No abstract

Most Companies are more customer centric than they were before. By adopting this strategy, it made the electronic commerce growing and enhance buyers experience. from the other side, Companies started to explore this customer experience-data generated-to extract knowledge about their customer to be well managed-eCRM-instead of classic customer relationship Management-CRM. Large quantity of Data motivated the companies to look for changes, and ask for more functionality, and this are what influenced software editors to adapt their solutions and implement the power of data. Nowadays, data available-Big Dataput the existing systems and architectures under question and push us to rethink the logical layer to explore this data. Following the data vague, puts a need to reconsider and study the strength of eCRM/CRM existing solutions and architectures. The main contribution of this paper is to propose architecture built on Trust-Based recommendation able to provide to companies better accuracy, coverage, novelty and diversity during the sales process.

Customer reviews of products and services play a key role in the customers' decision to buy a product or use a service. Customers' preferences and choices are influenced by the opinions of others online; on blogs or social networks. New customers are faced with many views on the web, but they can't make the right decision. Hence, the need for sentiment analysis is to clarify whether opinions are positive, negative or neutral. This paper suggests using the Aspect-Based Sentiment Analysis approach on reviews extracted from tourism websites such as TripAdvisor and Booking. This approach is based on two main steps namely aspect extraction and sentiment classification related to each aspect. For aspect extraction, an approach based on topic modeling is proposed using the semi-supervised CorEx (Correlation Explanation) method for labeling word sequences into entities. As for sentiment classification, various supervised machine learning techniques are used to associate a sentiment (positive, negative or neutral) to a given aspect expression. Experiments on opinion corpora have shown very encouraging performances.

No abstract

A blockchain is a technology that allows the storage and transmission of information without a control body. Technically, it is a distributed database in which the information sent by users is verified and grouped into blocks, thus forming a chain. Thanks to the secure encryption of the data and the fact that new transactions are linked to the previous ones, it is almost impossible to modify the old records without modifying the following ones. On the other hand, the control of the blockchain by more than half of the nodes in the network (by consensus) makes it impossible to falsify the data in the blockchain. However, this public/private, anonymous, and unforgeable ledger that is the blockchain contains a set of information (metrics, logs, etc.) that can provide clues for an efficient monitoring and allow the reinforcement of the security of the blockchain that could be discussed in the future with the advent of quantum machines.