Rating-based Systems Research Articles

Empowering Recommendations with NLP: Exploiting Textual Reviews for Enhanced Rating-Based Systems

This research paper proposes a rating-based recommender system that leverages Natural Language Processing (NLP) techniques to enhance the accuracy and effectiveness of recommendations. Traditional recommender systems primarily rely on numerical ratings provided by users to make predictions. However, these ratings often lack detailed information about user preferences and suffer from sparsity and inconsistency issues. By incorporating NLP, we aim to extract valuable insights from textual reviews and improve the recommendation process. Our system utilizes sentiment analysis, topic modelling, and text embeddings to capture the implicit information in reviews and generate more personalized and context-aware recommendations. The experimental results demonstrate the superior performance of the proposed rating-based recommender system compared to conventional approaches.

Teaching Notes The Forgotten Women of Pre-Code: An Annotated Filmography and Bibliography

In recent years, “pre-code” films have been re-discovered and applauded by film scholars and feminists. The term refers to the period between 1929 and 1934 when many Hollywood studios openly disregarded the censorship restrictions of the Hays Code. Named after censorship czar William H. Hays, the Code forbade nudity, cursing, sexual innuendo, miscegenation and a host of other behaviors deemed inappropriate or offensive by religious groups. However, the impact of the Great Depression and the lack of a unified standard among state censorship boards allowed Hollywood to take risks in what could be shown. The result was a series of uncompromising, insightful, and controversial films that explored mature relationships between the sexes, offered realistic depictions of society’s downtrodden, and challenged social and sexual mores. After Franklin Delano Roosevelt took office and the economy recovered slightly, religious groups finally took on what they saw as the excesses of Hollywood. In 1934, the Catholic Legion of Decency mobilized a successful boycott against the studios, ordering millions of Roman Catholics to abstain from movie attendance until their demands were met. Amid a climate of antiSemitism, the studio moguls were forced to capitulate. The result was the reinstitution of the Production Code, a censorship board that dominated all aspects of Hollywood filmmaking until the 1960s, when a rating-based system replaced it. This filmography is intended to broaden a contemporary audiences’ knowledge and appreciation of pre-code film and explore the ways in which these works can be used to support teaching and learning on campus. Contrary to the perception that the 1940s was the heyday of the “woman’s picture,” the pre-code era has been recognized by some film historians to have provided actually more opportunities to portray women as independent, strong-minded, and courageous in their explorations of sexuality. This annotated filmography and selective bibliography examines how and why these films might be considered appropriate for any college film collection and offers recommendations for collection development. Only silent films and films unavailable on DVD have been omitted, reflecting the

Reliability of nominations and two-dimensional rating scale methods for sociometric status determination

Sociometric status classifications of 205 youths (average age 11.3 years) were assessed on three occasions with 1-year intervals, using the current nomination methods and a two-dimensional rating-based procedure (SSrat). The goal of the study was to compare the stability of the methods involved. Preliminarily, it is shown that (a) the data allowed the stabilities of the different systems to be interpreted as test-retest reliabilities, and (b) usual practice of implementing a uniform cutoff limit for all sociometric categories without further justification was inadvisable. Consequently, the reliability of the various methods was established with varying cutoff limits. The rating-based systems were shown to yield classifications with test–retest reliability over 0.50, considerably higher than the nomination methods. We also demonstrated that the rating scale system provides the opportunity to monitor the choice of more appropriate cutoff criteria. As for theoretical implications, the present study provides support for Coie's theory [Coie, J.D., 1990. Toward a theory of peer rejection. In S.R. Asher, J.D. Coie. Peer rejection in childhood. New York: Cambridge University Press, pp. 365–401] of acquiring peer status as a stabilizing process. It also shows that the relatively low test–retest reliability of the nomination procedures employed may, in part, be responsible for the lack of empirical support for Coie's view in previous studies.