Edinburgh based data specialists Wallscope were selected to respond to a Creative Informatics Challenge Project set by Scottish podcast network The Big Light, who were looking to evaluate and improve their data processes, contextually linking and visualising the data in their growing archives to develop personalised user experiences that deepen user engagement with podcast content.
We asked the team at Wallscope to tell us more about the process they followed and the outcomes achieved through their work with The Big Light. Here’s what they had to say…
At Wallscope we believe that knowledge should be accessible to all. Our aim is to empower organisations to navigate and link ever-increasing volumes of information, and to present this in an understandable and engaging way.
Using our knowledge and experience of data solutions, we support organisations who seek sustainable transformation and innovation. Through knowledge mapping processes we generate insights that power change to support both organisational goals, improve existing processes and embed new technologies.
The Big Light’s Challenge Project gave us an opportunity to explore areas of possible innovation in the broader podcasting space and identify new opportunities for business development at both The Big Light and Wallscope.
We used an existing series of podcasts from The Big Light’s archive to explore how they could be best repurposed and repackaged. Currently, standard podcasting formats are rather ‘basic’ and are specifically tied to the operational requirements of large podcasting services such as Apple, ACAST, Spotify and others, all of which use a standardised RSS feed format. Because of this, we decided to focus on the human-level interactions which come from the intimate (analogue) space where audio is listened to using headphones to provide an ‘aural’ experience.
While the delivery is driven by digital technologies, listening is a ‘lived’ experience that can take place in many settings.
Podcasting is an active area of growth that leans into the power of the spoken word to tell fictional and non-fiction stories. Storytelling is a powerful point of engagement, impacting and reaching into many lives with a high level of personalisation. However the podcasting sector can struggle to identify key insights into how listeners interact with the discovery of content, the actual listening experience and the demographics of their listeners.
This project presented us with an opportunity to explore the ‘temporal’ dimension associated with audio-form files and how ‘time’ is key to how stories are found, told, recorded and consumed.
Early on in the Challenge Project process we realised that we needed to establish an efficient process for cataloguing and indexing audio-format files, like books in a library, allowing them to be easily ‘found’ and accessed through descriptive and illustrative materials directly referenced and associated to not just one file, but across a series of files.
We also utilised the metadata data within each file, which can be edited and augmented with elements such as ‘tags’ and copyright, providing insight into the quality of a file as well as the length of a recording.
We reviewed a range of technologies associated with data access, retrieval, insight and playback. We were specifically interested in areas of Artificial Intelligence and Machine Learning that could be used for automating the cataloguing of audio files and unlocking the audio content to improve listener engagement and possible commercial outcomes.
We were keen to identify an area for innovation at a sector-level, but we struggled to find this as the podcasting sector is very bound-up in modes of delivery based on large scale platforms, where the economics clearly run towards the platforms. The barriers to entry for actual podcast creators are low and the risks are borne by them. Taking this into account we identified the creation of quality recorded and produced original stories with compelling narratives and dialogue which contain lived, experienced, learnt or fictional content as the key area for innovation.
We then looked at how the ‘temporal’ value of storytelling could be used to better engage users across the listening experience, and provide points of knowledge expansion, engagement and ‘socialisation’ of content in web-based environments, creating a richer and more meaningful experience for listeners that runs in parallel with podcasting platforms.
THINK ANALOGUE, DO DIGITAL
We looked backwards at the concept of a wax-tablet as a means of recording areas and points of interest. In ancient times these analogue devices were used in a range of settings. One important use was by students who would listen to the voices of the Ancient philosophers as they spoke their ‘minds’. Yet if the wax was too soft or too hard, the listeners would struggle to organise and record their own sense of structure and ideas. (Digital technologies do not present that problem.)
Wax cylinders were used to record audio by capturing sounds on a timeline, this could be played back and listened to. (Digital technologies allow audio to be readily and accurately reproduced and distributed.)
These reflections enabled us to consider the array of technologies we needed to align so we could allow storytelling to prompt and enable better learning from spoken word experiences that are semantically meaningful – a bit like creating a ‘bookmark’ experience for the listener which is significant in terms of both ‘time’ and ‘content’.
Processing the transcript in ‘chunks’, we applied AI techniques which allowed Named Entity Recognition (NER), Natural Language Processing (NLP) and Machine Learning (ML) to expose, expand and interact with key identifiers such as people, places, organisation, concepts and topics.
These data points could then be programmatically used in many ways, allowing listeners to express, share and expand their knowledge. This would allow a community of interest to be supported by presenting the timeline as a point for interaction- based on the power of the spoken word. It could be based on a location, a topic, a brand, a person or an organisation, or all of these. It can be placed within the data fabric of the Web, and expand the value of podcasting as part of a wider learning endeavour, as well as within a digital marketing budget. Or, to better support learners by combining the power of listening and reading, which could be used to learn a new language, educate a workforce or in other self-directed learning environments.
LISTEN – LIKE – SHARE – JOIN – SEARCH – DISCOVER – LEARN – ENJOY!
Through our Challenge Project we were able to configure existing technologies to provide a platform level service which can be provided as an Application Programming Interface (API) Service, giving The Big Light the potential to provide a new ‘offer’ in the podcast market, creating a ‘community’ around active listeners, which can prosper through bespoke websites, as well as providing improved integrations with existing digital marketing activities.
You can find out more about Wallscope’s work at www.wallscope.co.uk.
Applications for our Round 6 Challenge Holders are open now! If your creative or cultural organisation is looking to power new ideas in its work using data or data-driven technologies, submit your Challenge Holder application by 5pm on Wednesday 3rd November. If you’d like to discuss your Challenge Project idea with the Creative Informatics team, get in touch at firstname.lastname@example.org.