María del Carmen Rodríguez-Hernández

Pull-based recommendations in mobile environments

In the field of Context-Aware Recommendation Systems (CARS), only static contextual information is usually considered. However, the dynamic contextual information would very helpful in mobile computing scenarios. Despite this interest, the design and implementation of flexible and generic frameworks to support an easy development of context-aware mobile recommendation systems have been relatively unexplored. In this paper, we describe a framework that facilitates the development of CARS for mobile environments. We mainly focus on the development of the elements needed to support pull-based recommendations and the experimental evaluation of the proposed system. We have designed a framework to support mobile context-aware recommendations.We have described in detail the pull-based recommendation module.We have performed an extensive study of the state of the art.We have performed an experimental evaluation that compares different paradigms.The proposed architecture is generic, extensible, and adaptable to the requirements.

Towards a Context-Aware Mobile Recommendation Architecture

Nowadays, the huge amount of information available may easily overwhelm users when they need to take a decision that involves choosing among several options. On the one hand, it is necessary to identify which items are relevant for the user at a particular moment and place. On the other hand, some mechanism would be needed to rank the different alternatives. Recommendation systems, that offer relevant items to the users, have been proposed as a solution to these problems. However, they usually target very specific use cases (e.g., books, movies, music, etc.) and are not designed with mobile users in mind, where the context and the movements of the users may be important factors to consider when deciding which items should be recommended.

DataGenCARS: A generator of synthetic data for the evaluation of context-aware recommendation systems

Abstract Context-Aware Recommender Systems (CARS) have started to attract significant research attention in the last years, due to the interest of considering the context of the user in order to offer him/her more appropriate recommendations. However, the evaluation of CARS is a challenge, due to the scarce availability of appropriate datasets that incorporate context information related to the ratings provided by the users. In this paper, we present DataGenCARS, a complete Java-based synthetic dataset generator that can be used to obtain the required datasets for any type of scenario desired, allowing a high flexibility in the obtention of appropriate data that can be used to evaluate CARS. The generator presents features such as: a flexible definition of user schemas, user profiles, types of items, and types of contexts; a realistic generation of ratings and attributes of items; the possibility to mix real and synthetic datasets; functionalities to analyze existing datasets as a basis for synthetic data generation; and support for the automatic mapping between item schemas and Java classes. Moreover, an experimental evaluation illustrates the interest and the benefits provided by DataGenCARS.

Exploiting Linguistic Analysis on URLs for Recommending Web Pages: A Comparative Study

Nowadays, citizens require high level quality information from public institutions in order to guarantee their transparency. Institutional websites of governmental and public bodies must publish and keep updated a large amount of information stored in thousands of web pages in order to satisfy the demands of their users. Due to the amount of information, the “search form”, which is typically available in most such websites, is proven limited to support the users, since it requires them to explicitly express their information needs through keywords. The sites are also affected by the so-called “long tail” phenomenon, a phenomenon that is typically observed in e-commerce portals. The phenomenon is the one in which not all the pages are considered highly important and as a consequence, users searching for information located in pages that are not condiered important are having a hard time locating these pages.

A review of the role of sensors in mobile context-aware recommendation systems

Recommendation systems are specialized in offering suggestions about specific items of different types (e.g., books, movies, restaurants, and hotels) that could be interesting for the user. They have attracted considerable research attention due to their benefits and also their commercial interest. Particularly, in recent years, the concept of context-aware recommendation system has appeared to emphasize the importance of considering the context of the situations in which the user is involved in order to provide more accurate recommendations. The detection of the context requires the use of sensors of different types, which measure different context variables. Despite the relevant role played by sensors in the development of context-aware recommendation systems, sensors and recommendation approaches are two fields usually studied independently. In this paper, we provide a survey on the use of sensors for recommendation systems. Our contribution can be seen from a double perspective. On the one hand, we overview existing techniques used to detect context factors that could be relevant for recommendation. On the other hand, we illustrate the interest of sensors by considering different recommendation use cases and scenarios.

