Ranking Events in Google Search Results

Sharing is caring!

Ranking Events without Links

This summer, Google was granted a patent that describes how the search engine might rank events based upon data that might indicate the popularity of those events, without relying on things such as the number of links pointed to pages about those events. The patent involves ranking events that occur in physical locations.

Examples of the kinds of events talked about in this patent include such things as music concerts, art exhibits, and athletic contests, all happening for specified periods of times at specified physical locations, such as concert halls, galleries, stadiums, or museums.

Since many events in a geographic region can happen at the same time or at overlapping times, interested individuals may at times find it difficult to determine which events to attend. For example, individuals may be unaware that events of interest are scheduled to occur or may have difficulty identifying the most interesting events when multiple events are occurring.

This ranking events patent lays out a general process flow to describe how the method in the patent works. It starts with receiving data about a physical location, and events taking place there during a certain time period and computing signal scores for those events based upon things such as a mention of the event and a popularity score for the event based upon those signal scores.

Signal Scores for Events

Additional signals for ranking events can include:

1) generating an initial ranking of events based on the popularity scores;
2) computing a respective modified popularity score for each of the events based on the initial ranking, and
3) generating the ranking of events occurring in the physical location by ranking events according to the modified popularity scores.

The patent describes the process of computing a popularity score for events, including:

1) obtaining information about a category for each event;
2) computing demotion values for each of the events, which is based on higher ranked events in the same category and
3) generating the respective modified popularity score for each event by applying the demotion value for the event to the popularity score for the event.

Computing a first signal score for the event can include:

1) determining, of the Internet sites including at least one mention of the event, a number that have been classified as ticket selling sites; and
2) computing the first signal score based at least in part on the count of Internet sites including at least one mention of the event and the number that have been classified as ticket selling sites.

Computing a plurality of signal scores for each of the events can also include:

1) determining whether the event has an official web page; when the event has an official web page,
2) determining a peak number of user selections of the official web page over a predetermined duration of time;
3) determining a measure of relevance of the official web page to the event; and
4) computing a second signal score of the plurality of signal scores for the event based at least in part on the peak number of user selections and the relevance measure.

Other signal score factors for ranking events can include:

1) obtaining data identifying one or more entities that are relevant to the event;
2) determining a measure of global popularity of each of the entities; and
3) computing a third signal score of the plurality of signal scores for the event based at least in part on the measures of global popularity of the entities that are relevant to the event.

Additional signal scores can also include:

1) obtaining data identifying one or more event categories that are relevant to the event;
2) determining whether any of the event categories that are relevant to the event have been classified as promoted or demoted categories; and
3) computing a fourth signal score of the plurality of signal scores for the event based at least in part on whether any of the event categories have been classified as promoted or demoted categories.

More signal scores may include:

1) obtaining search results for a search query that includes a first term identifying the physical location and a second term indicating an interest in events occurring in the physical location;
2) determining a position in a ranking of the search results of a highest-ranked search result that mentions the event;
3) determining a frequency with which the event is mentioned in the search results; and
4) computing a fifth signal score of the plurality of signal scores for the event based at least in part on the position of the highest-ranked search result that mentions the event and the frequency with which the event is mentioned in the search results.

Computing additional signal scores can include: 1) obtaining data identifying a venue hosting the event; 2) obtaining data identifying a seating capacity of the venue, and 3) computing a sixth signal score of the plurality of signal scores for the event based at least in part on the seating capacity of the venue.

The following advantages are described by the patent in following the approach it describes.

1) Events in a given location can be ranked so that popular or interesting events can be easily identified.
2) The ranking can be adjusted to ensure that highly-ranked events are diverse and different from one another.
3) Events matching a variety of event criteria can be ranked so that popular or interesting events can be easily identified.
4) The ranking can be provided to other systems or services that can use the ranking to enhance the user experience. For example, a search engine can use the ranking to identify the most popular events that are relevant to a received search query and present the most popular events to the user in response to the received query.
5) A recommendation engine can use the ranking to provide information identifying popular or interesting events to users that match the users’ interests.

