development, in particular, to ontology mapping and evolving. Ontology is a formal explicit specification of a shared conceptualization. Ontology itself is not a static model so that it must have the potential to capture changes of meanings and relations. As such, mapping and evolving ontologies is part of an essential task of ontology learning and development.

Ontologies have proven to be a powerful tool for many tasks such as natural language processing and information filtering and retrieval. However their development is an error prone and expensive task. One approach for this problem is to provide automatic or semi-automatic support for ontology construction. This work presents the Probabilistic Relational Hierarchy Extraction (PREHE) technique, an approach for extracting concept hierarchies from text that uses statistical relational learning and natural language processing for combining cues from many state-of-the-art techniques. A Markov Logic Network has been developed for this task and is described here. A preliminary evaluation of the proposed approach is also outlined.

English. Automatic detection of antonymy is an important task in Natural Language Processing (NLP). However, currently, there is no effective measure to discriminate antonyms from synonyms because they share many common features. In this paper, we introduce APAnt, a new Average-Precision-based measure for the unsupervised identification of antonymy using Distributional Semantic Models (DSMs). APAnt makes use of Average Precision to estimate the extent and salience of the intersection among the most descriptive contexts of two target words. Evaluation shows that the proposed method is able to distinguish antonyms and synonyms with high accuracy, outperforming a baseline model implementing the co-occurrence hypothesis.

Ontologies are often viewed as the answer to the need for interoperable semantics in modern information systems. The explosion of textual information on the Read/Write Web coupled with the increasing demand for ontologies to power the Semantic Web have made (semi-)automatic ontology learning from text a very promising research area. This together with the advanced state in related areas, such as natural language processing, have fueled research into ontology learning over the past decade. This survey looks at how far we have come since the turn of the millennium and discusses the remaining challenges that will define the research directions in this area in the near future.

Since the manual construction of ontologies is time-consuming and expensive, an increasing number of initiatives to ease the construction by automatic or semi-automatic means have been published. Most initiatives combine a certain level of NLP techniques with machine learning approaches to find concepts and relationships. However, a challenging issue is to quantitatively evaluate the usefulness or accuracy of the techniques and combinations of techniques when applied to ontology learning. We are developing a framework for acquiring an ontology from a large collection of domain texts. This framework provides support for evaluating different NLP and machine learning techniques when they are applied to ontology learning. Our initial experiment supports our assumptions on the usefulness of our approach.

Ontologies are an important instrument for structuring knowledge in a way that it can be easily accessed by computers and used for a wide range of purposes. Despite their importance, however, ontology development and maintenance is still expensive and time consuming, because human intervention is often still required. In recent decades, Ontology Learning (OL) has been attempting to automate these processes in order to reduce the costs. In the present paper, we provide a survey about the state-of-the-art in OL and we briefly mention some of the open issues in the field.

Domain ontology plays an important role in annotating web resources with proper semantic information. The underlying assumption behind this work is that the noun phrases appearing in the headings of a document as well as the document's hierarchical structure can be used to discover the concepts and is-a relations between them in the documents' domain. In order to verify this assumption a methodology was proposed, and a system was implemented and applied on a set of Arabic agricultural extension documents. The system ...

Ontologies play a vital role in many Web and internet related applications. However the demand on ontologies does not always meet with their availability and the manual construction of a high quality ontology is an expensive and time-consuming process. This work, presents a system for accelerating the ontology building process via semi-automatically learning a hierarchal ontology given a set of domain specific web documents and a set of seed concepts. The methods are tested with Web documents in the domain of Agriculture. The ontology is constructed through the use of two complementary approaches. The first approach utilizes the structure of phrases appearing in HTML headings while the second uses the hierarchical structure of the HTML headings for identifying new concepts and their taxonomical relationship between seed concepts and between each other. The presented system has been used to build an ontology in the agricultural domain using a set of Arabic extension documents. The resulting ontology was evaluated against a modified version of the AGROVOC ontology..

Ontology is an important emerging discipline that has the huge potential to improve information organization, management and understanding. It has a crucial role to play in enabling content-based access, interoperability, communications, and providing qualitatively new levels of services on the next generation of Web transformation in the form of the Semantic Web.

Similarity measures play a key role in the Semantic Web perspective. Indeed, most of the ontology related operations such as ontology learning, ontology alignment, ontology ranking and ontology population are grounded on the notion of similarity. In the last few years several similarity functions have been proposed for measuring both concept similarity and ontology similarity. However, they lack of a comprehensive formal characterization that is able to explain their behavior and value added, in particular when the ontologies are formulated in description logics languages like OWL-DL. Concept similarity functions need to be able to deal with the high expressive power of the ontology representation language, and to convey the underlying semantics of the ontology to which concepts refer. We propose a semantic similarity measure for complex Description Logics concept descriptions that elicits the underlying ontology semantics. Furthermore, we theorize a set of criteria that a measure has to satisfy in order to be compliant with a semantic expected behavior.

