Exploring Search (in the Context of the Browser) with Prospector

Prospector is a new series of experiments from Mozilla Labs focused on analyzing, experimenting and prototyping improvements on how you search and discover content with Firefox.

To start, we’re focusing on three main areas: websites you have visited, tabs you are currently viewing, and pages you haven’t visited yet. In each of these, there are different aspects of traversing through existing behavior, extracting relevant information and discovering better ways to search.

For example, if you search and find a page with what you are looking for, Firefox can be made aware of the words and sites used to search as well as which results were visited and for how long. Using that raw data, Firefox could learn to better streamline that search behavior and help you get what you want even faster – all directly in Firefox, without sending data to a server.

Firefox has always tried to help users find the websites they want to get to by including the search box on every window as well as defaulting to a search engine on new windows. Firefox 3 added the Awesome Bar for improved history search with suggestions of websites a user will likely re-visit. This behavior of highlighting items that a user frequently visits is something we see as just the first step in adapting Firefox to the current user’s behavior.

The Prospector project will contain a mixture of lightweight prototypes that try out various ideas as well as a few deeper explorations that might change multiple aspects of the Firefox user interface. These will be made available as add-ons that don’t require a browser restart and can be installed in Firefox 4. Most prototypes will not send any personal data to a server but focus on creating functionality directly in Firefox.

We aim to identify useful and exciting new ways for you to find what you are looking for faster and better – while putting you in control of your experience.

P.S.: Please note that Prospector is a Mozilla Labs exploration and thus not part of the Firefox roadmap.

9 responses

  1. Bella Lutzi wrote on :

    Ultimately, I founded the information I was wanting out for. We have now been taking out assignments on this subject, and for four days I preserve obtaining web-sites which are supposed to have what I am hunting for, only to get disappointed using the lack of what I wished. I wish I could have found your web pages sooner! I had about 25% of what I utilised to get in have to have of and your website has that, plus the rest of what I crucial to finish my researching. We now have activated to this internet site right here I like that you simply will notice unique report content that you’ll be capable to hardly discover elsewhere. One very good thing, you possibly can obtain nonetheless these sorts of sites, make certain you go on! I can no more time see the well-liked media. It can be there a great deal rubbish printed, I bear it no very much extra easily. A actually pleasant website and beneficial write-up. I invest nights inside of the entire world huge net learning blogs, about tons of numerous subjects. I should preliminary of all give kudos to whoever recognized your web sites and second of all to you for composing what i can only describe as an post. I honestly think there is a skill to writing articles or blog posts or blog site posts that only several posses and frankly you may have it. The mixture of insightful and fantastic subject material is totally exceedingly hard to find employing the substantial quantity of website round the online globe.Constantly hold a really very good give very good results!

  2. Jeffrey wrote on :

    I too want to know how to participate.

    I've been working on some ideas for a search client/manager that implements more OpenSearch specifications such as "response elements". This would make managing and controlling the way we send queries and receive results more integrated with the browser. (as opposed to just linking us to a results page)

    1. spencerhold wrote on :

      i really like this idea. I too think that opensearch was only half-way.
      would love to help with this project.

  3. Cybunk wrote on :

    Exited ! How do I participate ?

  4. fung0 wrote on :

    This sounds like a cool idea. The coolest part being the promise that it "will not send any personal data to a server." This MUST always remain an option, and I hope and trust that Firefox will continue to champion that approach.

    Too many trivial features nowadays, instead of being handled locally on my largely-idle quad-core CPU, are becoming pretexts for harvesting user information. Endless hollow promises of privacy can't make me trust this sort of behavior. If I can't avoid it in Firefox, I'll be doing my prospecting elsewhere.

    1. Edward Lee wrote on :

      @fung0: Privacy has always been a top priority with Firefox, and we would like to maintain that with our Labs experiments. But just to remind you, just like Firefox's search box will ask search providers to list word suggestions for what you've typed, our experiments may need to talk to servers to retrieve rich information. But again like the search box, it's a user opt-in behavior to trigger that functionality.

  5. Rodolfo wrote on :

    I applaud this idea, we need better tools for our tabs, history and ultimately our bookmarks.
    I think Prospector should use projects like Zootool http://zootool.com and Microsoft Pivot http://www.getpivot.com/ for inspiration.

    1. Edward Lee wrote on :

      @Rodolfo: Thanks for the links. Interesting ideas of collecting data/metadata about pages and making connections to other pages by those related ideas. This could definitely help improve search as typically you need to know the right key words to find the pages you want, and if you type the wrong one, you might get nothing. Whereas getting to the right area can remind you of what to look for next.

  6. Erik wrote on :

    How do I participate?