Advancing Topic Ontology Learning through Term Extraction

Abstract

This paper presents a novel methodology for topic ontology learning from text documents. The proposed methodology, named OntoTermExtraction is based on OntoGen, a semi-automated tool for topic ontology construction, upgraded by using and an advanced terminology extraction tool in an iterative, semiautomated ontology construction process. This process consists of (a) document clustering to find the nodes in the topic ontology, (b) term extraction from document clusters, (c) populating the term vocabulary and keyword extraction, and (d) choosing the concept names by comparing the best ranked terms with the extracted keywords. The approach is illustrated on a case study analysis of the ILPNet2 publications data. IntroductionOntoGen [1, 2] is a semi-automated, data-driven ontology construction tool, focused on the construction and editing of topic ontologies. In a topic ontology, each node is a cluster of documents, represented by keywords (topics), and nodes are connected by relations (typically, the the SubConcept-Of relation). The system combines text mining techniques with an efficient user interface aimed to reduce user's time and the complexity of ontology construction. In this way it presents a significant improvement in comparison with present manual, and relatively complex ontology editing tools, such as Protégé [3], whose use is hindered by the lack of ontology engineering skills of domain experts constructing the ontology. Concept naming suggestion (i.e. description of a document cluster through a set of relevant terms) plays a central part of the OntoGen system. Concept naming helps the user at evaluating clusters and organizing them hierarchically. This facility is provided by employing unsupervised and supervised methods for generating the suggestions. Despite the well-elaborated and user-friendly approach to concept naming, as currently provided by OntoGen, the approach was limited to single-word keyword suggestions, and by the use of very basic text lemmatization in the OntoGen text preprocessing phase. This paper aims at improving the ontology construction process through improved concept naming, using terminology extraction as implemented in the advanced TermExtractor tool [4,5]. The improved ontology construction process, proposed in this paper, consists of the following steps:• document clustering to find the nodes in the topic ontology, • terminology extraction from document clusters,

Bridging Concept Identification for Constructing Information Networks from Text Documents

“…Heuristic AUC Interval (6) freqDomnProdRel 93,71% 0,40% (21) outFreqRelSum 95,33% 0,35% (13) simDomnRatioMin 93,58% 0,00% (19) outFreqRelRF 95,24% 0,55% (7) tfidfSum 93,58% 0,00% (20) outFreqRelSVM 95,06% 1,26% (9) tfidfDomnProd 93,47% 0,39% (18) outFreqRelCS 94,96% 1,30% (5) freqDomnProd 93,42% 0,44% (17) outFreqSum 94,96% 0,70% (3) freqRatio 93,35% 5,23% (8) tfidfAvg 94,87% 0,00% (23) appearInAllDomn 93,31% 6,69% (15) outFreqRF 94,73% 1,53% (12) simDomnProd 93,27% 0,00% (16) outFreqSVM 94,70% 2,06% (1) freqTerm 93,20% 0,50% (14) outFreqCS 94,67% 1,80% (2) freqDoc 93,19% 0,50% (4) freqDomnRatioMin 94,36% 0,62% (11) simAvgTerm 92,71% 0,00% (10) tfidfDomnSum 93,85% 0,35% (22) random 50,00% 50,00%…”

“…Other interesting approaches to identifying concepts include methods such as KeyGraph [13], which extract terms and concepts with minimal assumptions or background knowledge, even from individual documents. Other alternatives are using domain ontologies which could be, for example, semi-automatically retrieved by a combination of tools such as OntoGen and TermExtractor [9].…”

Concept Extraction Based on Hybrid Approach Combined with Semantic Analysis

“…If term consistently used across domain documents with high frequency [27], it could be regarded as domain concept. Domain Consensus could be defined as: For term t in domain documents D=(d 1 ,d 2 ,……, d n ):…”

“…(2) Except for the natures mentioned above, there are still other nature, including Structural Relevance, and Miscellaneous [27], but unfortunately, they are not operational, so they had not been taken into consideration. Another important nature is Lexical Cohesion [27,28,29] which measures terms with length above 2.In this paper, we hasn't adopted, for concepts extracted out shows that terms are almost short.…”

“…Another important nature is Lexical Cohesion [27,28,29] which measures terms with length above 2.In this paper, we hasn't adopted, for concepts extracted out shows that terms are almost short.…”