A collaborative filtering approach to mitigate the new user cold start problem.

The new user cold start issue represents a serious problem in recommender systems as it can lead to the loss of new users who decide to stop using the system due to the lack of accuracy in the recommenda- tions received in that first stage in which they have not yet cast a significant number of votes with which to feed the recommender system?s collaborative filtering core. For this reason it is particularly important to design new similarity metrics which provide greater precision in the results offered to users who have cast few votes. This paper presents a new similarity measure perfected using optimization based on neu- ral learning, which exceeds the best results obtained with current metrics. The metric has been tested on the Netflix and Movielens databases, obtaining important improvements in the measures of accuracy, precision and recall when applied to new user cold start situations. The paper includes the mathematical formalization describing how to obtain the main quality measures of a recommender system using leave- one-out cross validation.

Social-media monitoring for cold-start recommendations

Generating personalized movie recommendations to users is a problem that most commonly relies on user-movie ratings. These ratings are generally used either to understand the user preferences or to recommend movies that users with similar rating patterns have rated highly. However, movie recommenders are often subject to the Cold-Start problem: new movies have not been rated by anyone, so, they will not be recommended to anyone; likewise, the preferences of new users who have not rated any movie cannot be learned. In parallel, Social-Media platforms, such as Twitter, collect great amounts of user feedback on movies, as these are very popular nowadays. This thesis proposes to explore feedback shared on Twitter to predict the popularity of new movies and show how it can be used to tackle the Cold-Start problem. It also proposes, at a finer grain, to explore the reputation of directors and actors on IMDb to tackle the Cold-Start problem. To assess these aspects, a Reputation-enhanced Recommendation Algorithm is implemented and evaluated on a crawled IMDb dataset with previous user ratings of old movies,together with Twitter data crawled from January 2014 to March 2014, to recommend 60 movies affected by the Cold-Start problem. Twitter revealed to be a strong reputation predictor, and the Reputation-enhanced Recommendation Algorithm improved over several baseline methods. Additionally, the algorithm also proved to be useful when recommending movies in an extreme Cold-Start scenario, where both new movies and users are affected by the Cold-Start problem.

Connecting social media to e-commerce:cold-start product recommendation using microblogging information

In recent years, the boundaries between e-commerce and social networking have become increasingly blurred. Many e-commerce websites support the mechanism of social login where users can sign on the websites using their social network identities such as their Facebook or Twitter accounts. Users can also post their newly purchased products on microblogs with links to the e-commerce product web pages. In this paper, we propose a novel solution for cross-site cold-start product recommendation, which aims to recommend products from e-commerce websites to users at social networking sites in 'cold-start' situations, a problem which has rarely been explored before. A major challenge is how to leverage knowledge extracted from social networking sites for cross-site cold-start product recommendation. We propose to use the linked users across social networking sites and e-commerce websites (users who have social networking accounts and have made purchases on e-commerce websites) as a bridge to map users' social networking features to another feature representation for product recommendation. In specific, we propose learning both users' and products' feature representations (called user embeddings and product embeddings, respectively) from data collected from e-commerce websites using recurrent neural networks and then apply a modified gradient boosting trees method to transform users' social networking features into user embeddings. We then develop a feature-based matrix factorization approach which can leverage the learnt user embeddings for cold-start product recommendation. Experimental results on a large dataset constructed from the largest Chinese microblogging service Sina Weibo and the largest Chinese B2C e-commerce website JingDong have shown the effectiveness of our proposed framework.

Collaborative filtering and deep learning based recommendation system for cold start items

Recommender system is a specific type of intelligent systems, which exploits historical user ratings on items and/or auxiliary information to make recommendations on items to the users. It plays a critical role in a wide range of online shopping, e-commercial services and social networking applications. Collaborative filtering (CF) is the most popular approaches used for recommender systems, but it suffers from complete cold start (CCS) problem where no rating record are available and incomplete cold start (ICS) problem where only a small number of rating records are available for some new items or users in the system. In this paper, we propose two recommendation models to solve the CCS and ICS problems for new items, which are based on a framework of tightly coupled CF approach and deep learning neural network. A specific deep neural network SADE is used to extract the content features of the items. The state of the art CF model, timeSVD++, which models and utilizes temporal dynamics of user preferences and item features, is modified to take the content features into prediction of ratings for cold start items. Extensive experiments on a large Netflix rating dataset of movies are performed, which show that our proposed recommendation models largely outperform the baseline models for rating prediction of cold start items. The two proposed recommendation models are also evaluated and compared on ICS items, and a flexible scheme of model retraining and switching is proposed to deal with the transition of items from cold start to non-cold start status. The experiment results on Netflix movie recommendation show the tight coupling of CF approach and deep learning neural network is feasible and very effective for cold start item recommendation. The design is general and can be applied to many other recommender systems for online shopping and social networking applications. The solution of cold start item problem can largely improve user experience and trust of recommender systems, and effectively promote cold start items.

