Antonina Dattolo

The role of tags for recommendation: A survey

Social tagging is an innovative and powerful mechanism introduced with Web 2.0: it shifts the task of classifying resources from a reduced set of knowledge engineers to the wide set of Web users. Users of social tagging systems define personal classifications which can be used by other peers for browsing available resources. However, due to the absence of rules for managing the tagging process, and to the lack of predefined schemas or structures for inserting metadata and relationships among tags, current user generated classifications dop not produce sound taxonomies. This is a strong limitation which prevents an effective and informed resource sharing. For this reason researchers are modeling innovative recommender systems capable to better support tagging, browsing, and searching for new resources. This paper is a survey which discusses the role of tags in recommender systems: starting from social tagging systems, we analyze various techniques for suggesting content and we introduce the approaches exploited for proposing tags for classifying resources, considering both personalized and not-personalized recommendation.

Collaborative version control in an agent-based hypertext environment

Abstract In this work we discuss a number of issues for the design of hypertext systems in an agent-based model of computation. We examine how the “traditional” fundamental concepts which are at the basis of the design of hypertexts can be re-visited under a new perspective of collaborative expert agents. The paper presents how some principles of high-level concurrent programming are applied as new methodologies for the design and development of complex software, such as hypertext systems. By adopting an agent-based framework, we gain powerful control on version management that presents considerable difficulties for the development of hypertext systems; a general distributed version control mechanism is applied, without significant differences, both in single-user and in collaborative multiuser mode. In both cases, the underlying hypertext architecture is defined in terms of computational agents interacting each other in order to accomplish common goals. In this paper we present a first-level prototype implemented in a concurrent object-oriented language, realized on the top of the Common Lisp Object System.

An integrated approach to discover tag semantics

Tag-based systems have become very common for online classification thanks to their intrinsic advantages such as self-organization and rapid evolution. However, they are still affected by some issues that limit their utility, mainly due to the inherent ambiguity in the semantics of tags. Synonyms, homonyms, and polysemous words, while not harmful for the casual user, strongly affect the quality of search results and the performances of tag-based recommendation systems. In this paper we rely on the concept of tag relatedness in order to study small groups of similar tags and detect relationships between them. This approach is grounded on a model that builds upon an edge-colored multigraph of users, tags, and resources. To put our thoughts in practice, we present a modular and extensible framework of analysis for discovering synonyms, homonyms and hierarchical relationships amongst sets of tags. Some initial results of its application to the delicious database are presented, showing that such an approach could be useful to solve some of the well known problems of folksonomies.

Recommending New Tags Using Domain-Ontologies

Tagging is a representative activity of social Web, useful for organizing information into knowledge. This activity presents some open issues, due in the majority to the manual insertion of tags. On the other hand, domain ontology is a specification of the conceptualization of a domain in terms of concepts, attributes and relations. Domain ontologies have a good potential to improve information organization, management and understanding. In this paper, we propose an automated approach for recommending new tags for Web resources by using domain ontologies and key-phrases. The proposed approach is implemented in the PIRATES framework, a prototype system for personalized content retrieval, annotation, and classification. Our approach is then explained with a simple use-case scenario.

Structural patterns for descriptive documents

Combining expressiveness and plainness in the design of web documents is a difficult task. Validation languages are very powerful and designers are tempted to over-design specifications. This paper discusses an offbeat approach: describing any structured content of any document by only using a very small set of patterns, regardless of the format and layout of that document. The paper sketches out a formal analysis of some patterns, based on grammars and language theory. The study has been performed on XML languages and DTDs and has a twofold goal: coding empirical patterns in a formal representation, and discussing their completeness.

Supporting Personalized User Concept Spaces and Recommendations for a Publication Sharing System

Current publication sharing systems weakly support creation and personalization of customized user concept spaces. Focusing the attention on the user, SharingPapers, the adaptive publication sharing system proposed in this paper, allows users to organize documents in flexible and dynamic concept spaces; to merge their concept map with a social network connecting people involved in the domain of interest; to support knowledge expansion generating adaptive recommendations. SharingPapers presents a multi-agent architecture and proposes a new way of representing user profiles, their evolution and views of them.

A New Domain Independent Keyphrase Extraction System

In this paper we present a keyphrase extraction system that can extract potential phrases from a single document in an unsupervised, domain-independent way. We extract word n-grams from input document. We incorporate linguistic knowledge (i.e., part-of-speech tags), and statistical information (i.e., frequency, position, lifespan) of each n-gram in defining candidate phrases and their respective feature sets. The proposed approach can be applied to any document, however, in order to know the effectiveness of the system for digital libraries, we have carried out the evaluation on a set of scientific documents, and compared our results with current keyphrase extraction systems.

Visualizing personalized views in virtual museum tours

We propose a multi-agent adaptive system to support tours of virtual museums. The system stores userspsila personalized views in zz-structures, particular data structures capable of representing both hypertext information and contextual interconnections among different information. We present an extension of the standard zz-structure model in terms of computational agents. These agents cooperate and collaborate in order to help users visualizing their personalized views. The power of this new model resides in the (partially limited) level of freedom users have for the dynamical choice, based on some present interest or necessity, of their navigational path inside the virtual museum.

Sentiment Classification for the Italian Language: A Case Study on Movie Reviews

We consider the problem of tracking the opinion polarity, in terms of positive or negative orientation, expressed in documents written in natural language and extracted from a heterogeneous set of Web sources. More specifically, we focus our attention on the movie reviews domain. We are interested in evaluating the performance obtained by a set of high performance opinion polarity classifiers for the Italian language. Classification of polarity expressed by the input documents is achieved by means of several sets of specialized autonomous or interacting agents, devoted, respectively, to document gathering, classification and visualization. In particular the results of opinion analysis are represented by means of a graphical interface, where a multi agent based implementation of zz-structures is exploited to offer graph-centric views and navigation of results. The specific experimental evaluation performed so far shows an accuracy level, which is higher than previous results reported in the literature.

Active distributed framework for adaptive hypermedia

Navigation through large hypermedia information spaces is complex and is an important application area for adaptive hypermedia systems. User navigation can be best supported when the design of the hypermedia system is embedded in an evolutionary process model that takes into account the decentralization of data sources and the variety of users. The paper deals with distributed frameworks for open hypermedia systems; it focuses on the design work done to make adaptive an existing actor-based architecture for hypermedia. The approach follows the initial design approach used in the definition of the hypermedia platform, i.e. the actor-based computational model. We present in detail the new actor classes and the cooperative schemes which allow adaptation within the resulting architecture.