Poonam B. Thorat

Bio: Poonam B. Thorat is an academic researcher. The author has contributed to research in topics: Information filtering system & Recommender system. The author has an hindex of 1, co-authored 2 publications receiving 79 citations.

Survey on Collaborative Filtering, Content-based Filtering and Hybrid Recommendation System

Abstract: systems or recommendation systems are a subset of information filtering system that used to anticipate the 'evaluation' or 'preference' that user would feed to an item. In recent years E-commerce applications are widely using Recommender system. Generally the most popular E- commerce sites are probably music, news, books, research articles, and products. Recommender systems are also available for business experts, jokes, restaurants, financial services, life insurance and twitter followers. Recommender systems have formulated in parallel with the web. Initially Recommender systems were based on demographic, content-based filtering and collaborative filtering. Currently, these systems are incorporating social information for enhancing a quality of recommendation process. For betterment of recommendation process in the future, Recommender systems will use personal, implicit and local information from the Internet. This paper provides an overview of recommender systems that include collaborative filtering, content-based filtering and hybrid approach of recommender system.

Hybrid conversational recommender system using temporal difference

Abstract: Recommender systems or recommendation systems are a subset of information filtering system that used to anticipate the 'evaluation' or 'preference' that user would feed to an item. In recent years E-commerce applications are widely using Recommender system. Generally the most popular E-commerce sites are probably music, news, books, research articles, and products. A conversational recommender system uses critiquing as a feedback to efficiently retrieve a product. Critiquing is common and powerful form of feedback, where a user can express their feature preferences. The expectation is that in each cycle the system retrieves the product that best satisfy the user's preferences from a minimal information input. The Reinforcement Learning (RL) approaches will be used to improve retrieval quality based on combination of compatibility and similarity scores. Reinforcement Learning (RL) reward function approach will be used with existing critiquing strategies to provide compatibility that will take into account for both, the moment at which the user makes a review and the number of satisfied review. This work focuses on new and existing strategies for conversational recommender system that will have potential to improve retrieval quality.

A hybrid recommendation system with many-objective evolutionary algorithm

Abstract: Recommendation system (RS) is a technology that provides accurate recommendations to users However, it is not comprehensive to only consider the accuracy of the recommendation because users have different requirements To improve the comprehensive performance, this paper presents a hybrid recommendation model based on many-objective optimization, which can simultaneously optimize the accuracy, diversity, novelty and coverage of recommendation This model enhances the robustness of recommendations by mixing three different basic recommendation technologies Additionally, we solve it with many-objective evolutionary algorithm (MaOEA) and test it extensively Experimental results demonstrate the effectiveness of the presented model, which can provide the recommendations with more and novel items on the basis of accurate and diverse

A Social-Relationships-Based Service Recommendation System for SIoT Devices

Abstract: Social Internet of Things comes as a new paradigm of Internet of Things to solve the problems of network discovery, navigability, and service composition. It aims to socialize the IoT devices and shape the interconnection between them into social interaction just like human beings. In IoT scenarios, a device can offer multiple services and different devices can offer the same services with different parameters and interest factors. The proliferation of offered services led to difficulties during service filtering and customization, this problem is known as services explosion. The selection of a suitable service that fits the requirements of the applications and devices is a challenging task. Several works have addressed service discovery, composition, and selection in IoT. However, these works did not emphasize on the fact that incorporating the users’ social features can increase the efficiency of the recommended services and help us to offer context-aware services. In this article, we present a service recommendation system that takes advantage of the social relationships between devices’ owners, where the recommendation is based on the different relationships between the service requester and service provider. Experimental results show, in the context of IoT, that incorporating the users’ social relationships in service recommendation increases the accuracy and diversity of the offered services.