Ranking events
US 9424360 B2
Publication number US9424360 B2
Granted: Aug 23, 2016
Filing date: March 12, 2013
Also published as US20150161128
Inventors Kavi J. Goel, Toshihiro Yoshino, Yang-hua Chu, Hidetoshi Shimokawa, Slaven Bilac, Mingmin Xie, and Satoru Yamauchi
Original Assignee Google

Abstract:

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for ranking events. One of the methods includes receiving data identifying a physical location; obtaining data identifying a plurality of events occurring in the physical location during a particular time period; computing a respective plurality of signal scores for each of the events, wherein computing the respective plurality of signal scores for each of the events comprises computing a first signal score for each of the events based at least in part on a count of Internet sites that include at least one mention of the event; computing a respective popularity score for each of the plurality of events by combining the respective plurality of signal scores for the event; and generating a ranking of events occurring in the physical location during the particular time period based at least in part on the popularity scores.

The selection scores may include things such as Unique Mention scores, Ticketing Sites mentions, and Official Page (of the event) selections, and the patent describes how those factor into popularity scores.

An entity popularity score is a measure of popularity of entities that have been classified as being relevant to the event. Each entity may be given a topicality score and a confidence score, based upon the global popularity of the obtained entities. This can be based in part upon a number of search queries that include a reference to the entity that has been submitted to a search engine.

An Event Category Score might be used to determine whether the category has been previously classified as a promoted or demoted category. A trade show may be classified as a demoted category because they tend to appeal to a limited audience, while festivals and fairs may be classified as a promoted category because they tend to appeal to a broader audience but may not be well-publicized.

In order to compute the ranking score, the system obtains search results for the search query from a search engine. The obtained search results are ranked according to scores generated by the search engine. The system can then compute the ranking score based on a position in the ranking of search results of the highest-ranked search result that identifies a resource that mentions the event, the number of mentions of the event in resources identified by a pre-determined number of highest-ranked search results, or both.

A score based on the popularity of the venue where an event takes place may also play a role in the ranking score of an event. Events that are held in places that usually show more popular events may be assigned higher venue scores.

Sharing is caring!

36 thoughts on “Ranking Events in Google Search Results”

  1. Hi David,

    I ran across this one this morning, and thought it was interesting that it came up with a different way of ranking events that didn’t rely directly upon links, but was a way of ordering search results that had more to do with other metrics. I think we will see more things getting ranked in unique ways like this. 🙂

  2. Hi there Bill,

    You really shared nice ideas here, because the events I think are a huge profit for Seo’s

    Regards

  3. Great writeup Bill. I’ve always said that Google will eventually need to get away from links and this is an interesting case where they might test it out. I use event markup frequently, and it seems that there could be a large advantage in the future for data related to events.

    SEOs that can provide this data will be in high demand.

  4. Hi Ryan,

    It is interesting hearing about Google applying new signals as they describe in this patent. It will also be interesting seeing them come up with other metrics for ranking other things as well. SEO definitely is evolving.

  5. Hi Bill,
    Topicality, authority and other signals that are less easily manipulated all point to more congruent search results.
    Thanks
    Bren

  6. Hi Bren,

    The metrics involved in this approach to ranking events do appear to be less manipulatable. I predict we will see more of that happening over time.

  7. Thanks for the good information. I love reading your blogs daily. I am getting more information.

  8. Bill,

    I’m curious if you see this patent as describing a process that could be described as another “layer” if you will of ranking criteria on top of existing criteria (such as links, etc.). I’m mostly wondering that based on some of the scoring criteria you described including the first part of this paragraph:

    “In order to compute the ranking score, the system obtains search results for the search query from a search engine. The obtained search results are ranked according to scores generated by the search engine. The system can then compute the ranking score based on a position…”

    That sounds to me like as part of this process the system could essentially be utilizing data that has already gone through the typical ranking process that would include links–am I inferring something that isn’t there/misunderstanding?

    I certainly agree this is VERY interesting to see insights into how they might be ordering results without links in this situation.

    Thanks!

  9. Hi Todd,

    Yes, I would call this a reranking approach from Google. Another example of a reranking approach is Phrase-Based indexing, which uses signals such as relevance and PageRank to identify the top ranked pages for specific terms, and then identifies co-occurring phrases on those top ranked pages, and then may boost or reduce pages in rankings based upon their use of those co-occurring terms. The signals described in the patent do rely upon the older metrics being used first, to generate new information about new metrics. They haven’t abandoned the older metrics completely, but they do seem to be stepping away from them.

  10. Thanks so much Bill for the reply. I was thinking of a similar reranking example to what you gave and glad to hear what I was teasing out was accurate.

    Have a great rest of your week!

  11. Hi Todd,

    You’re welcome. Yes, I’ve written about a few different reranking approaches over the years, and it was good that you caught that was going on. Knowing how different metrics from the search engines might fit together is a good thing to be aware of.

