Movienet: a movie multilayer network model using visual and textual semantic cues

Abstract

Discovering content and stories in movies is one of the most important concepts in multimedia content research studies. Network models have proven to be an efficient choice for this purpose. When an audience watches a movie, they usually compare the characters and the relationships between them. For this reason, most of the models developed so far are based on social networks analysis. They focus essentially on the characters at play. By analyzing characters interactions, we can obtain a broad picture of the narration's content. Other works have proposed to exploit semantic elements such as scenes, dialogues, etc.. However, they are always captured from a single facet. Motivated by these limitations, we introduce in this work a multilayer network model to capture the narration of a movie based on its script, its subtitles, and the movie content. After introducing the model and the extraction process from the raw data, we perform a comparative analysis of the whole 6-movie cycle of the Star Wars saga. Results demonstrate the effectiveness of the proposed framework for video content representation and analysis.

Movie Script Similarity Using Multilayer Network Portrait Divergence

“…To even go further, we can enrich the model. Previous works have also formulated multimedia cues for movie analysis [16] including its visual features often important for recommendation systems [27]. We may further enrich the model, using for example sentiment analysis.…”

“…Network models have been increasingly used to support the analysis of stories [11], such as novels and plays [23] [26] [14], famous TV series [24], news [20] [21], and movies [18][15] [16]. However, most of these models only focus on one facet of the movie story.…”

“…Nevertheless, they have always captured the information in a single layer network or a bipartite graph. To enrich the representation, we have previously proposed a multilayer network model that captures key elements of the movie story [15] [16]. It encompasses the single network analysis based either on characters or scenes and proposes new topological analysis tools.…”

“…It combines semantically extracted characters, keywords, and locations, allowing to compute similarity between the corresponding networks. Starting from the previously proposed model [15] [16] the multilayer network is automatically extracted from the scripts of the movies to compare. At this point the similarity issue reduces to compute the similarities between the layers.…”

Quantifying the global film festival circuit: Networks, diversity, and public value creation

“…The international film festival circuit has been long discussed as a network in theory [6,11,41,59], and described by borrowing terminology from network science, such as "a network with nodes" [41] or "dense network" [59]. Network methods have been applied to the broader domain of film, such as to study collaboration between creatives in film production [45,[60][61][62] or facilitate the discovery of films via multi-layer network models [63]. So far, however, the applications of network analysis to film festival data have been limited in either regional scope [64] or the time period of a single year [20].…”