Sharif Digital Repository

Currently, the main portion of knowledge is stored in electronic texts and documents and for transferring that knowledge effectively, we must use proper methods to gather and retrieve relevant information. Ontologies provide means to produce structured documents and use intelligent search instead of keyword search. Ontology defines the common words and concepts used to describe and represent an area of knowledge. However, developing ontologies is a time consuming and labor work. So, many ontology developers try to facilitate and speed up this process by reusing other resources. In fact, thesaurus contains semantic information and hierarchical structure that make it an appropriate resource... 

The Putting human commonsense knowledge into computers has always been a long standing dream of artificial intelligence (AI). The cost of several tens of millions of dollars and time have been covered so that the computers could know about "objects falling, not rising.", "running is faster than walking". The large database was built, automated and semi-automated methods were introduced and volunteers' efforts were utilized to achieve this, but an automated, high-throughput and low-noise method for commonsense collection still remains as the holy grail of AI. The aim of this study was to build commonsense knowledge ontology using three approaches namely Hearst method, machine translation and... 

Several metrics have been proposed for recognition of relationships between elements of two Ontologies. Many of these methods select a number of such metrics and combine them to extract existing mappings. In this article, we present a method for selection of more effective metrics - based on data mining techniques. Furthermore, by having a set of metrics, we suggest a data-mining-like means for combining them into a better ontology alignment  

From ancient times to 19th century geometry symbolized the essence of mathematical thinking and method, but modern philosophy of mathematics seems to have marginalized the philosophical status of geometry. The roots of this transformation will be sought in the ascendance of logical foundations in place of intuitive primacy as the cornerstone of mathematical certainty in the late 19th century. Nevertheless, geometry and geometrical thinking, in multiple manifestations, have continued to occupy a central place in the practice of mathematics proper. We argue that this, together with advances in the neuroscience of mathematical processes, calls for an expansion of the present limited remit of... 

Ontology matching is a process for selection of a good alignment across entities of two (or more) ontologies. This can be viewed as a two-phase process of (1) applying a similarity measure to find the correspondence of each pair of entities from two ontologies, and (2) extraction of an optimal or near optimal mapping. This paper is focused on the second phase and introduces our evolutionary approach for that. To be able to do so, we need a mechanism to score different possible mappings. Our solution is a weighting mechanism named coincidence-based weighting. A genetic algorithm is then introduced to create better mappings in successive iterations. We will explain how we code a mapping as... 

Ontology Alignment is a process for finding related entities of different ontologies. This paper discusses the results of our research in this area. One of them is a formulation for a new structural measure which extends famous related works. In this measure with a special attention to the transitive properties, it is tried to increase recall with less harm on precision. Second contribution is a new method for compound measure creation without any need to the mapping extraction phase. Effectiveness of these ideas is discussed and quantitative evaluations are explained in this paper. Copyright © 2006, Hassan Abolhassani, Babak Bagheri Hariri, & Seyed H. Haeri  

Today, public awareness about the benefits of using ontologies in information processing and the semantic web has increased. Since ontologies are useful in various applications, many large ontologies have been developed so far. But various areas like publication, maintenance, validation, processing, and security policies need further research. One way to better tackle these areas is to partition large ontologies into sub partitions. In this paper, we present a new method to partition large ontologies. For the proposed method, three steps are required: (1) transforming an ontology to a weighted graph, (2) partitioning the graph with an algorithm which recognizes the most important concepts,... 

Information extraction and refinement systems rely on a set of extraction patterns in pattern based approaches in order to construct taxonomies and finding instances of concepts in a corpus. Pattern based methods have high precision and suffer from low recall. Pattern generalization and modeling techniques can increase matching power and decrease GAP between training and test data  

Fetching Lexico-Syntactic patterns from text rely on pairs of words (positive instances) that represent the target relation, and finding their simultaneous occurrence in text corpus. Due to existence of WordNet thesaurus (which contains the semantic relationship between words), collecting positive instances is easy. In non-english languages, it's hard to collect large number of positive instances in various contexts. We investigated some new ideas for collecting them in Persian language and finally run the best one and collected approximately 6,000 positive instances  

Taxonomy learning is a prerequisite step for ontology learning. In order to create a taxonomy, first of all, existing 'is-a' relations between words should be extracted. A known way to extract 'is-a' relations is finding lexicosyntactic patterns in large text corpus. Although this approach produces results with high precision but it suffers from low values of recall. Furthermore developing a comprehensive set of patterns is tedious and cumbersome. In this paper, firstly, we introduce an approach for developing lexico-syntactic patterns automatically using the snippets of search engine results and then, challenge the law recall of this approach using a combined model, which is based on... 

As semantic web grows, security concerns increase. One concern is controlling accesses to resources in this environment. In order to infer whether the access is allowed or not, different information of different entities including contextual information should be involved. From access control point of view, we divide the entities in semantic web into three categories: resources (objects), requesters (subjects), and environment (infrastructure, time, and location). In this paper, we present a semantic-based context-aware access control framework to be applied in semantic web, considered as a multi-domain environment. To handle context information in the framework, we propose a context... 

Recent evaluations of mapping systems show that lack of background knowledge, most often domain specific knowledge, is one of the key problems of mapping systems these days. In fact, at present, most state of the art systems, for the tasks of mapping large ontologies, perform not with such high values of recall (~ 30%), because they mainly rely on label and structure based similarity measures. Disregarding context knowledge in ontology mapping is another drawback that almost all current approaches suffer from. In this paper we use the semantic web as background knowledge and introduce a novel approach for capturing context knowledge from the ontology for improving mapping results. We have... 

Ontology Aligning is an answer to the problem of handling heterogenous information on different domains. After application of some measures, one reaches a set of similarity values. The final goal is to extract mappings. Our contribution is to introduce a new genetic algorithm (GA) based extraction method. The GA, employs a structured based weighting model, named "coincidence based model", as its fitness function. In the first part of the paper, some preliminaries and notations are given and then we introduce the coincidence based weighting. In the second part the paper discusses the details of the devised GA with the evaluation results for a sample dataset  

In a wireless sensor network, with numerous sensor nodes, a huge volume of data is produced. For the massive data generated by sensor networks to be understood by the machine, semantic web technologies such as ontology need to be applied. On the other hand, since the end users intend to access the high-level physical entities information monitored by the sensor network, the applicability of sensor networks can be enhanced through proposing a strategy to extract the data based on entities instead of extracting the raw sensor data. Hence, in this paper firstly, ontology will be presented for physical entities semantic modelling (PESM). Secondly, an appropriate strategy will be offered to... 

As Gregory Bateson, the British social scientists, once said "You can't live without an eraser", the process of learning is not a unidirectional way toward accumulation of knowledge, rather, organizations should critically evaluate the value of their knowledge and if necessary, actively forget no-longer-useful knowledge. A brief review of related literature shows that the concept of knowledge active forgetting (KAF) is a crucial aspect of organizational learning and knowledge management. However, many aspects of this topic deserve more attention by researchers. The main gap in the literature seems to be related to the linkage between peculiar aspects of knowledge (such as stickiness, context... 

Taxonomy learning is one of the major steps in ontology learning process. Manual construction of taxonomies is a time-consuming and cumbersome task. Recently many researchers have focused on automatic taxonomy learning, but still quality of generated taxonomies is not satisfactory. In this paper we have proposed a new compound similarity measure. This measure is based on both knowledge poor and knowledge rich approaches to find word similarity. We also used Neural Network model for combination of several similarity methods. We have compared our method with simple syntactic similarity measure. Our measure considerably improves the precision and recall of automatic generated taxonomies. ©...