The Evolving Internet: Some Implications, Strategies, and Techniques for More Effective Research MSU Product Center September 26, 2007 Professor Larry G. Hamm Presentation Outline • • • • • • • Introduction Search Engine Basics Business Search with Google News Search Social Search Basic Information Trapping The Future?? QUESTIONS? • Who is Tim Berners-Lee? • What happened for “research” in 1990? Current Number of Websites July 2007-489,774,269 Top Global Web Properties Ranked by Total Unique Visitors (000)* June 2007 Total Worldwide, Age 15+ - Home and Work Locations Number(000’s) Percent Reach Total Unique Internet Visitors --- 778,310 100% Google Sites Microsoft Sites Yahoo! Sites Time Warner Network eBay Wikipedia Sites Fox Interactive Media Amazon Sites Apple Inc. Adobe Sites CNET Networks Ask Network Viacom Digital Lycos Sites The Mozilla Organization 544,783 529.155 471,924 266,367 264,732 208,120 163,545 145,947 123,554 121,966 116,579 115,655 88,654 77,517 70,850 70 68 61 34 34 27 21 19 16 16 15 15 11 10 9 Share of Online Searches by Engine August 2007 Total U.S. Home, Work and University Internet Users Source: comScore qSearch Aug 07 Total Internet Population 100% Google Sites 56.5 Yahoo! Sites 23.3 Microsoft Sites 11.3 Ask Network 4.5 Time Warner Network 4.5 * Excludes traffic from public computers such as Internet cafes or access from mobile phones or PDAs. Share of Online Searches by Engine August 2007 Total U.S. Home, Work and University Internet Users Source: comScore qSearch Aug 07 Total Number of Searches (Million) 9820 Google Sites 5545 Yahoo! Sites 2290 Microsoft Sites 1106 Ask Network 438 Time Warner Network 441 * Excludes traffic from public computers such as Internet cafes or access from mobile phones or PDAs. Herbert Simon, Nobel Prize Economist: “What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention” SOURCE: “Designing Organizations for an Information-Rich World,” in Donald M. Lamberton, ed., The Economics of Communication and Information (Cheltenham, England: Edward Elgar, 1997). The Source of Power? • Knowledge is no longer the “scarce” resource. • Attention is the “limiting factor”! • Implications: – Global--- Decisions on what is brought into global consciousness – Research --- Discipline to direct and control your attention The Role of ATTENTION THEREFORE: • “The most important function of attention is not taking information in, but screening it out.” Introduction The Meaning of Relevance Definition: The degree to which a search record (piece of information) meets the researchers’ query. • PROBLEM - Relevance to Search Engine and Researcher Are DIFFERENT • To a researcher: Does the result help answer the intent of the query? • To a Search Engine: Does the result meet the search engine’s ranking algorithm? Summary and Conclusion • Precision searching requires the process of consciously narrowing and eliminating the gap between researcher’s and search engine’s RELEVANCY • Knowledge of the search process and the characteristics of information sources are required to attack search engine relevance. • Intuition is required by the researcher to focus on formulating the search statements. Search Engine Basics • • • • The Invisible versus the Visible Web Defining and Identifying Search Engines How Search Engines Work Why Google? The Invisible Web • Great amounts of information exist than is not accessible via internet search engines • Much was formatted digitally but not ‘indexed’ (see latter lecture) • “Google Books” project is the grandest attempt to date to ‘shrink’ the invisible web. • Invisible Web information is differentiated by: – ACCESS – MODE of creation The Invisible Web (continued) • Information is differentiated by the nature of ACCESS to it: – 1.Publicly available --- Libraries – 2.Semi-public --- ‘Private’ Libraries i.e. MSU Libraries – 3.Private data --- Only available for purchase or through reciprocity The Invisible Web (continued) Types of Private Data – • Private data sets open to anyone with a checkbook (Mintel) • Restricted private data sets --- to contributors (Trade Association) • Proprietary data of individual firms/public institutions (Freedom of Information Act) • Spy data (commercial and public) • Private data interfaces with ‘Searchable’ data when private data firms use “free sample” or “versioning” marketing strategies The Invisible Web (continued) Differentiated by MODE of creation: • PRIMARY versus SECONDARY Data – Primary Data is data collected/generated through direct observation, survey, or poll – Secondary Data is data that is ‘repackaged’ primary data – Secondary data results from an ‘editing’ process – Evaluating secondary data requires an identification and evaluation of the base source(s) • Always go to the “ORIGINAL SOURCE”!!! The Visible Web Defining and Identifying Search Engines What is a search engine? • Definition – A search engine is an enormous database of websites compiled by a