A leading television broadcaster wanted to streamline the way they managed their large volume of non-textual media data. Improving ease of use and access was an important consideration, as was cost savings. All this data (video and audio) was stored and managed between different production companies, podcast production studios, major communication agencies, multimedia libraries, etc.
Between the information provided by the producers, the data from the broadcasting center and the content itself, we have a wealth of information about the programs at our disposal. It’s just a matter of being able to make use of it! – the Company’s Chief Data and AI Officer
The main challenge for the television broadcaster was the high level of complexity involved in locating information about content. Each TV program was associated with dozens of data items (such as program title, description, broadcast date, age limit, cast, etc.), which were split across the group’s various departments. These Items were essential for content management but were often difficult to access.
Every time I need information about a particular piece of content, I have to find the right person and wait for their response. I don’t want to have to go and ask different departments for information – the Company’s Chief Data and AI Officer
Another challenge was that the existing knowledge discovery solution was based on keywords. This presented typical search challenges such as returning irrelevant results, missing relevant documents that didn’t contain the exact terms used in the search query, matching keywords literally without understanding the context or meaning behind them, difficulty with complex queries, etc. This led to repetitive, error-prone work by the user, and meant that data wasn’t being managed effectively.
To address these needs, a custom data programs platform based on a knowledge graph solution was developed by Perfect Memory and Ontotext. Powered by Ontotext’s Resource Description Framework (RDF) knowledge graph database (GraphDB) and Perfect Memory’s Cognitive Asset Management solution, this automates the indexing of content and documents and map information to business ontologies. The platform enables different types of users (editors, creative staff, program directors, marketers, lawyers, etc.) to easily find, share, and utilize the resources they need and extract information from them.
In conjunction with the content management departments, the broadcaster leveraged this knowledge graph powered solution to interlink the various data sources, standardize formats, and compile this information to make it available to the teams. This included:
The combined solution from Perfect Memory and Ontotext conceives semantics as the exploitation of all data by converting it into information, and therefore the exploitation of a company’s knowledge by the people working in it – the Company’s Chief Data and AI Officer
Harmonizing all this data brings little value if internal teams are unable to use the platform in their everyday work. Therefore, the custom platform has been designed to be accessible and intuitive, enabling data product delivery. It allows data managers to be autonomous in the way they orchestrate data recovery, data processing and data sharing.
The platform makes it easy to document all data items about each TV program and to manage a significant quantity of tags. It also introduces a frame-accurate player. All this increases the accuracy and relevance of search results and optimizes user experience. Now, users spend more time uncovering valuable insights, and less time finding (potentially faulty) information.
Perfect Memory was able to create a BtoB interface as intuitive as a BtoC interface. – the Company’s Chief Data and AI Officer
The introduction of data and content lifecycle governance – from creation to monetization, transformation, enrichment, and sharing – is a real paradigm shift in the broadcast world. Perfect Memory’s platform, powered by Ontotext’s GraphDB, covers all of these stages. The joint solution makes it easy to:
Our main goal is to bring together all the data entrusted to us by the producers, together with data enriched by our artificial intelligence solutions in order to measure our level of knowledge for each program – the Company’s Chief Data and AI Officer
With the new Data Programs platform developed by Perfect Memory and powered by Ontotext GraphDB, the television broadcaster was able to streamline the management of their non-textual media data:
Apart from these figures, the expected return on investment can be best measured by how much easier it is for teams to locate data, irrespective of format, to honor editorial rights periods, share with others who need the insights, and be sure they are using accurate information to drive decisions.