Mario Cataldi

Emerging topic detection on Twitter based on temporal and social terms evaluation

Twitter is a user-generated content system that allows its users to share short text messages, called tweets, for a variety of purposes, including daily conversations, URLs sharing and information news. Considering its world-wide distributed network of users of any age and social condition, it represents a low level news flashes portal that, in its impressive short response time, has the principal advantage. In this paper we recognize this primary role of Twitter and we propose a novel topic detection technique that permits to retrieve in real-time the most emergent topics expressed by the community. First, we extract the contents (set of terms) of the tweets and model the term life cycle according to a novel aging theory intended to mine the emerging ones. A term can be defined as emerging if it frequently occurs in the specified time interval and it was relatively rare in the past. Moreover, considering that the importance of a content also depends on its source, we analyze the social relationships in the network with the well-known Page Rank algorithm in order to determine the authority of the users. Finally, we leverage a navigable topic graph which connects the emerging terms with other semantically related keywords, allowing the detection of the emerging topics, under user-specified time constraints. We provide different case studies which show the validity of the proposed approach.

CoSeNa: a context-based search and navigation system

Most of the existing document and web search engines rely on keyword-based queries. To find matches, these queries are processed using retrieval algorithms that rely on word frequencies, topic recentness, document authority, and (in some cases) available ontologies. In this paper, we propose an innovative approach to exploring text collections using a novel keywords-by-concepts (KbC) graph, which supports navigation using domain-specific concepts as well as keywords that are characterizing the text corpus. The KbC graph is a weighted graph, created by tightly integrating keywords extracted from documents and concepts obtained from domain taxonomies. Documents in the corpus are associated to the nodes of the graph based on evidence supporting contextual relevance; thus, the KbC graph supports contextually informed access to these documents. In this paper, we also present CoSeNa (Context-based Search and Navigation) system that leverages the KbC model as the basis for document exploration and retrieval as well as contextually-informed media integration.

Personalized emerging topic detection based on a term aging model

Twitter is a popular microblogging service that acts as a ground-level information news flashes portal where people with different background, age, and social condition provide information about what is happening in front of their eyes. This characteristic makes Twitter probably the fastest information service in the world. In this article, we recognize this role of Twitter and propose a novel, user-aware topic detection technique that permits to retrieve, in real time, the most emerging topics of discussion expressed by the community within the interests of specific users. First, we analyze the topology of Twitter looking at how the information spreads over the network, taking into account the authority/influence of each active user. Then, we make use of a novel term aging model to compute the burstiness of each term, and provide a graph-based method to retrieve the minimal set of terms that can represent the corresponding topic. Finally, since any user can have topic preferences inferable from the shared content, we leverage such knowledge to highlight the most emerging topics within her foci of interest. As evaluation we then provide several experiments together with a user study proving the validity and reliability of the proposed approach.

Good location, terrible food: detecting feature sentiment in user-generated reviews

A growing corpus of online informal reviews is generated every day by non-experts, on social networks and blogs, about an unlimited range of products and services. Users do not only express holistic opinions, but often focus on specific features of their interest. The automatic understanding of “what people think” at the feature level can greatly support decision making, both for consumers and producers. In this paper, we present an approach to feature-level sentiment detection that integrates natural language processing with statistical techniques, in order to extract users’ opinions about specific features of products and services from user-generated reviews. First, we extract domain features, and each review is modelled as a lexical dependency graph. Second, for each review, we estimate the polarity relative to the features by leveraging the syntactic dependencies between the terms. The approach is evaluated against a ground truth consisting of set of user-generated reviews, manually annotated by 39 human subjects and available online, showing its human-like ability to capture feature-level opinions.

The d-index: Discovering dependences among scientific collaborators from their bibliographic data records

The evaluation of the work of a researcher and its impact on the research community has been deeply studied in literature through the definition of several measures, first among all the h-index and its variations. Although these measures represent valuable tools for analyzing researchers’ outputs, they usually assume the co-authorship to be a proportional collaboration between the parts, missing out their relationships and the relative scientific influences. In this work, we propose the d-index, a novel measure that estimates the dependence degree between authors on their research environment along their entire scientific publication history. We also present a web application that implements these ideas and provides a number of visualization tools for analyzing and comparing scientific dependences among all the scientists in the DBLP bibliographic database. Finally, relying on this web environment, we present case and user studies that highlight both the validity and the reliability of the proposed evaluation measure.

