Case 39 Full Movie Dailymotion

[Alarico Boyett]

Theycontinued their work in Optimising the recommender engine with vector databases and opinion mining later with emphasis on ranking videos based on features like freshness, real views ratio, watch ratio, and aspect ratio to enhance user engagement and optimise watch time per user on the home feed. Furthermore, the team continued to focus on diversifying user interests by grouping videos based on interest and using stratified sampling to ensure a balanced experience for users.

By now it was clear to the Dailymotion team that the future initiatives will involve overcoming obstacles related to data processing, sentiment analysis, and user experience to provide meaningful and diverse recommendations. The main challenge stayed at the candidate generation process, textual embeddings, opinion mining, along with optimising the efficiency and accuracy of these processes and tackling the complexities of large-scale content curation.

Video content is inherently high-dimensional, comprising various features such as audio, visual, textual, and contextual elements.

Qdrant excels in efficiently handling high-dimensional data and out-of-the-box support for all the models with up to 65536 dimensions, making it well-suited for representing and processing complex video features with choice of any embedding model.

As the volume of video content and user interactions grows, scalability becomes paramount. Qdrant is meticulously designed to scale vertically as well as horizontally, allowing for seamless expansion to accommodate large volumes of data and user interactions without compromising performance.

Efficient video recommendation systems rely on identifying similarities between videos to make relevant recommendations. Qdrant leverages advanced HNSW indexing and similarity search algorithms to support fast and accurate retrieval of similar videos based on their feature representations nearly instantly (20ms for this use case)

Vector representations in Qdrant offer various Quantization and memory mapping techniques to efficiently store and retrieve vectors, leading to reduced storage requirements and computational overhead compared to alternative methods such as content-based filtering or collaborative filtering.

Qdrant allows to retrieve vector point / videos by ids while the Vertex Matching Engine requires a vector input to be able to search for other vectors which was another really important feature for Dailymotion

Figure shows the streaming architecture of the data processing pipeline that processes everytime a new video is uploaded or updated (Title, Description, Tags, Transcript), an updated embedding is computed and fed directly into Qdrant.

To find out if the library has a particular video, go to the Online Catalog and do a title search for the video. In the results list, videos are identified with this icon: . You can limit the results to videos only by adding the DVD/VHS filter. A search will find both DVDs and streaming video (see below). DVDs will have call numbers and streaming video will be identified as "electronic resource" in the results list and a link to the video in the catalog record.

At any time during the process of finding a particular video/s, feel free to ask for help directly from your Subject Librarian. If during your search, you find a video not held by the University Libraries that you want to use in a class, contact your Subject Librarian about your request or use the library's Purchase Request form.

Check Online Catalog for Documentaries and Feature Films

Begin by checking the Online Catalog to see if the University Libraries already has a streaming video and/or a DVD of the video. (See instructions in the box on the left.) The Online Catalog provides records on all videos currently accessible through the University Libraries, including both stream video and video on physical media, such as DVDs. All videos available through streaming databases such as Academic Video Online and Films on Demand or purchased individually and available on the Kanopy and Swank platforms can be found using the Online Catalog.

Check Academic Video Online and Films on Demand for Documentaries

Although all licensed videos should be in the Online Catalog, it is recommended that you check the two largest video databases, Academic Video Online and Films on Demand, for documentaries in case there is a glitch in online catalog access. If a video is found in one of these databases, you can easily embed or link to the video in Blackboard

Check Kanopy for Documentaries

If you do not find a streaming version of a documentary in the Online Catalog or in the larger video databases, you can check to see if the video is available for lease on the Kanopy:

Check Swank Digital Campus for Feature Films

If you do not find a streaming version of a feature film in the Online Catalog, you can check to see if the video is available via Swank Digital Campus:

Check YouTube for Documentaries

Before giving up on finding a video, check to see if the video is freely available on YouTube. Several faculty video requests have been fulfilled in the past by YouTube. Captions in YouTube are of relatively poor quality so issues of accessibility should be considered when using YouTube videos. Make sure the video is available through a legitimate source. Other potential free sources are Tubi and Dailymotion. Tubi includes free, premium, on demand video streaming with ads, and Dailymotion is a French video-sharing platform much like YouTube.

Check Pay-Per-View Services for Feature Films

In cases where access to streaming video is desired and no copy is available though other sources, it may be possible to access a video through a commercial streaming video Pay-Per-View services such as Amazon Prime, Hulu, iTunes, etc. Faculty and students can often rent or purchase films through one of these services for a low cost. Faculty sometimes require that students pay for a film in this manner as they would a textbook. There are a number of websites that identify streaming options for videos and provide price comparisons, including: Justwatch and Reelgood.

Check Google

Another option for finding a video not held by the University Libraries is to simply search Google and check the results list for references to a website or vendor that might provide access to the video. Again, contact your Subject Librarian if you are unable to find a freely available copy of the video.

At any time during the process of finding videos on your topic, feel free to ask for help directly from your Subject Librarian.

Check Online Catalog for Documentaries and Feature Films

Begin by doing a keyword search in the Online Catalog to see if the University Libraries has videos on your topic. The Online Catalog provides information about all licensed videos currently accessible through the University Libraries, including both streaming video and video on physical media, such as DVDs. When doing the keyword search, limit your results using the DVD/Videos filter on the right. This filter will limit your results to both streaming video and DVDs.

Check SmartSearch for Documentaries and Feature Films

You can also do a keyword search in SmartSearch. You can limit your results under Content Type on the left. Some possible options are "Streaming Video", "Streaming Audio", and "Video Recording" (for DVDs).

Check Academic Video Online, Films on Demand, and Subject-Specific Video Collections

Do a keyword search in the two multidisciplinary video databases, Academic Video Online and Films on Demand, as well as any of the other video collections found under the Video/Audio by Subject tab on the Streaming Video/Audio page of this guide that are appropriate for your topic. For example, if you are looking for dance videos, search Dance Online: Dance Videos. If you are looking for a nursing training video, search Nursing Education in Video.

3a8082e126