The second edition of Big Data Spain (bigdataspain.org), a landmark event about massive data processing applications and technologies, took place this November in Madrid. The strength of this industry was highlighted by the increasingly critical applications of big data in sectors such as banking and telecommunications and some new technological and business trends: systems that facilitate online analysis of large volumes of data and new paradigms of query to ease the development and visualization to communicate results effectively.
Without doubts, one of the key areas of big data application is the processing of social content to obtain all types of insights, a field where the typical requirements of Volume, Variety and Velocity (the three V's of big data) come to a new dimension. In particular, the phenomenon of social TV, in which viewers wish to augment their television consumption experience through the conversation in social media, the sharing of information and the exploration of content related to their preferences. It is so much so that the big social networks attribute most of their traffic during the prime time to the conversations about TV shows.
But social TV is not only important for viewers. Broadcasters and brands have joined the phenomenon not only to participate in the conversation, but also to understand and leverage it: to profile their audience, analyze its reactions, provide complementary content, insert relevant advertising, analyze the impact of their programs/campaigns or foster the audience's loyalty.
Social TV is not the typical scenario of large volume of structured data: unstructured, multimedia and multilingual content are much more difficult to treat. Storage platforms are not enough: it is necessary to structure and extract the meaning from that content before being able to analyze it and extract mentions, trends and relationships.
In the presentation "Real-time Semantic Search Engine for Social TV Streams" of the past Big Data Spain, Daedalus (daedalus.es) presented its developments on a platform of real-time analysis of data streams for social TV. This platform combines advanced storage and retrieval of unstructured information capabilities with semantic treatment for the extraction of elements of meaning and connection to external information.
For this purpose Daedalus uses its semantic APIs service in the cloud, Textalytics - Meaning as a Service (textalytics.com), which provides high level functionality to detect mentions of people or brands, identify topics, extract concepts, opinions, relationships, user profiles, enrich with linked data, etc.
"Social TV is not only a problem of big data. Only by extracting automatically the meaning of what is being said It is possible to provide viewers with tools to navigate the tangle of conversations, contextually relevant information and recommendations adapted to their profiles.”, states César de Pablo, semantic APIs evangelist at Daedalus. "Likewise, brands and broadcasters can understand the conversation and extract valuable insights about their audience, way beyond the mere audience rating."
You can access the presentation and the video of the talk on the Blog of Daedalus.