What started with a 32-hour marathon session in May is, in June, the beginning of one of the most dedicated efforts to solve the problem of online book discovery. The project is called Evoke and it’s being driven by a team of only three: entrepreneur and social scientist Jill Axline, developer and entrepreneur Lisa Maione, and coder Jason Pearson.
So, how will Evoke, the winner of the first industry “hackathon,” solve online book discovery? In a word: characters.
“The platform works to humanize online book discovery by setting [book] characters in relationship with one another based on various types of qualitative data,” explained Axline. “These data include readers’ emotional responses to characters; perceived relationships with characters; and attributions of roles and characteristics to characters.” Based on the information given in these elements, Evoke generates a personalized “character map,” which is meant to visualize similar characters in different books. Said Axline: “The character map, as a discovery tool, essentially allows readers to transcend literary genres while still retaining the type of person, relationship, or experience that facilitates their immersion into a narrative world and fuels their enjoyment of a story.”
Discovery is only one facet of Evoke, however: engagement with the characters is also a prominent feature. The platform will feature character-specific pages, including galleries of annotation composites (user-generated compilations of text), images, video, and audio. The goal is to spur a community dialogue around each character and enable “a character’s collective identity to empower a number of diverse perspectives in one visually stunning environment,” in Axline’s words.
That all may sound technical and theoretical, but Evoke’s character-first principle is looking to sidestep the difficulties prior online book-discovery efforts have experienced by putting humanity—and emotion (hence the name)—first. “Given the subjective, fluid, and highly emotional nature of what we’re proposing with Evoke, I understand the difficulties surrounding ‘data capture’ in this area,” said Axline, who noted that even though behavioral measures like purchasing and rating systems are easier to measure, they ignore the passionate, and often erratic, nature of emotional connection. “What we’re doing is really difficult and not altogether intuitive if you lack a nuanced understanding of the social science behind the emotions, roles, and relationships that are central to mediated interaction,” Axline said. “Ultimately, I think online recommendation is currently operating with the lowest hanging fruit and avoiding components of the human experience of book discovery that are not immediately translatable to a computer algorithm.”
Axline’s teammates, Maione and Pearson, echoed her stance and, in addition to emotional connection, stressed immersion. “If one is able to Google it and get a satisfactory answer, it is probably not a very interesting question,” said Maione. “And, you likely won’t ‘discover’ anything. To discover is to find something unexpectedly in the course of a search.”
After winning the Publishing Hackathon at BEA in late May and the $10,000 prize that came along with it, Evoke is on its way to creating a workable demo. The first character map is already being planned, with thousands of characters drawn from young adult books—Evoke’s first area of focus. In the map, the characters themselves, as well as their initial classifications, will be derived from a combination of scraping metadata and API feeds from publishers and reader input that is collected seamlessly as users navigate the Evoke site, explained Axline, who added that the way characters are “matched” is Evoke’s “secret sauce.” But among the talk of patenting its system, establishing itself as a company, and more, Evoke isn’t losing sight of the core of its mission. “The way to a story’s heart is through its characters,” said Maione. “Evoke will connect you to a character that you will relate to in a way that you enjoy. Every character leads to another.”
