This article presents YAGO, a large ontology with high coverage and precision. YAGO has been automatically derived from Wikipedia and WordNet. It comprises entities and relations, and currently contains more than 1.7 million entities and 15 million facts. These include the taxonomic Is-A hierarchy as well as semantic relations between entities. The facts for YAGO have been extracted from the category system and the infoboxes of Wikipedia and have been combined with taxonomic relations from WordNet. Type checking techniques help us keep YAGO’s precision at 95% – as proven by an extensive evaluation study. YAGO is based on a clean logical model with a decidable consistency. Furthermore, it allows representing n-ary relations in a natural way while maintaining compatibility with RDFS. A powerful query model facilitates access to YAGO’s data.

Abstract. Textual patterns have been used effectively to extract information from large text collections. However they rely heavily on textual redundancy in the sense that facts have to be mentioned in a similar manner in order to be generalized to a textual pattern. Data sparseness thus becomes a problem when trying to extract information from hardly redundant sources like corporate intranets, encyclopedic works or scientific databases. We present results on applying a weakly supervised pattern induction algorithm to Wikipedia to extract instances of arbitrary relations. In particular, we apply different configurations of a basic algorithm for pattern induction on seven different datasets. We show that the lack of redundancy leads to the need of a large amount of training data but that integrating Web extraction into the process leads to a significant reduction of required training data while maintaining the accuracy of Wikipedia. In particular we show that, though the use of the Web can have similar effects as produced by increasing the number of seeds, it leads overall to better results. Our approach thus allows to combine advantages of two sources: The high reliability of a closed corpus and the high redundancy of the Web. 1

State-of-the-art research on automated learning of ontologies from text currently focuses on inexpressive ontologies. The acquisition of complex axioms involving logical connectives, role restrictions, and other expressive features of the Web Ontology Language OWL remains largely unexplored. In this paper, we present a method and implementation for enriching inexpressive OWL ontologies with expressive axioms which is based on a deep syntactic analysis of natural language definitions. We argue that it can serve as a core for a semi-automatic ontology engineering process supported by a methodology that integrates methods for both ontology learning and evaluation. The feasibility of our approach is demonstrated by generating complex class descriptions from Wikipedia definitions and from a fishery glossary provided by the Food and Agriculture Organization of the United Nations.

Abstract. We present a class of models that are discriminatively trained to directly map from the word content in a query-document or documentdocument pair to a ranking score. Like Latent Semantic Indexing (LSI), our models take account of correlations between words (synonymy, polysemy). However, unlike LSI our models are trained with a supervised signal directly on the task of interest, which we argue is the reason for our superior results. We provide an empirical study on Wikipedia documents, using the links to define document-document or query-document pairs, where we obtain state-of-the-art performance using our method. Key words: supervised, semantic indexing, document ranking 1

Abstract. In this paper, an experiment is presented for the automatic annotation of several semantic relationships in the Wikipedia, a collaborative on-line encyclopedia. The procedure is based on a methodology for the automatic discovery and generalisation of lexical patterns that allows the recognition of relationships among concepts. This methodology requires as information source any written, general-domain corpora and applies natural language processing techniques to extract the relationships from the textual corpora. It has been tested with eight different relations from the Wikipedia corpus. 1

This paper describes automatic construction a freely-available wordnet for French (WOLF) based on Princeton WordNet (PWN) by using various multilingual resources. Polysemous words were dealt with an approach in which a parallel corpus for five languages was word-aligned and the extracted multilingual lexicon was disambiguated with the existing wordnets for these languages. On the other hand, a bilingual approach sufficed to acquire equivalents for monosemous words. Bilingual lexicons were extracted from Wikipedia and thesauri. The results obtained from each resource were merged and ranked according to the number of resources yielding the same literal. Automatic evaluation of the merged wordnet was performed with the French WordNet (FREWN). Manual evaluation was also carried out on a sample of the generated synsets. Precision shows that the presented approach has proved to be very promising and applications to use the created wordnet are already intended. 1.

Abstract. Extracting semantic relations is of great importance for the creation of the Semantic Web content. It is of great benefit to semi-automatically extract relations from the free text of Wikipedia using the structured content readily available in it. Pattern matching methods that employ information redundancy cannot work well since there is not much redundancy information in Wikipedia, compared to the Web. Multi-class classification methods are not reasonable since no classification of relation types is available in Wikipedia. In this paper, we propose PORE (Positive-Only Relation Extraction), for relation extraction from Wikipedia text. The core algorithm B-POL extends a state-of-the-art positive-only learning algorithm using bootstrapping, strong negative identification, and transductive inference to work with fewer positive training examples. We conducted experiments on several relations with different amount of training data. The experimental results show that B-POL can work effectively given only a small amount of positive training examples and it significantly outperforms the original positive learning approaches and a multi-class SVM. Furthermore, although PORE is applied in the context of Wikipedia, the core algorithm B-POL is a general approach for Ontology Population and can be adapted to other domains. Keywords: Relation Extraction, Ontology Population, Positive-Only Learning 1

Abstract. Most new words, or neologisms, bubble beneath the surface of widespread usage for some time, perhaps even years, before gaining acceptance in conventional print dictionaries [1]. A shorter, yet still significant, delay is also evident in the life-cycle of NLP-oriented lexical resources like WordNet [2]. A more topical lexical resource is Wikipedia [3], an open-source community-maintained encyclopedia whose headwords reflect the many new words that gain recognition in a particular linguistic sub-culture. In this paper we describe the principles behind Zeitgeist, a system for dynamic lexicon growth that harvests and semantically analyses new lexical forms from Wikipedia, to automatically enrich WordNet as these new word forms are minted. Zeitgeist demonstrates good results for composite words that exhibit a complex morphemic structure, such as portmanteau words and formal blends [4, 5]. 1

In this article we present Supervised Semantic Indexing (SSI) which defines a class of nonlinear (quadratic) models that are discriminatively trained to directly map from the word content in a query-document or document-document pair to a ranking score. Like Latent Semantic Indexing (LSI), our models take account of correlations between words (synonymy, polysemy). However, unlike LSI our models are trained from a supervised signal directly on the ranking task of interest, which we argue is the reason for our superior results. As the query and target texts are modeled separately, our approach is easily generalized to different retrieval tasks, such as crosslanguage retrieval or online advertising placement. Dealing with models on all pairs of words features is computationally challenging. We propose several improvements to our basic model for addressing this issue, including low rank (but diagonal preserving) representations, correlated feature hashing (CFH) and sparsification. We provide an empirical study of all these methods on retrieval tasks based on Wikipedia documents as well as an Internet advertisement task. We obtain state-of-the-art performance while providing realistically scalable methods.

Abstract. Semantics can be integrated in to search processing during both document analysis and querying stages. We describe a system that incorporates both, semantic annotations of Wikipedia articles into the search process and allows for rich annotation search, enabling users to formulate queries based on their knowledge about how entities relate to one another while simultaneously retaining the freedom of free text search where appropriate. The outcome of this work is an application consisting of semantic annotators, an extended search engine and an interactive user interface. 1