Cold-Start Emissions testing of Snowmobiles Using Ethanol and Gasoline

Increasing prices for fuel with depletion and instability in foreign oil imports has driven the importance for using alternative and renewable fuels. The alternative fuels such as ethanol, methanol, butyl alcohol, and natural gas are of interest to be used to relieve some of the dependence on oil for transportation. The renewable fuel, ethanol which is made from the sugars of corn, has been used widely in fuel for vehicles in the United States because of its unique qualities. As with any renewable fuel, ethanol has many advantages but also has disadvantages. Cold startability of engines is one area of concern when using ethanol blended fuel. This research was focused on the cold startability of snowmobiles at ambient temperatures of 20 °F, 0 °F, and -20 °F. The tests were performed in a modified 48 foot refrigerated trailer which was retrofitted for the purpose of cold-start tests. Pure gasoline (E0) was used as a baseline test. A splash blended ethanol and gasoline mixture (E15, 15% ethanol and 85% gasoline by volume) was then tested and compared to the E0 fuel. Four different types of snowmobiles were used for the testing including a Yamaha FX Nytro RTX four-stroke, Ski-doo MX Z TNT 600 E-TEC direct injected two stroke, Polaris 800 Rush semi-direct injected two-stroke, and an Arctic Cat F570 carbureted two-stroke. All of the snowmobiles operate on open loop systems which means there was no compensation for the change in fuel properties. Emissions were sampled using a Sensors Inc. Semtech DS five gas emissions analyzer and engine data was recoded using AIM Racing Data Power EVO3 Pro and EVO4 systems. The recorded raw exhaust emissions included carbon monoxide (CO), carbon dioxide (CO2), total hydrocarbons (THC), and oxygen (O2). To help explain the trends in the emissions data, engine parameters were also recorded. The EVO equipment was installed on each vehicle to record the following parameters: engine speed, exhaust gas temperature, head temperature, coolant temperature, and test cell air temperature. At least three consistent tests to ensure repeatability were taken at each fuel and temperature combination so a total of 18 valid tests were taken on each snowmobile. The snowmobiles were run at operating temperature to clear any excess fuel in the engine crankcase before each cold-start test. The trends from switching from E0 to E15 were different for each snowmobile as they all employ different engine technologies. The Yamaha snowmobile (four-stroke EFI) achieved higher levels of CO2 with lower CO and THC emissions on E15. Engine speeds were fairly consistent between fuels but the average engine speeds were increased as the temperatures decreased. The average exhaust gas temperature increased from 1.3-1.8% for the E15 compared to E0 due to enleanment. For the Ski-doo snowmobile (direct injected two-stroke) only slight differences were noted when switching from E0 to E15. This could possibly be due to the lean of stoichiometric operation of the engine at idle. The CO2 emissions decreased slightly at 20 °F and 0 °F for E15 fuel with a small difference at -20 °F. Almost no change in CO or THC emissions was noted for all temperatures. The only significant difference in the engine data observed was the exhaust gas temperature which decreased with E15. The Polaris snowmobile (semi-direct injected two-stroke) had similar raw exhaust emissions for each of the two fuels. This was probably due to changing a resistor when using E15 which changed the fuel map for an ethanol mixture (E10 vs. E0). This snowmobile operates at a rich condition which caused the engine to emit higher values of CO than CO2 along with exceeding the THC analyzer range at idle. The engine parameters and emissions did not increase or decrease significantly with decreasing temperature. The average idle engine speed did increase as the ambient temperature decreased. The Arctic Cat snowmobile (carbureted two-stroke) was equipped with a choke lever to assist cold-starts. The choke was operated in the same manor for both fuels. Lower levels of CO emissions with E15 fuel were observed yet the THC emissions exceeded the analyzer range. The engine had a slightly lower speed with E15.