    You have a great week, too.

  12. Thanks Bill,

    I just bookmarked this blog and will be checking back. I am a full time SEO and love the theory that attempts to make sense of what Google tries to implement, often with a bang, and then revises or retracts quietly, and now mostly on weekends. 😉

    Google rankings irrespective of links, is like implementing martial law to stave off a few late night burglaries. Personally I can live with being unsure about the quality of the Russian Viagra I bought with Bitcoin or whether those were in fact real goose feathers in the jacket I purchased online from that ‘any domain with “Canada” in it’ site that’s always up. I can live with all of that in exchange for local search results that are basically perfect as far as I can tell. I’ve never looked online to order dinner, ordered Chinese food, picked it up, only to get home and realize I’d actually purchased pornography. Damn those local spammers!!

    Rankings that ignore the logic of links would likely be somewhere between terrible all the way down to just a little better than Bing. And Google doesn’t seem all the fussed about the use of networks that manipulate results through linking. “Private” blog networks are openly for sale.

    Ultimately no one likes to play a game when they know others are cheating. But is it cheating? White hat SEO and spot on on page optimization ranks sites almost as fast as building out big networks of costly domains on dodgy and expensive hosting. It sounds like a lot of work and expensive, which sounds a lot like SEO that actually ranks sites.

    Thanks,

    Richard Conover
    Think Tank SEO

  13. The future of seo is certainly interesting. Surely Google will be planning to move away from links at some point. Time will tell but an insightful article Bill. Cheers

  14. Hi Matt,

    There are a lot of interesting things happening in the world of SEO and promoting websites; and there are a lot of possible changes that will be coming along as well. That is one of the reasons why I try to keep up with patents like this one, to see what might be a little down the road from us. I do believe that Google will be moving away from links at some point.

  15. Yes for sure Bill, I suppose for their sake they will want to but a long and difficult task its probably been for them seeing as it has been originally built upon links. Reading those patents gives me nightmares 🙂

  16. Hi Matt,

    Google has had a lot of success with processes such as PageRank, But it does seem like searchers are enjoying new things from the search engines such as question answering, where the search engines provide answers directly, and searchers don’t have to dig though a website to find the facts and information they are looking for.

    I find myself fascinated with the futures that patents provide hints of, and wondering which might come to pass can be pretty exciting.

  17. Correct me if I am wrong, but as I understand your interpretation of the patent, they are saying that they will use mentions of events on ticket selling sites versus links to determine popularity?

    What would be interesting to see is which ticket sites do they use? i.e. are they going to favor big brand ticket sites or will even smaller affiliate ticket sites serve as an indication, especially if they are ranking for the event themselves?

    I don’t know, this seems to me to be an easy thing to game in the end and I’m not so sure it was well thought out.

  18. Hi Clint,

    I have included a link to the patent, so you can read that, and try to interpret it for yourself, if you would like

    It does appear that they are trying to determine popularity by counting how often an event is mentioned on as many sites as it is, rather than using PageRank to determine how popular an event may be. They don’t say, but I imagine it’s possible that they might not include some sites in their count if those sites are attempts to manipulate this process. That wouldn’t be a surprise.

  19. Pretty incredible stuff. While backlinks are probably always going to play a major role, do you think this same methodology could be used in the local business ranking algorithms. I know Google has changed a lot in their local search algo after internet marketers found out how easy it was to manipulate.

  20. Hi bill,

    A very good article for newbies like me. No one can estimate the upcoming algorithm of Google. Once they will take backlinks, once speed, once Responsiveness as their algorithm. All bloggers must be aware of new algorithms of Google. Thank you for this wonderful post. It would be very helpful for me.

  21. Hi Richard,

    Thank you. It’s good seeing you appreciate my blog. There are a lot of questions and a lot of approaches in action at Google. It is good seeing all of the experimentation taking place.

  22. Every time they will add more and more additional signals to rank a page. I think, using backlinks is old school, backlinks can be manipulated, but it is very difficult to manipulate those signals.

  23. Hi Rafael,

    I think the additional signals that they are adding to rank events gives value to the events that they list – by making sure that they are showing events that people might be most interested in seeing. Yes, those signals are more difficult to fake than links might be, and much more meaningful, too.

  24. Thanks to share with us The knowledge grap can be very interesting for the SEO but , Google don’t communicate to much about it…I am sharing this post…

Comments are closed.