Search as an Editorial Tool - New Forms of High-Value Content Aggregation
As the search engine marketplace consolidates there are new ways of looking at content that go beyond traditional search engine technologies into forms and formats that place more emphasis on enabling editorial capabilities and complete content products.
Yesterday’s search results page is becoming one of many filtered feeds that support today’s mined analytics, rich editorial output and on-the-fly aggregation tools. How do search technologies create marketable content value today - and how do more valuable content products get built using search technologies?
Moderator
Peter Jackson, Chief Scientist, Thomson Corporation
Panelists
Tom Aley, President and CEO, Generate Inc.
Robin Johnson, CEO, Financial Times Search
Bruce Molloy, CEO, Connotate Technologies, Inc.
Patrick Spain, CEO, Highbeam Research, Inc.
Peter: Focus is going beyond the portal to applications that can allow alerts, summarization.
Johnson: Moving to contexts in workflow, in businesses but also in consumer markets. Some queries are impossible to do in Google, so we need to develop facets, disambiguate things like Ford Modeling Agency from Ford Motor Company. No summarizations of information on Internet, nothing that says on balance this much is positive and negative, just some high-end tools providing this for business. Have to focus consumer on the nature of the question, can’t always be translated into keywords.
Patrick: Free reference site, subscription HighBeam site, Newser, graphic content highlighting. Started to figure out if Web 2.0 is applicable for our business. Concluded that most people aren’t good writers but many are good collectors and recommenders, can vote for content and use search tools to find things elsewhere in product. In Newser we write all story summaries, now users will be able to write about stories that they’re enthusiastic about.
Peter: How do you make sources of information more active?
Molloy: Intelligent software agents monitor a site for changes and create alert streams, can make these sources much more useful. Search can go only far, what do you do with the documents? You’ll cut, paste and compile. Need a search component but also need an automation component to aggregate content more effectively.
Aley: I don’t want to get out of an article to find related information, our tooolbar widget enables people to find rich data in contcxt. We’ve all been saying for years that users need content in their workflow, to be able to embed that information in any page is a way to turn any page into a contextual portal.
Peter: What are the real-world scenarios for workflow that these products help.
Johnson: Imagine a tool that can help a journalist to prepare for an interview, if you have a tool that you could use to rearrange content is difficult. You can’t just do it with keyword search, experimenting with semantic navigation, keeping track of content relationships. We think that this is the way to make workflow come alive.
Peter: What’s feasible right now?
Molloy: Our technology enables end-users to identify data elements in a page rapidly, in terms of workflow, AI mining technology combined with this end-user interface can empower the end-user and bypass I.T. to get their sophisticated needs met. They can go a step further and put their mining agents in a common library that other people in their organization can use and modify to create even more value.
Peter: The “m” word, metadata?
Aley: Technology has come so far in past 5-10 years, intelligent crawling of 70 million domains to extract unstructured content into a structured format. Many publishers and enterprises have terabytes of their own unstructured content, we can extract and integrate it all. “Keyword search is so yesterday.” With robust metadata the query to be interpreted more effectively.
Johnson: You don’t have metadata, you won’t get there. “Customers like you asked these questions” can be created with metadata, hard to do without it.
Aley: More sophisticated, now seeing that an entity is an automotive ad agency instead of an automobile manufacturer.
Peter: What kinds of tools do you create to help users?
Spain: Looked at how we could use data exhaust to help user experience. All stories tagged on scale of hard or soft news, an interface control lets them choose ratio. Interesting - 10:1 people set the controls for hard news but 3:1 look at soft news. Looking at controls that will help people understand what other people are looking at.
Peter: Standards? [comment: have heard too many questions about standards today.]
Johnson: IBM’s open standard may allow some data exchange, but not much adoption.
Molloy: Hard to get people to agree to metadata standards, but with that being said it’s possible to impose metadata standards, but we don’t believe that a grand taxonomy will emerge, we allow industries or companies to create their own standards or map to your consortium’s standards.
Peter: What would you do in the future if the technology was there, how close are we?
Johnson: Annotate on the fly, managing copyright can make this tricky
Spain: Amazed at how different every Web site is, many different ways that archives are managed,
Aley: Looking at dynamic “version of you” and put it in the system, if I have a LinkedIn page, Facebook page, several Zoominfo profiles, can we integrate this. Being able to integrate on the fly, slamming together software and data has been the dream, as the Web world evolves we’re seeing that it’s achievable.
Molloy: See a move from search to business intelligence tools. [Slide of antique “computing division” with dozens of people entering data into mechanical calculators] we see agents extracting but also providing value add processing for alerts and other functions.
Question: How truly aspirational is this?
Aley: Many publishers have legacy platforms, can overlay infrastructure that they have, can work with Generate to free that informatoin without interrupting underlying operations. When we dynamically understand what’s on the page we can match entities and match it with behind the firewall content.
Molloy: We see a real eagerness from publishers to understand this, see need to catch up with search engines, the idea of monitoring and aggregate components on the Web is one model, or create your own kind of an alerting system. Editorial awareness and situational awareness, being able to monitor everything and choose what needs to be seen. Not only a cost reduction factor but also for new revenue streams, new products, new inventory.
Johnson: All leveraging off of creating engagement, make things more sticky, have to invest in what will make them do this, this is what will make C-levels invest.
Question: Proprietary metadata as a competitive value?
Spain: At the end of the day all you may have is metadata.
Aley: Extremely valuable to our clients, DMOZ helps to accelerate.
Molloy: Metadata important, tagging and folksonomies may add value also.
Question: Copyright presents issues, to what extent does copyright and licensing agreements present barriers?
Aley: We process data that’s either proprietary or public domain, very black and white in U.S., privacy laws overseas make it harder to expose this information. Big focus is to maintain copyright.
Peter: Publishers and vendors are cozy right now, but vendors are creating content and content value. Is it always going to be as cozy as this?
Johnson: Isn’t always entirely cozy when there’s intellectual property involved. Software companies survive on reselling ideas gained from working with clients. FT survives with technology, coopetition has required non-competes and to lay out how we dance when a new idea is discovered.
Molloy: A blurred line.
Aley: Reed Elsevier with scientific journals, owned market share for 80 years, Reuters buys ClearForest buys software companies, publishers thinking about whether they want to buy technology from a publishing competitor.
Spain: Web has changed fundamentally what fair use is. It’s all about chunks of data, chunks are hard to protect with copyright. The smaller it is the more valuable it is oftentimes.
Great panel, search is transforming editorial operations radically, the database is now, the metadata and other contextual content features are key to proprietary advantage, own the context and you own the publishing relationship.
- John Blossom, Shore Communications
The Information Industry Summit is the digital information industry’s flagship conference. It provides strategic guidance to senior business leaders representing publishers, content technology companies, bankers, analysts and press.
Jason Kichline is ScribeMedia's project manager and producer of ScribeMedia's Emmy Award winning series Reporting AIDS. He likes typos, fast food and MacGyver like solutions to life's nagging problems.
Discussion
No comments for “Search as an Editorial Tool - New Forms of High-Value Content Aggregation”
Post a comment