Abatement of hydrocarbons by acid ZSM-5 and BETA zeolites under cold-start conditions

Simulated cold-start tests have been carried out to evaluate the performance of H-ZSM-5 and H-BETA zeolites as hydrocarbon traps under simulated gasoline car exhaust gases, paying special attention to the effect of water on their behaviour. It is concluded that the hydrothermal treatment of the zeolites in the acidic form contributes to the better performance of these materials as hydrocarbon traps since the stabilization of the zeolites takes place. Moreover, the decrease of the surface acidity of the zeolites results in an increase of the Si/Al ratio, which contributes to the decrease of the water affinity for adsorption sites. Thus, the competition with hydrocarbon molecules in the exhaust for the adsorption sites is reduced which increases their trap efficiency. The stabilized H-ZSM-5 is the zeolite that showed the best performance with a propene offset temperature of 240 °C, which should be high enough for the three-way catalyst to carry out its role as catalytic converter.

CuH-ZSM-5 as Hydrocarbon Trap under Cold Start Conditions

Cold start tests are carried out to evaluate the performance of copper-exchanged zeolites as hydrocarbon traps under simulated gasoline car exhaust gases, paying special attention to the role of copper in the performance of these zeolites. It is concluded that the partial substitution of the protons in the parent H-ZSM-5 zeolite is highly beneficial for hydrocarbon trapping due to the formation of selective adsorption sites with specific affinity for the different exhaust components. However, it is also observed that uncontrolled exchanging process conditions could lead to the presence of CuO nanoparticles in the zeolite surface, which seem to block the pore structure of the zeolite, decreasing the hydrocarbon trap efficiency. Among all the zeolites studied, the results point out that a CuH-ZSM-5 with a partial substitution of extra-framework protons by copper cations and without any detectable surface CuO nanoparticles is the zeolite that showed the best performance under simulated cold start conditions due to both the high stability and the hydrocarbon retaining capacity of this sample during the consecutive cycles.

BETA Zeolite Thin Films Supported on Honeycomb Monoliths with Tunable Properties as Hydrocarbon Traps under Cold-Start Conditions

A high percentage of hydrocarbon (HC) emissions from gasoline vehicles occur during the cold-start period. Among the alternatives proposed to reduce these HC emissions, the use of zeolites before the three-way catalyst (TWC) is thought to be very effective. Zeolites are the preferred adsorbents for this application; however, to avoid high pressure drops, supported zeolites are needed. In this work, two coating methods (dip-coating and in situ crystallization) are optimized to prepare BETA zeolite thin films supported on honeycomb monoliths with tunable properties. The important effect of the density of the thin film in the final performance as a HC trap is demonstrated. A highly effective HC trap is prepared showing 100 % toluene retention, accomplishing the desired performance as a HC trap, desorbing propene at temperatures close to 300 °C, and remaining stable after cycling. The use of this material before the TWC is very promising, and works towards achieving the sustainability and environmental protection goals.

Optimizing the performance of catalytic traps for hydrocarbon abatement during the cold-start of a gasoline engine

A key target to reduce current hydrocarbon emissions from vehicular exhaust is to improve their abatement under cold-start conditions. Herein, we demonstrate the potential of factorial analysis to design a highly efficient catalytic trap. The impact of the synthesis conditions on the preparation of copper-loaded ZSM-5 is clearly revealed by XRD, N2 sorption, FTIR, NH3-TPD, SEM and TEM. A high concentration of copper nitrate precursor in the synthesis improves the removal of hydrocarbons, providing both strong adsorption sites for hydrocarbon retention at low temperature and copper oxide nanoparticles for full hydrocarbon catalytic combustion at high temperature. The use of copper acetate precursor leads to a more homogeneous dispersion of copper oxide nanoparticles also providing enough catalytic sites for the total oxidation of hydrocarbons released from the adsorption sites, although lower copper loadings are achieved. Thus, synthesis conditions leading to high copper loadings jointly with highly dispersed copper oxide nanoparticles would result in an exceptional catalytic trap able to reach superior hydrocarbon abatement under highly demanding operational conditions.

Collaborative filtering and deep learning based hybrid recommendation for cold start problem

Recommender systems (RS) are used by many social networking applications and online e-commercial services. Collaborative filtering (CF) is one of the most popular approaches used for RS. However traditional CF approach suffers from sparsity and cold start problems. In this paper, we propose a hybrid recommendation model to address the cold start problem, which explores the item content features learned from a deep learning neural network and applies them to the timeSVD++ CF model. Extensive experiments are run on a large Netflix rating dataset for movies. Experiment results show that the proposed hybrid recommendation model provides a good prediction for cold start items, and performs better than four existing recommendation models for rating of non-cold start items.

