The technology in the field of digital media generates huge amounts of tex- tual information every day, so mechanisms to retrieve relevant information are needed. Under these circumstances, many times current web search engines do not provide users with the information they seek, because these search tools mainly use syntax based techniques. However, search engines based on semantic and context information can help overcome some of the limitations of current alternatives. In this paper, we propose a system that takes as input a list of plain keywords pro- vided by a user and translates them into a query expressed in a formal language without ambiguity. Our system discovers the semantics of user keywords by consulting the knowledge represented by many (heterogeneous and distributed) ontologies. Then, context information is used to remove ambiguity and build the most probable query. Our experiments indicate that our system discovers the user's information need bet- ter than traditional search engines when the semantics of the request is not the most popular on the Web.

/pdf/discovering-the-semantics-of-user-keywords-28gj73glml.pdf

Discovering the Semantics of User Keywords

Searching for information in long videos can be a time-consuming experience. In this paper, we describe OnAIR, an ontology-aided information retrieval system applied to retrieve clips from video collections. We used a video collection compiled from interviews with Ana Teixeira, a Brazilian artist. The interviews were made by Paula P. Braga, the domain expert. The interview is developed in the domain of contemporary art and the system uses a domain ontology to expand the queries with related terms. We tested the system with a battery of queries, and we veri.ed that the ontology contributes to the e.ciency improvement in terms of the relevance of retrieved documents. We designed the system to work in a domain-independent way, allowing us to move to other domains by just changing the underlying ontologies and video collections.

/pdf/an-information-retrieval-application-using-ontologies-35g5zx657i.pdf

An information retrieval application using ontologies

Recent research in Cloud Computing and Peer-to-Peer systems for Video-on-Demand (VoD) has focused on multimedia information retrieval, using cloud nodes as video streaming servers and peers as a way to distribute and share the video segments. A key challenge faced by these systems is providing an efficient way to retrieve the information segments descriptor, composed of its metadata and video segments, distributed among the cloud nodes and the Peer-to-Peer (P2P) network. In this paper, we propose a novel Cloud Computing and P2P hybrid architecture for multimedia information retrieval on VoD services that supports random seeking while providing scalability and efficiency. The architecture comprises Cloud and P2P layers. The Cloud layer is responsible for video segment metadata retrieval, using ontologies to improve the relevance of the retrieved information, and for distributing the metadata structures among cloud nodes. The P2P layer is responsible for finding peers that have the physical location of a segment. In this layer, we use trackers, which manage and collect the segments shared among other peers. We also use two Distributed Hash Tables, one to find these trackers and the other to store the information collected in case the tracker leaves the network and another peer needs to replace it. Unlike previous work, our architecture separates cloud nodes and peers responsibilities to manage the video metadata and its segments, respectively. Also, we show via simulations, the possibility of converting any peer to act as a tracker, while maintaining system scalability and performance, avoiding using centralized and powerful servers.

A hybrid cloud-P2P architecture for multimedia information retrieval on VoD services

With the semantic Web progress, encoding of knowledge bases as ontologies has increased. Information retrieval applications are employing this knowledge organization to enhance quality of results by returning documents semantically related and relevant to initial user's query. The proposed fuzzy information retrieval model retrieves information providing a framework to encode a knowledge base composed of multiple related ontologies whose relationships are expressed as fuzzy relations. This knowledge organization is used in a novel method to expand the user initial query and to index the documents in the collection. The model allows the ontologies, as well as the relationships among their concepts, to be represented independently. Experimental results show that the proposed model presents better overall performance when compared with another classical fuzzy-based approach for information retrieval.

