Search Video Content Using Watson AI

Share this post:

With most folks working remotely, most organizations are using video conferencing technologies for meetings and trainings. These videos are a goldmine of information that will need to be made used for various purposes. 

While videos are easy to consume, it is also difficult to find specific piece of information that is hidden in some part of a video.  Thousands of hours of videos are being generated in various institutions/businesses that have important information that will be useful for them down the road.  Think of a student/employee trying to remember a specific input by a teacher/manager mentioned in an online class/meeting at some point. 

  1. First find the specific video. Currently searching video content depends on tags, descriptions and/or titles; which are prone to human error.
  2. Even if they do find the video, they would have to watch throughthe entire video (or at least a part of it) to get to the specific information.  That is time consuming and thus ineffective
  3. Especially in longer videos, search results more often than not are contextually incorrect

What if there was a way to find specific clips within a longer video as an answer to a question?   So

  • one can have the full video available as is (maybe 10-60 minutes) and
  • consumers are able to find a specific clip or part of the video that specifically caters to whatever they are looking for

IBM Watson is an AI platform that allows one to tap into unstructured content (including videos) and convert them into usable insights.  There are a number of offerings within the platform providing various capabilities like Natural Language Processing (NLP); including Speech and Natural Language Understanding (NLU).  What’s even better is the ability to use these features without no-code, easy to use GUI based tooling.  

A couple years back, IBM Research made TED videos available for search as a demo with the power of Watson – using cutting-edge technologies like video processing, speech to text, and semantic understanding of language (like concepts, keywords, etc. extraction) to make video content more searchable. Below you see a snapshot of the solution showing the result of a search query: “What is the relationship between money and happiness?”. 

Watson was able to search for all TED videos that matched those concepts and also find the segment within each video where this was being discussed by the TED speaker.  By analyzing concepts within each video, Watson can curate a playlist of short video clips that offer many perspectives on the user’s question. When a user enters a question, it searches across the videos to build a nuanced answer to that question from the speakers who tackled it.

Below each clip is a timeline that shows more concepts that Watson found within the talk, so that users can “tunnel sideways” to follow another interest that’s contextually related, allowing for more, serendipitous, explorations. Users can keep leaping from one video to another, using concepts as a bridge to other topics that might interest them. For a video of the Watson TED video search demo, see below

Watch the video

IBM partner Veracitiz has recently used the same principles to build a video search demo for an online educational enterprise that wanted to enable a search engine for their subscribers; in Hindi and “Hinglish” (Hindi written in English/latin script) languages.

Watson was used as a context-based Search Engine for Videos.  You can go through the details in this video – Watch Video

To summarize, using Watson AI you can quickly build a solution that can help

Some key beneficiaries of such a solution are:

  1. Educational Institutions that provide online live/pre-recorded lecture, that makes it impossible for students to look for specific topics/chapters/lectures without wasting time & energy manually searching.
  2. News and Media companies that want to engage with their subscribers and help them find information hidden within their video libraries
  3. Any business with online meeting recordings, where their employees want to search for specific information
  4. And many more…

Please feel free to reach out to me for a free consultation.  LinkedIn

More stories

Insurance Company Brings Predictability into Sales Processes with AI

Generally speaking, sales drives everything else in the business – so, it's a no-brainer that the ability to accurately predict sales is very important for any business. It helps companies better predict and plan for demand throughout the year and enables executives to make wiser business decisions.

Continue reading

Never miss an incident with an application-centric AIOps platform

Applications are bound to face occasional outages and performance issues, making the job of IT Ops all the more critical. Here is where AIOps simplifies the resolution of issues, even proactively, before it leads to a loss in revenue or customers.

Continue reading

How ICICI Prudential Life Insurance is Scaling Customer Care and Leveraging AI to Personalize Experiences

Organisations are constantly challenged to meet dynamic customer requirements and rethink ways to engage with them on their terms and as per their convenience. With customers at the core of decision making and business success, organisations are tuning to digital capabilities that can support new-age services. When done well, after sales service boosts the overall customer experience by providing […]

Continue reading