software robot that seeks out and indexes websites. How does it work? • Sends a ‘spider’ or ‘crawler’ to visit a Web page, finds the information on the page. • The ‘crawler’ then sends its “finds” to an indexer which takes every word on a Web page, logs it, categorizes it and than stores the results in a huge databases. Defining and Identifying Search Engines What types of search engines exist? www.searchengineshowdown.com www.lib.berkeley.edu/TeachingLib/Guides/Internet/FindInfo.html • General All Purpose Search Engines (Big 4) – Google; YahooSearch; Live.com; Ask.com • Metasearch Engines – Search engines that search other search engines (S.E. ‘bot’) – • www.dogpile.com • www.clusty.com • www.kartoo.com Defining and Identifying Search Engines What types of search engines exist? (continued) Specialized Search Engines (Vertical Search Engines) – Search engines dedicated for specific subject areas or specific purposes. For research: www.lii.org • “Customized Search Engine ” – Now anyone can create one www.google.com/coop/cse/ --See www.customsearchguide.com The Visible Web How Search Engines Work Search Engine – RANKING ALGORITHMS • WHAT? – Ranking Algorithms are used to ORDER the search results • WHY DOES ORDER MATTER? Answer - ATTENTION because the researcher wants ‘help’ in deciding relevance for the searcher's needs HOW? - Most ranking algorithms are and continue to be ordered by the frequency of use of the searched “WORDS” Google created a new addition to their Ranking Algorithm • • The Visible Web How Google Works 1. The web server sends the query to the index servers. The content inside the index servers is similar to the index in the back of a book - it tells which pages contain the words that match the query. 3. The search results are returned to the user in a fraction of a second. 2. The query travels to the doc servers, which actually retrieve the stored documents. Snippets are generated to describe each search result. The Visible Web Conclusion An Overview of a Basic Search • Be very proficient with ONE search engine • Remember because of different software approaches and indexing, NO TWO SEARCH ENGINES WILL PRODUCE THE SAME RESULTS • When very focused and search is narrowed, identify and use other specific engines • Should the “Product Center” create their own? Business Search with Google • • • • Translating Web Language Underlying Search Logic Understanding Google Search Features Conclusion Translating Web Language Reading URL’s – Uniform Resource Locator • This the Web site’s address; i.e. Were a Web site lives • Example: http://online.wsj.com/article/SB114609925357637113.html • http: - Transfer Protocol (hypertext) – the way the information is transfer on the Web. – HTML – Hypertext Markup Language is current Web language – XML (eXtensible Markup Language) is coming as the vehicle for information trapping Translating Web Language (continued) Reading URL’s (continued) www.online.wsj.com (domain name) of the server • Domain Suffix (com) – Perhaps the first and most important things to examine – Assigned by ICANN – Internet Corporation for Assigned Names and Numbers – www.icann.org – – – Country Codes (.uk) follow domain suffix (.us) not used by most U.S. sites except with state/local government sites. Current Issues? Translating Web Language (continued) Reading URL’s (continued) Common DOMAIN SUFFIXES • • • • • .com - commercial site .edu - educational institution .gov - government agency in the U.S. .net - network with most assigned to ISP networks .org - non-profit/non-commercial organization (Caution: many companies are setting up “non-profits” to get .org domain suffixes to disguise their agendas) • OTHERS - .mil, .biz, .info, .coop, .pro Underlying Search Logic Boolean Logic Searches • Definition - Use of mathematical set theory to retrieve search information. • AND, OR, and NOT searches • See following Venn diagrams: Underlying Search Logic (continued) Boolean Logic Searches - AND The Visible Web Why Google? (continued) Boolean Logic Searches - OR Underlying Search Logic (continued) Boolean Logic Searches - NOT The Visible Web Why Google? Google Has Two Basic Strengths Over Other Search Engines • • Popularity Ranking Number of and Breadth of Features The Visible Web Why Google? (continued) “Popularity” Ranking – “The Google Creation” • • A page’s ranking includes a score for how many “other pages” link to it i.e. How ‘popular’ it is with other Web sites This is done on multiple levels. For Example: If page X and Y both have 100 pages linked to them, but the 100 Y pages have more links to them than do the 100 X pages, Y gets a higher score for ranking The Visible Web Why Google? (continued) “Popularity” Ranking – “The Google Creation” (continued) THE UNDERLYING ASSUMPTION: • • A Web page that has more pages linked indirectly (like a pyramid scheme) to it implies that more pages find it relevant implying that it will be more relevant to you. Analogy – Your popularity is ranked within high school by how many friend your friends have and how many friends those friends have and so on. The Visible Web Why Google? (continued) “Popularity” Ranking – “The Google Creation” (continued) “THE GOOGLE BIAS” • New pages won’t have as many links as established pages; therefore a lower ranking. • Analogy: New friends might be better than the old friends. The Visible Web Why Google? (continued) Google’s Breadth of Features • Home Page Features – One of the Cleanest/Clutter Free Page • • Advanced Search Features Business research useful features are highlighted here The Visible Web Why Google? (continued) Google’s Advanced Search Features • Advanced features allow searchers to narrow their queries to very specific searches • Narrowed searches allow the gaps between ‘researcher’ and ‘search engine’ RELEVANCY to close much quicker • With precision query formulation, the search will be faster and more useful • 8 highlighted advanced features The Visible Web Why Google? (continued) 1.Google uses a modified Boolean Search Searches can be done from Google Home Page or from Advanced Features Page The Visible Web Why Google? (continued) • • • • • “Phrase Searching” Google automatically “ANDS” words Accepts one or more “OR’s” Use a minus sign in front of term to “NOT” it Google will not search on very common “STOP” words like “a”, “it”, and “the”. The Visible Web Why Google? (continued) • 2.Option to retrieve only a specific file format – (pdf), (ps), (xls), (ppt), (doc), (rtf) – Very useful if searching for a certain ‘type’ of data. For example: xls. and financial data. The Visible Web Why Google? (continued) 3. Date restrictions 4. Window to limit retrieval to title or URL fields 5. Box for limiting to (or excluding) a particular DOMAIN or URL The Visible Web Why Google? (continued) 6. Page Specific Searches: – for pages similar one to the entered URL – for pages that link to the entered URL Links to “Topic-Specific Searches” – for pages similar one to the entered URL – for pages that link to the entered URL 8. Domain specific searches for .gov, .mil, and .edu 7. Everything About Google?? • http://www.google.com/intl/en/help/refinesear ch.html#domain • http://www.google.com/intl/en/help/operato rs.html • http://www.google.com/intl/en/help/cheatsh eet.html • http://www.google.com/intl/en/help/feature s.html • http://www.google.com/options/ The Visible Web The Greatest Google Feature?? • Skip the Title - Click the cache? WHY? – – – – – Google ‘Highlights’ (different color) keywords/phrases No pop-ups that are attached to Web pages Faster – Google’s servers are the best in the world Allows for ‘text only’ versions Allows access when the current site is ‘unavailable’ The Visible Web The Greatest Google Feature?? • Further ‘Search’ Within the Result Generated Sites – If not in cache but titled page, use browser’s – “Find” button (Control+F) to show keywords/phrases – Use (Control+F) for NEW search with new words and phrases The Visible Web Conclusions • Is the desired information CONCEPTUAL or FACTUAL? • If Conceptual: Use in-depth research (library, books, scholarly journals, etc.) is most likely necessary to effectively frame the search. If Factual: A search engine web search can most likely proceed • • But always strive to find the “Original Source” The Visible Web Conclusions • Set a time limit - ‘Web Surfing’ can be addictive causing: – – – Tendencies to wonder off task Get attention ‘fatigue’ resulting in overlooking possible sources All other forms of destructive social and moral behaviors. News Searching What Do You Want: • Read news without ‘a paper, TV, or radio’? • Just see last second’s headline? • Find older stories? • Monitor an industry? • Other? NEWS SOURCE GENERATED INFORMATION Introduction • The evolution of news based information – Story telling town criers news posters/papers electronic news divisions the WEB • News is now a ‘commodity’ – Minimal costs of distribution – ‘Creation of news content’ is believed by many to be unrestricted (text messages, cell phone pictures, etc.) – Believed by many that with the ‘information democracy’ they have the “right” to create news and that their “news” is as legitimate as anyone else’s NEWS SOURCE GENERATED INFORMATION Introduction (continued) • News Differentiation Attention Merger of News & Entertainment – The Daily Show, The Colbert Report, etc. NEWS SOURCE GENERATED INFORMATION Five Specific Types of Web News Outlets • • • • • 1.Individual Online News Sites 2.Breaking News Aggregators 3.News Alert Services 4.Searchable News Data Bases 5.Industry News Sites NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 1.Individual Online News Sites – – – – – Definition – Migrations of existing established media outlets to Web based platform Examples: CNN.com, nytimes.com, onlinewsj.com Usually have graphics, delivery methods similar to parent outlet Mix of “free” and for fee