Towards Trajectory-Based Recommendations in Museums: Evaluation of Strategies Using Mixed Synthetic and Real Data

Abstract Recommendation systems, which suggest items that are of potential interest to the user (e.g., regarding which books to read, which movies to watch, etc.) have grown in popularity due to the ever-increasing amount of data available, that can lead to significant user’s overload. In particular, in recent years, extensive research has focused on the so-called Context-Aware Recommender Systems (CARS), which exploit context data to offer more relevant recommendations. In this paper, we study this problem with a use case scenario: recommending items to observe in a museum. We propose a trajectory-based and user-based collaborative filtering approach, that considers context data such as the location of the user and his/her trajectory to offer personalized recommendations. Besides, we exploit DataGenCARS, a dataset synthetic generator designed to construct datasets for the evaluation of context-aware recommendation systems, to build a mixed scenario based on both real and synthetic data. The experimental results show the advantages of the proposed approach and the usefulness of DataGenCARS for practical evaluation with a real use-case scenario.

Intelligent Transportation Systems – Maybe, But Where Are My Agents?

Significant advances in wireless communication technologies and mobile devices have led to their widespread use. For example, the so-called Intelligent Transportation Systems (ITS), which encompass a wide range of advanced applications for transportation, have attracted a lot of attention. In this context, one could think that software agents, which can have properties such as intelligence and autonomy, are expected to play a key role. But is this the case? Are they being used in work related to ITS and/or vehicular networks? Could they really provide benefits? In this paper, we analyze the state of the art and draw some conclusions about the potential interest of mixing these two fields.

Towards Keyword-based Pull Recommendation Systems

Due to the high availability of data, users are frequently overloaded with a huge amount of alternatives when they need to choose a particular item. This has motivated an increased interest in research on recommendation systems, which filter the options and provide users with suggestions about specific elements (e.g., movies, restaurants, hotels, books, etc.) that are estimated to be potentially relevant for the user. In this paper, we describe and evaluate two possible solutions to the problem of identification of the type of item (e.g., music, movie, book, etc.) that the user specifies in a pull-based recommendation (i.e., recommendation about certain types of items that are explicitly requested by the user). We evaluate two alternative solutions: one based on the use of the Hidden Markov Model and another one exploiting Information Retrieval techniques. Comparing both proposals experimentally, we can observe that the Hidden Markov Model performs generally better than the Information Retrieval technique in our preliminary experimental setup.

Push-Based Recommendations in Mobile Computing Using a Multi-Layer Contextual Approach

Nowadays, due to the high availability of heterogeneous data sources that can provide interesting information, users usually suffer from information overload. Therefore, the development of adaptive information systems that can offer personalized information and filter out irrelevant data for a user is required. Significant work has been developed to solve this problem in the area of the so-called recommendation systems. However, context information has only started to be considered recently to build recommendation systems, despite being key to obtain more accurate recommendations. Moreover, even with some context information, there is still a significant gap between the fields of mobile computing and recommendation systems. In this paper, we focus on push-based recommendations (i.e., recommendations not explicitly requested) for mobile users, as it represents the most challenging and effective approach for recommending items in mobile environments. As opposed to existing work, a generic model that fits different domains is proposed. This model is based on the definition of the concept of environment and manages the impact of dynamic events and all the actors involved in the mobile recommendation process.

Towards a Structured Representation of Results in an Information Retrieval System for Public Examination Calls

Nowadays, the huge amount of information available may easily overwhelm users. Information Retrieval techniques can help the user to find what he/she needs, but there are still challenges to solve within this research area. An example is the problem of minimizing the users search time to find specific information in unstructured texts within the retrieved documents, in different application domains. The use of supervised learning-based information extraction techniques can be a solution to this problem. However, a supervised learning model requires as input a large labeled dataset, generated manually by experts. Moreover, there are currently very few information extraction frameworks that allow to reduce or avoid the human effort needed to label such training datasets. In this paper, we present our work in progress towards the development of an information retrieval system that will display structured, centralized and updated information extracted from documents corresponding to calls for public examinations. In this scenario, the search engine should be able not only to display the documents relevant to the users query, but also specific data contained in the documents. In addition, we present a study of frameworks that can be used in this context as well as our preliminary experience with the use of the Snorkel framework. In the future, we plan to complete our proposal and also extend it for other types of documents published in Spanish official bulletins.