As the volume of video content for businesses grow, the problem of how to find appropriate content within an ever-growing library also arises. According to a report from Cisco, globally, IP video traffic is going to be 82 percent of all the IP traffic (both business and consumer) by 2022. This report also indicates that video consumption is essential for many organizations. Industries such as insurance, education, finance, healthcare, to name a few- are generating massive amounts of videos.
As per another report published by IDC, knowledge workers spend more than nine hours a week trying to find the information they need to do their job. An information worker earning $60,000 per year will spend about $14,251 per year searching for content. Also, not finding the related information costs the same organization an additional $5.3 million a year. A better video search system is worth millions of dollars in productivity and translates into a higher bottom-line.
VIDIZMO Search Solution
VIDIZMO, a Gartner recognized enterprise video content management system (EVCM), has made searching for video both within the platform and searching inside the video easy for any organization keeping in view the inherent challenges faced by enterprises in seeking the relevant content.
VIDIZMO’s video search within the platform is done by categories, tags, custom attributes (user-defined, and system-generated), and metadata generated from various sources such as Artificial Intelligence (A.I.). Whereas search inside the video is done through:
- Spoken Words
- Labels (picked up from objects, scenes, and visuals)
- Keywords Extraction
Search Within Platform
VIDIZMO’s search within the platform is done by searching relevant metadata, tags, video title, video description, author, categories, closed captions, and filters in the search bar. By entering any of these relevant details of the video, VIDIZMO’s powerful search tool will generate the appropriate result in the platform. A good example can be of the insurance agency, where a claim adjuster looking for a particular video uploaded by the claimant can search the video by entering the claim number entered by the claimant while uploading the video.
Content in VIDIZMO can also be searched by applying the appropriate categories and filters, thereby narrowing down the search result generated by the platform. Similarly, using multiple categories and filters simultaneously will enable VIDIZMO’s powerful faceted or multidimensional search, which can narrow down search results even further. In industries such as the automotive industry where videos are used for training automotive engineers in creating, modifying, and assembling whole parts of an automobile, viewers can reduce the time searching for video content by selecting the appropriate category for instance “engineering” and applying multiple filters such as “author”, “part type” etc, hence specifying the search results to enlist the most relevant content for viewers.
Another use for faceted search is in Law Enforcement agencies. One of the most challenging and time-consuming tasks for any Law Enforcement during an investigation is the search for incident data and metadata during the analysis of digital evidence. VIDIZMO’s Digital Evidence Management is positioned to help law enforcement agencies solve the most challenging task of searching digital evidence content. By selecting the search facets through choosing the right filters and categories, VIDIZMO automatically filters out irrelevant sub facets. Because the system generates only relevant evidence with every selection, it allows law enforcement agencies to drill down into searching results with precision and accuracy.
Search Inside Video
Using VIDIZMO’s search inside video tool, users can search inside the video for spoken words, images, faces, sentiments, labels, brands, and any keywords that appear on-screen during the entire video. All of these detected features by VIDIZMO are searchable in a search box that appears beside the video. This way, viewers can skim video content within a fraction of a second and find the most relevant parts that they are interested in. Also, these search in-video features can also be searched within the platform.
Spoken Words: Words spoken in the video are transcribed either automatically or manually (depends on the customer’s choice), which then appears as either closed caption or subtitles. These spoken words are then searchable in a search box that appears beside the video. In cases where videos are long, users can fast-forward to the exact moment in the video where their search was mentioned.
Faces: VIDIZMO Artificial Intelligence (AI) features detect facial features of the people present in the video. The detected faces are displayable at the side of the video, which, when clicked, will allow the viewers to watch the exact part where the exposed face appeared. This feature is useful in videos where there are multiple speakers and the viewer by clicking on the face of the speaker they want to listen to will reach the part in the video where the speaker spoke and thereby watch only the relevant section of the video instead of watching the entire video.
Search in video using VIDIZMO's facial recognition features
Labels: VIDIZMO, through its video insights, can identify visual objects and actions displayed, which are then marked as labels. These labels are then made clickable and searchable through the search box besides the video.
Topics: VIDIZMO’s A.I. feature can infer the main issues of a video through transcripts. This feature is useful for training, learning, and education where students or knowledge workers can identify the main topics of the video by looking at the tags generated by VIDIZMO and can view only the portion of the video where the relevant topic is discussed. Knowledge workers and students will, therefore, save time searching and viewing the appropriate content.
Keywords: VIDIZMO’s A.I. extracts keywords from speech and visual text, these insights can then be searched inside the video by either typing the keywords in a search box or clicking on the tags generated by the system.
Sentiments: In marketing, videos are used to engage audiences' attention because it creates an emotional connection via visuals and sounds. But to know the opinions and the process of emotional tone behind a series of words, marketers conduct sentiment analysis. VIDIZMO’s sentiment analysis identifies positive, negative, and neutral sentiments from speech and visual texts of a video.
So just by looking at a video, marketers can monitor content data. Where parts of the video when positive sentiments are expressed by the customer can be clipped and used as customer testimonials for use in retailing and promotions. VIDIZMO’s powerful search would allow marketers to see the exact parts of the video, where positive or negative sentiment is expressed. Marketers would need to click on the sentiments for instance ‘positive’ and click on the indicators that would display below the sentiments indicating at what timeline the sentiment was expressed. By clicking on the sentiment indicator, they would automatically reach the part where the relevant sentiment is expressed.
Traditionally, viewers would need to hunt for a topic within a video by scrolling bit-by-bit through a video timeline, hoping to come across the content they needed. Not only was this process inefficient, but time-consuming as well, resulting in considerable costs to organizations where knowledge workers would spend hours looking for the relevant content.
VIDIZMO’s powerful search is the most comprehensive way to search within the platform and search inside the video, thereby helping organizations find digital assets quickly, save time, reduce content search costs, and effectively manage a growing digital asset library.