Field selection for job categorization and recommendation to social network users

Nowadays, in the Web 2.0 reality, one of the most challenging task for companies that aim to manage and recommend job offers is to convey this enormous amount of information in a succinct and intelligent manner such to increase the performances of matching operations against users profiles/curricula and optimize the time/space complexity of these processes. With this goal, this paper presents a novel method to formalize the textual content of job offers that aims at identifying the most relevant information and fields expressed by them and leverage this compact formalization for job recommendation and profile matching in social network environments. This method has been then developed and tested in the industrial environment represented by Multiposting and Work4, world leaders in digital solutions of e-recruitment problems. In this study three classes of documents are considered: job offers, job categories and social network user profiles (as potential job candidates); each class contains several fields with textual information. The proposed representation method permits to dynamically identify those text fields, for each class, that could help a cross-matching strategy in order to preserve, from one hand, the matching/recommendation performances and, on the other hand, reduce the cost of these operations (due to a straightforward dimensionality reduction mechanism). We then evaluated and compared the presented approach showing significant improvements on both categorization and recommendation tasks by also drastically reducing their computational costs.

Lexical Mediation for Ontology-Based Annotation of Multimedia

In the last decade, the annotation of multimedia has evolved toward the use of ontologies, as a way to bridge the semantic gap between low level features of media objects and high level concepts. In many cases, the annotation terms refer to structured ontologies. Such ontologies, however, are often light scale domain oriented knowledge bases, whereas the employment of wide, commonsense ontologies would improve interoperability and knowledge sharing, with beneficial effects on search and navigation. In this chapter, we present an approach to the semantic annotation of media objects through a meaning negotiation approach that requires natural language lexical terms as interface and employs large scale commonsense ontologies. As a test case, we apply the annotation to narrative media objects, using a meta–ontology, called Drammar, to describe their structure. We present the annotation schema, the software architecture for integrating several large scale ontologies, and the lexical interface for negotiating the ontological term. We also describe an evaluation of the proposed approach, conducted through experiments with annotators.

The 10 million follower fallacy: audience size does not prove domain-influence on Twitter

With the advent of social networks and micro-blogging systems, the way of communicating with other people and spreading information has changed substantially. Persons with different backgrounds, age and education exchange information and opinions, spanning various domains and topics, and have now the possibility to directly interact with popular users and authoritative information sources usually unreachable before the advent of these environments. As a result, the mechanism of information propagation changed deeply, the study of which is indispensable for the sake of understanding the evolution of information networks. To cope up with this intention, in this paper, we propose a novel model which enables to delve into the spread of information over a social network along with the change in the user relationships with respect to the domain of discussion. For this, considering Twitter as a case study, we aim at analyzing the multiple paths the information follows over the network with the goal of understanding the dynamics of the information contagion with respect to the change of the topic of discussion. We then provide a method for estimating the influence among users by evaluating the nature of the relationship among them with respect to the topic of discussion they share. Using a vast sample of the Twitter network, we then present various experiments that illustrate our proposal and show the efficacy of the proposed approach in modeling this information spread.

Estimating domain-based user influence in social networks

Social networks and microblogging systems play a fundamental role in the diffusion of information. The information, from different sources, reaches each user through multiple connections, the study of which is indispensable for the sake of understanding the dynamics of its evolution and expansion. In this paper, we propose a system which enables to delve in the spread of information over a network along with the changes in the user relationships with respect to the domain of discussion. To cope up with the goal, considering Twitter as a case study, we analyse the tweets as the starting point or as the generators of the information which later flows through subsequent retweets. Furthermore, we integrate a N-Gram model classification approach for categorizing, under various domains, the information shared within the social network under consideration. We finally leverage this formalization to propose a domain-based model which aims to estimate the influence of a user, on a community, in the domain under consideration. In conclusion, using a sample of the Twitter network we then present a set of case studies and real case scenarios that show the validity of the proposed approach.

Topic Development Based Refinement of Audio-Segmented Television News

With the advent of the cable based television model, there is an emerging requirement for random access capabilities, from a variety of media channels, such as smart terminals and Internet. Random access to the information within a newscast program requires appropriate segmentation of the news. We present text analysis based techniques on the transcript of the news, to refine the automatic audio-visual segmentation. We present the effectiveness of applying the text segmentation algorithm CUTS to the news segmentation domain. We propose two extensions to the algorithm, and show their impacts through an initial evaluation.