Cristina Ioana Muntean

On Learning Prediction Models for Tourists Paths

In this article, we tackle the problem of predicting the “next” geographical position of a tourist, given her history (i.e., the prediction is done accordingly to the tourist’s current trail) by means of supervised learning techniques, namely Gradient Boosted Regression Trees and Ranking SVM. The learning is done on the basis of an object space represented by a 68-dimension feature vector specifically designed for tourism-related data. Furthermore, we propose a thorough comparison of several methods that are considered state-of-the-art in recommender and trail prediction systems for tourism, as well as a popularity baseline. Experiments show that the methods we propose consistently outperform the baselines and provide strong evidence of the performance and robustness of our solutions.

A trajectory-based recommender system for tourism

Recommendation systems provide focused information to users on a set of objects belonging to a specific domain. The proposed recommender system provides personalized suggestions about touristic points of interest. The system generates recommendations, consisting of touristic places, according to the current position of a tourist and previously collected data describing tourist movements in a touristic location/city. The touristic sites correspond to a set of points of interest identified a priori. We propose several metrics to evaluate both the spatial coverage of the dataset and the quality of recommendations produced. We assess our system on two datasets: a real and a synthetic one. Results show that our solution is a viable one.

Sentiment-enhanced multidimensional analysis of online social networks: perception of the mediterranean refugees crisis

We propose an analytical framework able to investigate discussions about polarized topics in online social networks from many different angles. The framework supports the analysis of social networks along several dimensions: time, space and sentiment. We show that the proposed analytical framework and the methodology can be used to mine knowledge about the perception of complex social phenomena. We selected the refugee crisis discussions over Twitter as a case study. This difficult and controversial topic is an increasingly important issue for the EU. The raw stream of tweets is enriched with space information (user and mentioned locations), and sentiment (positive vs. negative) w.r.t. refugees. Our study shows differences in positive and negative sentiment in EU countries, in particular in UK, and by matching events, locations and perception, it underlines opinion dynamics and common prejudices regarding the refugees.

Perception of social phenomena through the multidimensional analysis of online social networks

Abstract We propose an analytical framework aimed at investigating different views of the discussions regarding polarized topics which occur in Online Social Networks (OSNs). The framework supports the analysis along multiple dimensions, i.e., time, space and sentiment of the opposite views about a controversial topic emerging in an OSN. To assess its usefulness in mining insights about social phenomena, we apply it to two different Twitter case studies: the discussions about the refugee crisis and the United Kingdom European Union membership referendum. These complex and contended topics are very important issues for EU citizens and stimulated a multitude of Twitter users to take side and actively participate in the discussions. Our framework allows to monitor in a scalable way the raw stream of relevant tweets and to automatically enrich them with location information (user and mentioned locations), and sentiment polarity (positive vs. negative). The analyses we conducted show how the framework captures the differences in positive and negative user sentiment over time and space. The resulting knowledge can support the understanding of complex dynamics by identifying variations in the perception of specific events and locations.

Social Media Image Recognition for Food Trend Analysis

An increasing number of people share their thoughts and the images of their lives on social media platforms. People are exposed to food in their everyday lives and share on-line what they are eating by means of photos taken to their dishes. The hashtag #foodporn is constantly among the popular hashtags in Twitter and food photos are the second most popular subject in Instagram after selfies. The system that we propose, WorldFoodMap, captures the stream of food photos from social media and, thanks to a CNN food image classifier, identifies the categories of food that people are sharing. By collecting food images from the Twitter stream and associating food category and location to them, WorldFoodMap permits to investigate and interactively visualize the popularity and trends of the shared food all over the world.

RecTour: A Recommender System for Tourists

This paper presents a recommender system that provides personalized information about locations of potential interest to a tourist. The system generates suggestions, consisting of touristy places, according to the current position and history data describing the tourist movements. For the selection of tourist sites, the system uses a set of points of interest a priori identified. We evaluate our system on two datasets: a real and a synthetic one, both storing trajectories describing previous movements of tourists. The proposed solution has high applicability and the results show that the solution is both efficient and viable.

RankEval: An Evaluation and Analysis Framework for Learning-to-Rank Solutions

In this demo paper we propose RankEval, an open-source tool for the analysis and evaluation of Learning-to-Rank (LtR) models based on ensembles of regression trees. Gradient Boosted Regression Trees (GBRT) is a flexible statistical learning technique for classification and regression at the state of the art for training effective LtR solutions. Indeed, the success of GBRT fostered the development of several open-source LtR libraries targeting efficiency of the learning phase and effectiveness of the resulting models. However, these libraries offer only very limited help for the tuning and evaluation of the trained models. In addition, the implementations provided for even the most traditional IR evaluation metrics differ from library to library, thus making the objective evaluation and comparison between trained models a difficult task. RankEval addresses these issues by providing a common ground for LtR libraries that offers useful and interoperable tools for a comprehensive comparison and in-depth analysis of ranking models.

Sentiment Spreading: An Epidemic Model for Lexicon-Based Sentiment Analysis on Twitter

While sentiment analysis has received significant attention in the last years, problems still exist when tools need to be applied to microblogging content. This because, typically, the text to be analysed consists of very short messages lacking in structure and semantic context. At the same time, the amount of text produced by online platforms is enormous. So, one needs simple, fast and effective methods in order to be able to efficiently study sentiment in these data. Lexicon-based methods, which use a predefined dictionary of terms tagged with sentiment valences to evaluate sentiment in longer sentences, can be a valid approach. Here we present a method based on epidemic spreading to automatically extend the dictionary used in lexicon-based sentiment analysis, starting from a reduced dictionary and large amounts of Twitter data. The resulting dictionary is shown to contain valences that correlate well with human-annotated sentiment, and to produce tweet sentiment classifications comparable to the original dictionary, with the advantage of being able to tag more tweets than the original. The method is easily extensible to various languages and applicable to large amounts of data.

MUSETS: Diversity-Aware Web Query Suggestions for Shortening User Sessions

We propose MUSETS (multi-session total shortening) – a novel formulation of the query suggestion task, specified as an optimization problem. Given an ambiguous user query, the goal is to propose the user a set of query suggestions that optimizes a diversity-aware objective function. The function models the expected number of query reformulations that a user would save until reaching a satisfactory query formulation. The function is diversity-aware, as it naturally enforces high coverage of different alternative continuations of the user session. For modeling the topics covered by the queries, we also use an extended query representation based on entities extracted from Wikipedia. We apply a machine learning approach to learn the model on a set of user sessions to be subsequently used for queries that are under-represented in historical query logs and present an evaluation of the approach.