The social media trends are moving at a very fast pace to keep itself updated with everything that is happening around. This industry got a major boost in 2008 when a large market size was grabbed by Facebook. This industry opened up opportunities not only for people to showcase their talents and tell stories in their version, but also for brands to advertise their products and reach targeted users easily.
Short Video applications are a major contributor behind this social media boom which is going on. With the increased penetration of mobile phones and cheap internet, more number of users are coming online and becoming regular consumers of these video contents. Content creation has become so easy with the bundle of filters and settings offered by these apps, that users are generating a huge amount of content every minute.
As more and more user content is getting generated, it is very critical to also maintain the quality of the same and keep the platform clean and safe from abuse. Quality content not only attracts genuine users but also meets every country’s digital content guidelines for business sustainability.
A Short Video app vendor (currently one of the top short video app vendors) was about to launch their app for the first time globally. They were in desperate need of a company, who understands the social media market well and have enormous experience in curing user generated content. They were looking for a partner, who can moderate their platform’s content in near real time and also
guide them to build an automated system which can offload a major chunk of moderation as the size of generated video content grows along with the number of users.
When they approached Foiwe, we understood their requirement better than any other content moderation vendors. With our experience of working with global market leaders and exposure to global user content, we provided them a three-step solution.
Firstly, we helped them to create clear guidelines for their digital content in line with digital content rules set by the government bodies. Also, with our deep understanding of user content, we helped them to classify them into multiple groups for further categorization. In addition to this, we also helped them design their moderation platform with various inputs which are key for moderation and shorter turnaround time.
Secondly, we immediately setup a scalable dynamic team with experienced content moderators, who will work aggressively on their platform to maintain the applications content quality and also give valuable feedback to finetune the moderation platform. With the advantage of multiple language coverage and our inhouse collaboration platform, we even helped them to moderate their live stream along with short videos.
Thirdly we deployed a team, which worked closely with their automation team to create a dataset for their AI models. Along with labeling and emotion recognition, this team also helped them in creating datasets for NLP, ASR and object detection etc. Utilizing this high-quality dataset, they were able to create a robust AI model with high accuracy.
Combining our human intelligence solution with a robust AI model, our short video partner is able to tackle their huge volume of dynamic video contents created by users and also triggered by various events. This hybrid solution gave our partner a full proof solution for continuous learning and quality check for the AI model and also to handle any unseen video category parallel to their automation in near real time.