Ontology provides a shared and reusable piece of knowledge about a specific domain, and has been applied in many fields, such as Semantic Web, e-commerce and information retrieval, etc. However, building ontology by hand is a very hard and error-prone task. Learning ontology from existing resources is a good solution. Because relational database is widely used for storing data and OWL is the latest standard recommended by W3C, this paper proposes an approach of learning OWL ontology from data in relational database. Compared with existing methods, the approach can acquire ontology from relational database automatically by using a group of learning rules instead of using a middle model. In addition, it can obtain OWL ontology, including the classes, properties, properties characteristics, cardinality and instances, while none of existing methods can acquire all of them. The proposed learning rules have been proven to be correct by practice.

English. Automatic detection of antonymy is an important task in Natural Language Processing (NLP). However, currently, there is no effective measure to discriminate antonyms from synonyms because they share many common features. In this paper, we introduce APAnt, a new Average-Precision-based measure for the unsupervised identification of antonymy using Distributional Semantic Models (DSMs). APAnt makes use of Average Precision to estimate the extent and salience of the intersection among the most descriptive contexts of two target words. Evaluation shows that the proposed method is able to distinguish antonyms and synonyms with high accuracy, outperforming a baseline model implementing the co-occurrence hypothesis.

—Understanding or acquiring a user's information needs from their local information repository (e.g. a set of example-documents that are relevant to user information needs) is important in many applications. However, acquiring the user's information needs from the local information repository is very challenging. Personalised ontology is emerging as a powerful tool to acquire the information needs of users. However, its manual or semi-automatic construction is expensive and time-consuming. To address this problem, this paper proposes a model to automatically learn personalised ontology by labelling topic models with concepts, where the topic models are discovered from a user's local information repository. The proposed model is evaluated by comparing against ten baseline models on the standard dataset RCV1 and a large ontology LCSH. The results show that the model is effective and its performance is significantly improved.

Ontologies play a vital role in many web and internet related applications. However the demand on ontologies does not always meet with their availability and the manual construction of a high quality ontology is an expensive and a time consuming process. In this work, a system for accelerating the ontology building process via semi- automatically learning a hierarchal ontology given

Ontology engineering has been fully or partially practiced by knowledge engineers or knowledge workers, towards delivering either fully fledged conceptualizations of domains or providing lightweight ontology versions for less demanding but more frequent knowledge tasks. Domainspecific information can be shaped into ontologies either manually or (semi-)automatically using ontology learning techniques. The aim of the paper is to present a novice ontology learning approach that automatically constructs kickoff and useful ontologies from query logs. We place tasks related to the proposed learning approach in all phases of an ontology engineering life-cycle. By providing knowledge workers a useful kick-off ontology that is automatically built from "their needs (i.e. users' search interests)" in order to address "their needs (i.e. use of the kick-off ontology to query data precisely)", an approach that contributes as an incentive in the semantic content creation bottleneck is introduced.

The paper reports on the methodology and preliminary results of a case study in automatically extracting ontological knowledge from Italian legislative texts in the environmental domain. We use a fully-implemented ontology learning system (T2K) that includes a battery of tools for Natural Language Processing (NLP), statistical text analysis and machine language learning. Tools are dynamically integrated to provide an incremental representation of the content of vast repositories of unstructured documents. Evaluated results, however preliminary, are very encouraging, showing the great potential of NLP-powered incremental systems like T2K for accurate large-scale semi-automatic extraction of legal ontologies.

The efficient authoring of learning content is a central problem of courseware engineering. Courseware authors will appreciate the benefits of tools which automate various authoring tasks. We describe a system, OntAWare, which provides an environment comprising a set of software tools that support learning content authoring, management and delivery. This system exploits an opportunity provided by the emerging technologies of the Semantic Web movement, most notably knowledge-representation standards and knowledge-processing techniques. The system represents a combination of these newer developments with earlier work in areas such as artificial intelligence (AI) and intelligent tutoring systems (ITS).

—Concepts in web ontologies help machines to understand data through the meanings they hold. Furthermore, learning contexts and topics of web documents also have helped in better semantic-oriented structuring and retrieval of data on the web. In this short paper we present a novel approach for domain-independent open learning of domain concepts, context and topic of any given web document. Our approach is based on a computational version of the Construction-Integration (CI) model of text comprehension. Our proposed system mimics the way humans learn the meanings of textual units and identify domain concepts, contexts and topics in the form of semantic networks. We apply our system on a number of web documents with a range of topics and domains. The resulting semantic networks provide a quantitative and qualitative insights into the nature of the given web documents.

Fuzzy Formal Concept Analysis is a generalization of Formal Concept analysis (FCA) for modeling uncertain information and has been applied in recent years for supporting different activities of semantic web context. However Fuzzy FCA considers type-1 fuzzy rules and does not benefit the opportunities provided by type-2 fuzzy systems. The heterogeneity and imprecision of data on the internet in one hand and the need to extract intensively blurred information from the net for semantic web purposes, on the other hand, highlights the necessity for the development of type-2 fuzzy analytical system. Accordingly, this paper proposes an Interval Type-2 Fuzzy FCA system and applies it for similarity reasoning in the context of the semantic web.

