Innovation and AI: Frankfurt Book Fair offers insight into how the publishing industry is embracing digital transformation
75 years of the Frankfurt Book Fair and we were right in the middle of it – with 215,000 visitors, the 2023 Fair was a great success. The world’s largest book fair demonstrated how digital transformation is changing the publishing industry and increasing the demand for creative quality content. 580 years after its invention, the book is by no means an antiquated medium.
The publishing industry is increasingly adapting to new challenges such as competition from other media formats, growing environmental awareness and the demand for more transparency and interactivity. The boundaries between media are becoming increasingly blurred and offerings are becoming more multimodal, including in the context of AI.
We were able to present ourselves in one of the start-up pods during the fair and give an impulse on AI at the breakfast of the German Publishers & Booksellers Association thanks to our hosts, MFG Baden-Württemberg and our state government with THE LÄND.
The host, State Secretary Dr Patrick Rapp, emphasised the importance of the ethical and legal debate on AI and why he sees Baden-Württemberg and its IPAI as a world leader in the future. The Stuttgart colleagues from Springer Nature also gave a very good insight into the adaptation of AI and algorithms in their publishing work and how they started ten years ago. The latest highlight: a book about AI, written by AI. The ideal environment to exchange ideas on innovation with publishers.
But it’s not just generative AI that has found its way into the publishing world; Christian Hirth spoke on behalf of DeepVA about how we’ve been building AI solutions for broadcasting and its internal data management since 2018, and how large media companies are dealing with this development. Especially the practical examples and approaches of public broadcasters seemed to be of interest, which is no wonder – we were back to back with the Deutschlandfunk stand.
A small personal highlight: the Consul of the US Embassy left his large entourage of security personnel for a brief moment to take a private selfie at THE LÄND stand. A sign that “The Future of Publishing” made in THE LÄND is working.
A walk through the halls of the Book Fair led to some valuable conversations with (audiobook) publishers and software providers and clearly showed us that the industry has a great need for AI solutions and that the pressure to innovate is palpable.
The Frankfurt Book Fair 2023 was therefore a feast for all the senses and a sign that literature will always play an important role in our society, no matter what changes are taking place.
But our AI isn’t just for the broadcast and media industry. Wherever large amounts of data are generated and need to be processed, our AI capabilities are a valuable component in maximising the efficiency of the management and use of audio, image and video data.
DeepVA: Beyond the limits of ChatGPT - Holistic AI functionality combined with LLM for comprehensive data analysis
Comprehensive and sales-promoting additional information on audiobooks or book titles is an essential component of the internal management of (audiobook) publishers, i.e. of book and audiobook data, as well as of the sale of these in online shops.
This information is already provided to publishers in a standardised form as promotional or bibliographic information.
But how can I automatically extract information from my audiobooks into my internal systems and online stores that is more relevant to my customers' interests and leads to increased sales?
Unlike ChatGPT, which is currently unable to read metadata and content from the direct software, DeepVA is able to automatically recognise image, video and audio data and enrich it with metadata.
DeepVA offers a comprehensive solution for applying AI to transcripts to optimise the management and personalisation of content in your internal systems and online shops.
This solution includes the automated speech-to-text function, which automatically creates transcripts from audio data, which can then be automatically transformed into short summaries, marketing slogans, short portraits, blurbs, forewords, or many other desired texts.
When linked to a LLM – or Large Language Model – DeepVA provides additional options for generating customised information from data and creating enhanced text.
In combination with our other AI services, such as face recognition, speaker identification, visual concepts recognition, landmark recognition and logo recognition, any additional information can be extracted from audio and video data.
For example, an audiobook file can be transcribed to identify the speaker, entities, key messages, special highlights, emotions, content information, slogans and more.
This information helps audiobook publishers to generate more reader-relevant and sales-promoting information, as well as to optimise and automate their keywording in internal systems or online platforms.
What results can be achieved?
- Efficient management of book and audiobook content: Automated keywording and content analysis enables time-saving management and optimisation of the product collection in internal CMS/DAM systems and online shops.
- Improved shopping experience: Customers can find books and audiobooks more easily and receive personalised recommendations, leading to increased sales.
- Detailed additional and marketing-relevant text through the combination of speech-to-text, LLM connection and AI services such as face or object recognition.
- A better understanding of reading behaviour: DeepVA enables the analysis of customer reading behaviour to continuously improve the offer and increase customer satisfaction.