services Most have archives with most of non-current for fees (NYT’s recent decision!) NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 2.Breaking News Aggregators – – – – Definition – Sites that pull material from multiple online news sources Usually limited to recent “Headline” material Use to do keyword search for relevant news articles Use when the individual site does not cover all possible relevant (geographic/minor stories) information NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 2.Breaking News Aggregators (continued) • • • Personalize one of the general portal sites (Google News, My Yahoo) and make it your “start page” Go to a “news service” site like BBC, CNN, MSNBC, etc. Go to favorite newspaper and set up an RSS feed NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 3.News Alert Services – – – – Definition – Same as breaking news aggregators except for a “User Profile” can be created Delivery method is via e-mail or Web site Useful if your particular interest is a company, product, topic, etc. Issues include: • • • Completeness of what is delivered (original source, abstract) Search provisions and degree of advanced features Frequency of the Update NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 4.Searchable News Data Bases – Definition – archive oriented (as opposed to headline) multiple news source aggregators – Best are “fee based” (MSU Library) • • • • • Dialog LexisNexis Dow Jones (Factiva) ProQuest Others NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • 4.Searchable News Data Bases (continued) – Web “free” sites – See Suggestions below NEWS SOURCE GENERATED INFORMATION Web News Outlets (continued) • • 5.Industry News Sites Definition – Industry specific news sites – – • Have combined features of several types above: – – – • Created by Trade Associations Trade Publications (their migration to the Web) News Alert Service Breaking News Aggregators If ‘News’ source based, may have archives Examples: www.foodinstitute.com. A Few Suggested News Sites Google News Archive Search news.google.com/archivesearch • Claims to go back 300 years • Time, WSJ, NYTimes, The Guardian, The Washington Post • Sources from ProQuest, Factiva, HighBeam, etc. • Some full articles are free, most are fee • Timelines • Advanced Search features A Few Suggested News Sites See: www.onstrat.com/news/newssearchchart.html For a comparison of: Yahoo, Google, daypop, rocketnews, findroy, feedster, topix A Few Suggested News Sites • www.monitor.bbc.uk/weekahead.shtml • www.wn.com • www.einnews.com – Subscription business information and online news service which draws from 35,000 sources – Covers 240 countries categorized by country and topic – Headlines Only!! Use to identify and than go to library sources A Few Suggested News Sites News Resource Guides: • www.kidon.com/media-link Provides info and link to sources and indicates the presence of streaming audio and video • www.abyznewslinks.com links to newspapers, broadcast stations, internet services, etc. • www.metagrid.com List of 8000 online magazines newspapers worldwide • www.newswealth.com Unique categories of miscellaneous ‘news’ sources A Few Suggested News Sites Front Pages: www.newseum.org/todaysfrontpages • 581 front pages from 54 countries • Alphabetical main page with “Sort by Region” geographic listing • Thumbnail view www.pressdisplay.com • Front Page Free – 7 day free trial • Full images of news pages of 500 newspapers from 70 countries for SUBSCRIPTION • Zoom and in paper search feature A Few Suggested News Sites Radio/TV Sources: www.radio-locator.com • Links to over 10,000 radio stations and over 2500 audio streams from radio stations in 130 countries www.tvradioworld.com • From over 200 countries Some Conclusions and Cautions • There is great redundancy so be very selective and methodical • One way is to “personalize” your news (Selfconfirmation bias) • The nature of news creation and distribution means that there will be more broken links • Spend time becoming an “Information Trapper!” Social Search What is social search? • No industry standard definition yet. • “Internet wayfinding tools informed by human judgment” • “Informed” can mean many thingsincluding egregiously uninformed. Social Search Algorithmic Search is “Social” • Algorithms are written by humans who make choices • Now. Search engines observe human behavior – click paths, popular, URL’s, etc which are used to modify the algorithm (Yahoo’s 14 tetragigs/day) • “Personalization efforts are becoming more evident. Social Search Why now? • Algorithmic search has plateaued • Humans are still better at some things • Rise in cocreation and collaboration via Web 2.0 • Recall status of wikipedia • Social Networking – 69% of females(56% males) ages 17-25 use Facebook – 38% females (14% males) ages 17-25 use MySpace – 70% ages 18-21 uses social networks Social Search Issues--• Scale and scope issues – How to keep up and