A reliable linguistic filter for Farsi term extraction

Ontology plays a very important role in supporting knowledge-based applications. In cloud computing, ontology learning technology is facing new challenges in dealing with heterogeneous data sources from different domains and researchers, which may contain various particular concepts and relations. Traditional ontology learning frameworks usually focus only on the extraction of concepts and taxonomic relations from the multi-structured corpus. However, former researches rarely studied the interactions during ontology learning process among different researchers. Lack of interactions among people who build ontology in different domains may cause inconsistent ontology. Besides, lack of incentive during the ontology building process will also result in low efficiency. To address these challenges, this paper specifies a novel solution to perform ontology learning. The solution includes a service-oriented ontology interaction framework, a service-oriented ontology learning strategy. It shows that it advances ontology learning to a higher level of performance and portability with a number of experiments in demo system.

We report on TExSIS, a flexible bilingual terminology extraction system that uses a sophisticated chunk-based alignment method for the generation of candidate terms, after which the specificity of the candidate terms is determined by combining several statistical filters. Although the set-up of the architecture is largely language-independent, we present terminology extraction results for four different languages and three language pairs. Gold standard data sets were created for French-Italian, French-English and French-Dutch, which allowed us not only to evaluate precision, which is common practice, but also recall. We compared the TExSIS approach, which takes a multilingual perspective from the start, with the more commonly used approach of first identifying term candidates monolingually and then aligning the source and target terms. A comparison of our system with the LUIZ approach described by Vintar (2010) reveals that TExSIS outperforms LUIZ both for monolingual and bilingual terminology extraction. Our results also clearly show that the precision of the alignment is crucial for the success of the terminology extraction. Furthermore, based on the observation that the precision scores for bilingual terminology extraction outperform those of the monolingual systems, we conclude that multilingual evidence helps to determine unithood in less related languages.

The C/NC value method combines linguistic and statistical information for the automatic extraction of multi-word technical terms from machine readable corpora. C-value enhances the commonly used statistical measure of frequency of occurrence for term extraction, making it sensitive to a particular type of multi-word terms, nested terms. Nested terms are those which also exist as substrings of other terms. Consider the following term: mitogen-activated protein kinase. Valid substrings of the longer term are also extracted, e.g. mitogen-activated protein, protein kinase. NC/value incorporates contextual information in the form of statistical (weights) and linguistic information improving the performance of C/value. Deeper forms of contextual information (semantic knowledge) have also been used. The measures have been applied to medical corpora in English and have been also adapted for Japanese. We applied the C/NC value into a collection of 2000 abstracts from MEDLINE. The domain area was that of nuclear receptors. Before applying the first measure (C-value) we tagged the corpus using MXPOST tagger, freely available from the University of Pennsylvania. The tagger was not trained for our corpus. The wordforms were stemmed using Porter’s stemming algorithm. Stemming was used to deliver better results for statistical measures. The same corpus was run on a different tagger based on constraint based grammars, ENGCG. We used a domain specific stop list which was produced after the first results of C/NC value, containing single words (e.g. dramatic, data) and multiword units (e.g. brief review). The following linguistic filters were applied, based on commonly used term formation patterns:

Requirements engineering, the first phase of any software development project, is the Achilles' heel of the whole development process, as requirements documents are often inconsistent and incomplete. In industrial requirements documents, natural language is the main presentation means. This results in the fact that the requirements documents are imprecise, incomplete, and inconsistent. A viable way to detect inconsistencies and omissions in documents is to extract system models from them. In our previous work we developed approaches translating textual scenarios to message sequence charts (MSCs) and textual descriptions of automata to automata themselves. It turned out that these approaches are highly sensitive to proper definition of terms (communicating objects for MSCs, states for automata). The goal of the presented paper is a systematic comparison of different term extraction heuristics, as a preliminary stage of MSC or automata extraction. The extracted terms were declared to communicating objects (in the case of MSCs) or to states (in the case of automata). The heuristics were compared on the basis of correctness of resulting MSCs and automata. We came to the conclusion that named entity recognition is the best performing technique for term extraction from requirements documents.

We design a new technique for the distributional semantic modeling with a neural network-based approach to learn distributed term representations (or term embeddings) – term vector space models as a result, inspired by the recent ontology-related approach (using different types of contextual knowledge such as syntactic knowledge, terminological knowledge, semantic knowledge, etc.) to the identification of terms (term extraction) and relations between them (relation extraction) called semantic pre-processing technology – SPT. Our method relies on automatic term extraction from the natural language texts and subsequent formation of the problem-oriented or application-oriented (also deeply annotated) text corpora where the fundamental entity is the term (includes non-compositional and compositional terms). This gives us an opportunity to changeover from distributed word representations (or word embeddings) to distributed term representations (or term embeddings). The main practical result of our work is the development kit (set of toolkits represented as web service APIs and web application), which provides all necessary routines for the basic linguistic pre-processing and the semantic pre-processing of the natural language texts in Ukrainian for future training of term vector space models.

Ontology learning is a technology. Ontology learning can be used to establish ontology automatically or semi-automatically by introducing the ontology engineering and machine learning technology and many other sciences and technologies. The ontology learning technology which is proposed in our paper is to reduce the time of building an entire ontology. Our paper presents an Ontology Learning model which will enhance the efficiency of extraction concept, and enhance the efficiency of ontology building. It includes several aspects, and area concept extraction is the main aspect of all. The model combines personalized recommendation with concept extraction and realizes a more accurate and stable domain concept extraction method. We describe these techniques and report the results of the experiment examining its effectiveness and efficiency.

