Search Expert Develops Rating System To Help Advertisers Buy Trusted Media

Pat Condo, a 35-year search veteran and founder of Seekr Technologies, believes his team has developed a search engine and scoring system that will provide a higher level of trust and reliability for marketers buying advertising that serves up across the internet and on websites.

Seekr Technologies, an internet technology company, launched the beta version of its search engine, Seekr, this week. The goal is to provide access to reliable information for people searching for answers, as well as a place for brands to advertise products and services alongside trusted content.

The engine will eventually support advertising, but it initially launched with what Condo calls the Seekr Score for news.

The Seekr Score rates each news article’s quality. Individual news articles containing political content are rated as right, center, or left through the Political Lean Indicator

The AI technology does this by extracting and analyzing the text for expressions, words, and semantics typically associated with a political position.

“Our content rating, filtering and other things we will launch in the future has become very interesting to brands,” he said. “Their biggest concern is the suitability of the content the ad falls alongside, keeping their brand identity [safe], and not being deemed political.”

Reports will support advertisers and publishers. It will, for example, include ZIP code-related information such as the percentage of people in the area that leans left or right politically, or what type of content they like best. It should help brands better understand the demographics, he said.

Seekr has processed an estimated 9 million queries since January 2022 -- when the company opened its Alpha testing.

Media rating systems from companies like NewsGuard that look for bias and misinformation exist, but the Seekr Score, in the future, will automatically evaluate the reliability of information across the web, from news to video, as well as from individual journalists.

The platform, built on an independent index, uses what the company calls Lite-Web Technology to serve news and the best web search results. 

It provides a scoring and filtering system to help brands as well as people searching make informed decisions on what they buy, consume, trust, and share. The goal is to have a common rating system that provides advertisers and consumers with a way to evaluate all web-based content.

Today, Seekr’s engine returns the majority of the queries. For long queries, it aggregates answers from search engines such as Microsoft Bing. “We pull back queries from them if we cannot answer them with our own index,” Condo said. “Over the next six months, a greater number of answers to queries will come from us, between eighty and ninety percent. The remainder will come from others.”

Seekr’s technology crawls the web, categorizes news content, and by analyzing the article, applies an algorithm built on machine learning to determine the score.

The score looks for a variety of behaviors using the ML modeling. For example, the score indicates credibility.

Condo founded Seekr Technologies in March 2021, built on a “great percentage” of the search technology from NTENT. Seekr built the scoring and filtering technology, and the two combined in the fourth quarter of 2021.

In the future, Seekr may offer its technology to media companies that want to capture and filter content to provide more accurate information to their readers.

Several years ago, Seekr developers examined variety of studies on best practices in journalism, with the help of experts in journalistic principles, to help create an initial set of algorithms targeted at “high-value signals," with dozens more planned to come.

These signals would determine the quality of an article’s content -- looking for things such as whether or not the headline matches the body copy, whether sources are cited, and whether bias is present. The technology would also determine whether the quote was accurate, as well as looking at other biases.

Pattern matching is a major part of the process. “We know satire, parody, comedy and opinion are complex, and there’s a chance they can be mis-rated,” he said. “Eventually, we will learn all the patterns and scoring will continue to exceed, at a human level, analyzing it better than we can.”

Condo also has plans to ramp up marketing efforts through a variety of social and paid media including influencers.

With his many years of expertise, Condo feel confident when ads launch on the platform, brands will advertise in search and news sections.

In the 1990s, Condo founded Excalibur Technologies, one of the first search companies to go public on Nasdaq. Within five years, he sold it to Intel for about $1.2 billion. The two companies formed a new company called Convera, a supplier of search technology for the defense industry. The technology enables content owners to produce and securely sell their audio and video online. Pieces of the company were later sold to Lockheed Martin, and Microsoft. Vertical Search Works came next.


 

5 comments about "Search Expert Develops Rating System To Help Advertisers Buy Trusted Media".
Check to receive email when comments are posted.
  1. John Grono from GAP Research, March 18, 2022 at 8:04 p.m.

    Very interesting.  


    • WHOis in 1982 searched for domains.

    • 1989 CERN's Tim Berners-Lee "invented" the web

    • Archie in 1990 was a pre-web content search of FTP

    • In 1992 Tim Berners-Lee created the Virtual Library

    • In June 1993 Matthew Gray released the first known web robot

    • In September 1993 W3Catalog was the first web search engine released

    • In October 1993 Aliweb (Archie Like Indexing for the Web) was announced and released in May 1994

    • In January 1994 Jerry Wang and David Filolaunched Yahoo!Directory



    TBL's "invention" of the WWW is just 33 years ago.

  2. Laurie Sullivan from lauriesullivan, March 19, 2022 at 10 a.m.

    Search and computers existed prior to TBL's "invention."

  3. Ed Papazian from Media Dynamics Inc, March 19, 2022 at 11:27 a.m.

    Laurie, I can see how this service might be helpful for political advertising and, maybe for advertisers trying to avoid extreme left or right websites  but I'm having difficulty seeing how it can determine whether the content is credible, "fake" or whatever.  For example if a post on a website is being sarcastic about some issue and, hence, seems to be repeating slogans that might type the writer as a fan of the cause or credo that he/she is bashing---when, actually, the opposite is intended---how does this service catagorize the post? It may be that there is a complicated  word association methodology that could do this but even so, who is determining what the correct interpretations are?

  4. Laurie Sullivan from lauriesullivan, March 19, 2022 at 12:18 p.m.

    I'm sure it's not a perfect science, but the company is working with a variety of linguistic experts and others. He said experts have done studies on this. I thought readers would find this interesting. We will see more of this type of technology. Who is to say the content is credible? A very difficult question to answer. 

  5. John Grono from GAP Research replied, March 19, 2022 at 6:15 p.m.

    Indeed Laurie.   Back at least to 1982.   I was referring to the lingua franca understanding of 'Search'.   In fact man-kind has 'searched' for centuries and in fact millenia.

Next story loading loading..