what is the level of “control and policies”? • Tagging – How to you get to common understanding? – Folksonomy (also known as collaborative tagging , social classification, social indexing, social tagging, and other names) is the practice and method of collaboratively creating and managing tags to annotate and categorize content. – Ambiguity of language (‘orange’) – Others? • Social search will probably work best for non-text content (photos, music, video, widgets, etc.) Social Search Some Selected Types of Social Search • Shared bookmarks and Web Pages • Tag Engines (blogs and RSS) • Collaborative directories • Personalized vertical search engines Shared Bookmarks • The most basic and probably least useful type of social search • http://del.icio.us/ • http://www.shadows.com/ • http://myweb2.search.yahoo.com/ • http://www.furl.net/ • http://www.diigo.com/community Tag Engines Sometimes call “taggregators” primarily search blogs and RSS feeds • http://technorati.com/ - The #1 • http://www.ask.com/?tool=bls – Could be the best • http://www.blogpulse.com/ - Monitors and is a Nielsen firm Collaborative directories Directories created by teams of volunteers • Open Directory Project (AOL) – Has become dated and stale • http://www.prefound.com/ • http://www.stumbleupon.com/ - Appears quite good • http://www.mahalo.com/ - Mostly currently popular material • http://www.linkedin.com/ -Professional Networking Personalized Verticals It is no longer difficult or laborious to create a specialized search engine – • http://www.google.com/coop/cse/ • http://www.eurekster.com/ • http://rollyo.com/ Social Search Conclusions • Social Search will grow in importance • People are less predictable than algorithms – unlimited potential or problems? Basic Information Trapping • Information Trapping is the process of setting monitors – traps – to cature information from the flow of the Web and have it sent to you. • Termed coined by Tara Calishain • http://www.researchbuzz.com Basic Information Trapping Info Trapping Pros • Faster Results – As it happens, not weeks latter • More Results – Don’t have to remember to check • Saves You Time – Not constantly duplicating searches Info Trapping Cons • The sheer volume can overwhelm you. What Is Trappable? • • • • • • • News Stories Web Sites Conversations Multimedia Tag Directories Blogs Anything with an RSS Feed Basic Information Trapping This is where ‘the action is’ for: • Consumer research • Image management • Political planning and advertising • Social profiling • Etc. How Do You ‘Trap’? • • • • RSS Feed Readers Web Page Monitors E-Mail Alerts ‘Trapline’ Allocation – 70% RSS – 20% Web pages – 10% e-mail alerts Basic Information Trapping RSS Feeds • Definition - an XML-based specification that allows a Web site to instantly and automatically distribute its content (news and now more) to other sites • Accessing - requires specialized software be installed by the researcher Basic Information Trapping RSS Feeds (continued) • What is the value of RSS Feeds? Prequalification • By setting the profile, the user ‘edits’ what information comes into the attention space • However, the researcher still has an obligation to do the editing – Guard against self-conformation biases – Must have a ‘focused’ relevancy strategy Basic Information Trapping WEBLOGS a.k.a. BLOGS • Definition - a form of personal journalism where an individual purporting to have knowledge of and interest in a specific topic posts his/her views on the topic on the Web. • Typical Characteristics of Blogs include: – daily postings – recommended links – often have “chatrooms” for forums and discussions – popular blogs now generate advertising Basic Information Trapping E-mail Alerts are straight forward – • Most run on an RSS platform • Are now readily available Basic Information Trapping Info Trapping is a separate training session Some possible tools include: • http://www.aignes.com/ (WebSite Watcher) Free Trail than fee service • http://www.trackle.com/ Modest subscription fee • http://www.rocketnews.com/info/portal.jsp • http://www.boardtracker.com/ (Conversations) • http://boardreader.com/ (Conversations) • http://find.yuku.com/ Web 2.0 (Conversations) Basic Information Trapping Some possible tools (continued): • • • • http://www.everyzing.com/ (Multimedia/Podcasts) http://technorati.com/ (everything blogs) http://www.icerocket.com/ (blogs) http://www.zuula.com/ (Beta version of a Metasearch engine for Info Trapping) Basic Information Trapping • Requires a fair amount of work • Absolutely requires you have a very specific search query • Requires some advanced skills for managing the “Trapline” The Future (NOW) of Internet Search? • “Blended” or “Universal” Search are becoming the norm • “Personalization’ of Search because of algorithm interaction with “YOUR” actual search actions • “Mobilization” will take everything where you are • The battle between Web 1.0 vs. Web 2.0 philosophy