Most approaches to ontology learning combine techniques from different areas (hybrid approaches) to increase the efficiency of the ontology learning process. However, the results from the ontology learning process do not fully satisfy the users at present. An important problem is that there is a lack of quantitative and comparative data about the efficiency of techniques and technique combinations applied to ontology learning. In this paper we present a quantitative comparison of technique combinations for concept extraction and a software system (OntoLancs) to support the evaluation of techniques. By applying OntoLancs, users are able to assist the process of building ontologies by semi- automatically acquiring concepts from large-scale domain document collections and experiment with different combinations of knowledge acquisition techniques to refine and organize domain concepts into a taxonomy. Quantitative and comparative studies about the performance of several techniques and user experiences indicate the applicability and usefulness of our approach.

Ontology is a representation term used to describe and represent a domain of knowledge. Manually ontology development is currently considered complex, requiring a lot of time and effort. This research was proposed to develop methods to build automatic domain ontology bilingual in Indonesian and English by using corpus and ontology design patterns (ODPs) in tuberculosis disease. In this study, the methods used were to combine ontology learning from text and correspond with ontology design patterns to decrease the role of expert knowledge. The methods in this research consist of six stages: (i) Term and relation extraction (ii) Corresponding with Tuberculosis glossary (iii) Corresponding the ontology design patterns (iv) Score computation similarity term and relations with ODPs (v) Ontology Building (vi) Ontology evaluation. The results of ontology construction were 361 terms and 44 relations with 260 terms were added. The calculation accuracy of ontology construction was 71%. Ontology construction had higher complexity and shorter time as well as decreases the role of the expert knowledge which proof that the automatic ontology evaluation is better than manual ontology construction.

Ontology learning from unstructured text has become a critical task for knowledge-driven applications in Big Data and Artificial Intelligence. While significant advances have been made in the automatic extraction of concepts and relations using neural and Transformer-based models, the generation of formal Description Logic axioms required for constructing logically consistent and computationally tractable ontologies remains largely underexplored. This paper puts forward a novel pipeline for automated axiom generation through schema mapping. Our paper introduces three key innovations: a deterministic mapping framework that guarantees logical consistency (unlike stochastic Large Language Models); guaranteed formal consistency verified by OWL reasoners (unaddressed by prior statistical methods); and a transparent, scalable bridge from neural extractions to symbolic logic, eliminating manual post-processing. Technically, the pipeline builds upon the outputs of a Transformer-based fusion model for joint concept and relation extraction. We then map lexical relational phrases to formal ontological properties through a lemmatization-based schema alignment step. Entity typing and hierarchical induction are then employed to infer class structures, as well as domain and range constraints. Using RDFLib and structured data processing, we transform the extracted triples into both assertional (ABox) and terminological (TBox) axioms expressed in Description Logic. Experimental evaluation on benchmark datasets (Conll04 and NYT) demonstrates the efficacy of the approach, with expert validation showing high acceptance rates (>95%) and reasoners confirming zero inconsistencies. The pipeline thus establishes a reliable, scalable foundation for automated ontology learning, advancing the field from extraction to formally verifiable knowledge base construction.

Traditional text mining systems employ shallow parsing techniques and focus on concept extraction and taxonomic relation extraction. This paper presents a novel system called CRCTOL for mining rich semantic knowledge in the form of ontology from domain-specific text documents. By using a full text parsing technique and incorporating both statistical and lexico-syntactic methods, the knowledge extracted by our system is more concise and contains a richer semantics compared with alternative systems. We conduct a case study wherein CRCTOL extracts ontological knowledge, specifically key concepts and semantic relations, from a terrorism domain text collection. Quantitative evaluation, by comparing with a state-of-the-art ontology learning system known as text-to-onto, has shown that CRCTOL produces much better precision and recall for both concept and relation extraction, especially from sentences with complex structures.

Enhancing relevant concepts extraction for ontology learning using domain time relevance

At present, the focus of ontology learning is on the extraction of concepts and relations. The relation extraction is divided into hierarchical relation extraction and non-hierarchical relation extraction, and hierarchical relation extraction is the basis of non-hierarchical relation extraction. In this paper, the method of mixed hierarchical clustering is used to extract and classify domain concepts from textual corpora in different fields of Uyghur language. Due to the limitations of the text content selected in this paper and the difficulty of merging nodes by hierarchical clustering, although the results are not ideal, they basically conform to the ontology hierarchy form. The experimental results show that the method is feasible, and the factors such as increasing the vector dimension and the text content, can further improve the accuracy of the domain concept and relation.

Model tuning or prompt Tuning? a study of large language models for clinical concept and relation extraction

Creating ontologies is an essential while challenging task to be performed by either a human or a system: on one hand it is excessively burdensome for a human operator, on the other it is very complex also for a machine due to the not negligible amount of that it must be able to manage. In the last years, some attempts have been made to automate this process, but at present, due to the large number of aspects to be covered in the automatic creation of an ontology (such as Domain terminology extraction, Concept discovery, Concept hierarchy derivation, ) satisfactory solutions have not been reached yet. In order to produce efficient tools for both creation and enrichment of ontologies, the participation of the human in such a process still seems necessary. Our approach, that foresees a broader framework for ontology learning, is based by first on the automatic extraction of triples from heterogeneous sources, then on the presentation of the most reliable triples to the human operator for validation purposes. The system provides the user with a series of graphical representations that can give him an overview of the level of uncertainty of the automatically generated ontology. Then provides the user with the possibility to perform SPARQL what-if queries, (i.e. assuming as true the triples filtered according to the level of confidence, the source and the structure of the triples).

Arabic ontology extraction model from unstructured text