Lightweight user modeling – a case study

In the context of previous publications, we propose a new lightweight UM process, intended to work as a tourism recommender system in a commercial environment. The new process tackles issues like cold start, gray sheep and over specialization through a rich user model and the application of a gradual forgetting function to the collected user action history. Also, significant performance improvements were achieved regarding the previously proposed UM process.

Algorithmes de recommandation musicale

Ce mémoire est composé de trois articles qui s’unissent sous le thème de la recommandation musicale à grande échelle. Nous présentons d’abord une méthode pour effectuer des recommandations musicales en récoltant des étiquettes (tags) décrivant les items et en utilisant cette aura textuelle pour déterminer leur similarité. En plus d’effectuer des recommandations qui sont transparentes et personnalisables, notre méthode, basée sur le contenu, n’est pas victime des problèmes dont souffrent les systèmes de filtrage collaboratif, comme le problème du démarrage à froid (cold start problem). Nous présentons ensuite un algorithme d’apprentissage automatique qui applique des étiquettes à des chansons à partir d’attributs extraits de leur fichier audio. L’ensemble de données que nous utilisons est construit à partir d’une très grande quantité de données sociales provenant du site Last.fm. Nous présentons finalement un algorithme de génération automatique de liste d’écoute personnalisable qui apprend un espace de similarité musical à partir d’attributs audio extraits de chansons jouées dans des listes d’écoute de stations de radio commerciale. En plus d’utiliser cet espace de similarité, notre système prend aussi en compte un nuage d’étiquettes que l’utilisateur est en mesure de manipuler, ce qui lui permet de décrire de manière abstraite la sorte de musique qu’il désire écouter.

Social and content hybrid image recommender system for mobile social networks

One of the advantages of social networks is the possibility to socialize and personalize the content created or shared by the users. In mobile social networks, where the devices have limited capabilities in terms of screen size and computing power, Multimedia Recommender Systems help to present the most relevant content to the users, depending on their tastes, relationships and profile. Previous recommender systems are not able to cope with the uncertainty of automated tagging and are knowledge domain dependant. In addition, the instantiation of a recommender in this domain should cope with problems arising from the collaborative filtering inherent nature (cold start, banana problem, large number of users to run, etc.). The solution presented in this paper addresses the abovementioned problems by proposing a hybrid image recommender system, which combines collaborative filtering (social techniques) with content-based techniques, leaving the user the liberty to give these processes a personal weight. It takes into account aesthetics and the formal characteristics of the images to overcome the problems of current techniques, improving the performance of existing systems to create a mobile social networks recommender with a high degree of adaptation to any kind of user.

Uma abordagem baseada em filtragem colaborativa integrada a mapas de saliência para a recomendação de imagens

Nowadays, the amount of customers using sites for shopping is greatly increasing, mainly due to the easiness and rapidity of this way of consumption. The sites, differently from physical stores, can make anything available to customers. In this context, Recommender Systems (RS) have become indispensable to help consumers to find products that may possibly pleasant or be useful to them. These systems often use techniques of Collaborating Filtering (CF), whose main underlying idea is that products are recommended to a given user based on purchase information and evaluations of past, by a group of users similar to the user who is requesting recommendation. One of the main challenges faced by such a technique is the need of the user to provide some information about her preferences on products in order to get further recommendations from the system. When there are items that do not have ratings or that possess quite few ratings available, the recommender system performs poorly. This problem is known as new item cold-start. In this paper, we propose to investigate in what extent information on visual attention can help to produce more accurate recommendation models. We present a new CF strategy, called IKB-MS, that uses visual attention to characterize images and alleviate the new item cold-start problem. In order to validate this strategy, we created a clothing image database and we use three algorithms well known for the extraction of visual attention these images. An extensive set of experiments shows that our approach is efficient and outperforms state-of-the-art CF RS.

Collaborative Filtering with Data Classification : A Combined Approach to Hotel Recommendation Systems

Recommendation systems aim to help users make decisions more efficiently. The most widely used method in recommendation systems is collaborative filtering, of which, a critical step is to analyze a user's preferences and make recommendations of products or services based on similarity analysis with other users' ratings. However, collaborative filtering is less usable for recommendation facing the "cold start" problem, i.e. few comments being given to products or services. To tackle this problem, we propose an improved method that combines collaborative filtering and data classification. We use hotel recommendation data to test the proposed method. The accuracy of the recommendation is determined by the rankings. Evaluations regarding the accuracies of Top-3 and Top-10 recommendation lists using the 10-fold cross-validation method and ROC curves are conducted. The results show that the Top-3 hotel recommendation list proposed by the combined method has the superiority of the recommendation performance than the Top-10 list under the cold start condition in most of the times.