We present initial experimental results of an approach to learning ontological concepts from text. For each word to be learned, our system a) creates a corpus of sentences, derived from the web, containing this word; b) automatically semantically annotates the corpus using the OntoSem semantic analyzer; c) creates a candidate new concept by collating semantic information from annotated sentences; and d) finds in the existing ontology concept(s) "closest" to the candidate. In the long term, our approach is intended to support the continual mutual bootstrapping of the learner and the semantic analyzer as a solution to the knowledge acquisition bottleneck problem in AI.

Semantic Web offers many possibilities for future Web technologies. Therefore, it is a need to search for ways that can bring the huge amount of unstructured documents from current Web to Semantic Web automatically. One big challenge in searching for such ways is how to understand patterns by both humans and machine. To address this issue, we present an innovative model which interprets patterns to high level concepts. These concepts can explain the patterns' meanings  in a human understandable way while improving the information filtering performance. The model is evaluated by comparing it against one state-of-the-art benchmark model using standard Reuters dataset. The results show that the proposed model is successful. The significance of this model is three fold. It gives a way to interpret text mining output, provides a technique to find concepts relevant to the whole set of patterns which is an essential feature to understand the topic, and to some extent overcomes information mismatch and overload problems of existing models. This model will be very useful for knowledge based applications.

Knowledge extraction methods have not efficiently evolved towards new methods to automate the process of building multilingual ontologies as the main representation of structured knowledge. The need for ontologies that support massive contents suffers from this bottleneck. Whereas texts are written in natural language, ontologies are expressed in formal languages. This paper describes a new approach based on an interlingual meaning representation between natural and formal languages, where linguistic patterns are identified. The advantages of this method to reach an important level of automation for ontology building along and associated multilingual aspects are depicted. The process is described in detail in a case study.

The present-day web can be broken into smaller pieces that slowly but surely will be transformed into semantic web pieces. This paper describes an approach for eliciting ontology components by using specific-domain maps. The knowledge contained in a particular domain, any kind of digital archive, is portrayed by assembling and displaying its ontology components. The novelty of our approach is to offer clustering and visualization features not present in other techniques helping in the semi-automatic construction of ontologies by identifying components from digital archives. We describe here a case study that applies our approach on an academic domain. The specific-domain maps generated are presented and a conceptualized ontology created with the ontology componentes is introduced. Further processing may be carried out on the extracted knowledge to be embedded on the semantic web pages for software agents to use. The ultimate kind of semantic web software applications that we will be seeing are intelligent software agents searching the web to infer and extract new knowledge from the digital archives.

One of the goals of the knowledge puzzle project is to automatically generate a domain ontology from plain text documents and use this ontology as the domain model in computer-based education. This paper describes the generation procedure followed by TEXCOMON, the knowledge puzzle ontology learning tool, to extract concept maps from texts. It also explains how these concept maps are exported into a domain ontology. Data sources and techniques deployed by TEXCOMON for ontology learning from texts are briefly described herein. Then, the paper focuses on evaluating the generated domain ontology and advocates the use of a three-dimensional evaluation: structural, semantic, and comparative. Based on a set of metrics, structural evaluations consider ontologies as graphs. Semantic evaluations rely on human expert judgment, and finally, comparative evaluations are based on comparisons between the outputs of state-of-the-art tools and those of new tools such as TEXCOMON, using the very same set of documents in order to highlight the improvements of new techniques. Comparative evaluations performed in this study use the same corpus to contrast results from TEXCOMON with those of one of the most advanced tools for ontology generation from text. Results generated by such experiments show that TEXCOMON yields superior performance, especially regarding conceptual relation learning.

A concept hierarchy is an integral part of an ontology but it is expensive and time consuming to build. Motivated by this, many unsupervised learning methods have been proposed to (semi-) automatically develop a concept hierarchy. A significant work is the Guided Agglomerative Hierarchical Clustering (GAHC) which relies on linguistic patterns (i.e., hypernyms) to guide the clustering process. However, GAHC still relies on contextual features to build the concept hierarchy, thus data sparsity still remains an issue in GAHC. Artificial Immune Systems are known for robustness, noise tolerance and adaptability. Thus, an extension to the GAHC is proposed by hybridizing it with Artificial Immune Network (aiNet) which we call Guided Clustering and aiNet for Learning Concept Hierarchy (GCAINY). In this paper, we have tested GCAINY using two parameter settings. The first parameter setting is obtained from the literature as a baseline parameter setting and second is by automatic parameter tuning using Particle Swarm Optimization (PSO). The effectiveness of the GCAINY is evaluated on three data sets. For further validations, a comparison between GCAINY and GAHC has been conducted and with statistical tests showing that GCAINY increases the quality of the induced concept hierarchy. The results reveal that the parameters value found by using PSO significantly produce better concept hierarchy than the vanilla parameter. Thus it can be concluded that the proposed approach has greater ability to be used in the field of ontology learning.