Fuzzy Information Retrieval Model Based on Multiple Related Ontologies

It is a big challenge for an information retrieval system (IRS) to interpret the queries made by users, particularly because the common form of query consists of very few terms. Tolerance rough sets models (TRSM), as an extension of rough sets theory, have demonstrated their ability to enrich document representation in terms of semantic relatedness. However, system efficiency is at stake because the weight vector created by TRSM (TRSM-representation) is much less sparse. We mapped the terms occurring in TRSM-representation to terms in the lexicon, hence the final representation of a document was a weight vector consisting only of terms that occurred in the lexicon (LEX-representation). The LEX-representation can be viewed as a compact form of TRSM-representation in a lower dimensional space and eliminates all informal terms previously occurring in TRSM-vector. With these facts, we may expect a more efficient system. We employed recall and precision commonly used in information retrieval to evaluate the effectiveness of LEXrepresentation. Based on our examination, we found that the effectiveness of LEX-representation is comparable with TRSM-representation while the efficiency of LEX-representation should be better than the existing TRSM-representation. We concluded that lexicon-based document representation was another alternative potentially used to represent a document while considering semantics. We are tempted to implement the LEX-representation together with linguistic computation, such as tagging and feature selection, in order to retrieve more relevant terms with high weight. With regard to the TRSM method, enhancing the quality of tolerance class is crucial based on the fact that the ∗Address for correspondence: Faculty of Mathematics, Informatics and Mechanics, University of Warsaw, Banacha 2, 02-097 Warsaw, Poland 28 G. Virginia and H. S. Nguyen / Lexicon-based Document Representation TRSM method is fully reliant on the tolerance classes. We plan to combine other resources such as Wikipedia Indonesia to generate a better tolerance class.

Lexicon-based Document Representation

In this paper we propose a multilingual extension for OnAIR which is an ontology-aided information retrieval system applied to retrieve clips from a video collection. The multilingual extension basically involves allowing the user to search in several languages in a multilingual video collection. Particularly, the pair of languages we work in this paper are English and Portuguese. In order to perform query translation we use a statistical machine translation approach. Our experiments show that the multilingual system is capable of achieving almost the same quality of that obtained by the monolingual system.

Initial approaches on Cross-Lingual Information Retrieval using Statistical Machine Translation on user-queries

Looking for information in long videos can be a time-consuming ex- perience. The ONAIR system was developed in order to allow the users to find the information they want through queries in written natural language. We im- plemented the vector space model for information retrieval and used ontologies to improve the retrieval. In this paper we describe the system architecture, which makes strong use of patterns. We then describe an application of the system to the domain of Object-Oriented patterns in a course lectured by Joseph Yoder at the University of S˜ ao Paulo.

/pdf/a-pattern-based-tool-for-learning-design-patterns-4519v94n8j.pdf

A Pattern-based Tool for Learning Design Patterns

In this paper we propose a multilingual extension for OnAIR which is an ontology-aided information retrieval system applied to retrieve clips from a video collection. The multilingual extension basically involves allowing the user to search in several languages in a multilingual video collection. Particularly, the pair of languages we work in this paper are English and Portuguese. In order to perform query translation we use a statistical machine translation approach. Our experiments show that the multilingual system is capable of achieving almost the same quality of that obtained by the monolingual system. Resumo. Neste trabalho, propomos uma extensão multilingue para OnAir que é um sistema de recuperação de informação auxiliado por uma ontologia. O sistema é usado para recuperar clips de uma coleção de vı́deos. A extensão multilingue permite ao usuário fazer buscas em duas lı́nguas em uma coleção de vı́deo multilingue. Particularmente, o par de lı́nguas que trabalhamos neste artigo são Inglês e Português. Para realizar a conversão de consulta, usamos uma abordagem estatı́stica de tradução. As nossas experiências mostraram que o sistema multilingue é capaz de atingir quase a mesma qualidade do obtido pelo sistema monolingue.

http://ceur-ws.org/Vol-938/ontobras-most2012_paper2.pdf

Christian Paz-Trillo

Papers

An information retrieval application using ontologies

Initial approaches on Cross-Lingual Information Retrieval using Statistical Machine Translation on user-queries

A Pattern-based Tool for Learning Design Patterns

Initial Approaches on Cross-Lingual Information Retrieval Using SMT on User-Queries.