Webscraping using readLines and RCurl

There is a massive amount of data available on the web. Some of it is in the form of precompiled, downloadable datasets which are easy to access. But the majority of online data exists as web content such as blogs, news stories and cooking recipes. With precompiled files, accessing the data is fairly straightforward; just download the file, unzip if necessary, and import into R. For “wild” data however, getting the data into an analyzeable format is more difficult. Accessing online data of this sort is sometimes reffered to as “webscraping”. Two R facilities, readLines() from the base package and getURL() from the RCurl package make this task possible.

readLines

For basic webscraping tasks the readLines() function will usually suffice. readLines() allows simple access to webpage source data on non-secure servers. In its simplest form, readLines() takes a single argument – the URL of the web page to be read:

web_page <- readLines("http://www.interestingwebsite.com")

As an example of a (somewhat) practical use of webscraping, imagine a scenario in which we wanted to know the 10 most frequent posters to the R-help listserve for January 2009. Because the listserve is on a secure site (e.g. it has https:// rather than http:// in the URL) we can't easily access the live version with readLines(). So for this example, I've posted a local copy of the list archives on the this site.

One note, by itself readLines() can only acquire the data. You'll need to use grep(), gsub() or equivalents to parse the data and keep what you need.

# Get the page's source
web_page <- readLines("http://www.programmingr.com/jan09rlist.html")
# Pull out the appropriate line
author_lines <- web_page[grep("<I>", web_page)]
# Delete unwanted characters in the lines we pulled out
authors <- gsub("<I>", "", author_lines, fixed = TRUE)
# Present only the ten most frequent posters
author_counts <- sort(table(authors), decreasing = TRUE)
author_counts[1:10]
[webscrape results]


We can see that Gabor Grothendieck was the most frequent poster to R-help in January 2009.

The RCurl package

To get more advanced http features such as POST capabilities and https access, you'll need to use the RCurl package. To do webscraping tasks with the RCurl package use the getURL() function. After the data has been acquired via getURL(), it needs to be restructured and parsed. The htmlTreeParse() function from the XML package is tailored for just this task. Using getURL() we can access a secure site so we can use the live site as an example this time.

# Install the RCurl package if necessary
install.packages("RCurl", dependencies = TRUE)
library("RCurl")
# Install the XML package if necessary
install.packages("XML", dependencies = TRUE)
library("XML")
# Get first quarter archives
jan09 <- getURL("https://stat.ethz.ch/pipermail/r-help/2009-January/date.html", ssl.verifypeer = FALSE)
jan09_parsed <- htmlTreeParse(jan09)
# Continue on similar to above
...

For basic webscraping tasks readLines() will be enough and avoids over complicating the task. For more difficult procedures or for tasks requiring other http features getURL() or other functions from the RCurl package may be required. For more information on cURL visit the project page here.

Source: http://www.r-bloggers.com/webscraping-using-readlines-and-rcurl-2/

Screen scrapers: To program or to purchase?

Companies today use screen scraping tools for a variety of purposes, including collecting competitive information, capturing product specs, moving data between legacy and new systems, and keeping inventory or price lists accurate.

Because of their popularity and reputation as being extremely efficient tools for quickly gathering applicable display data, screen scraping tools or browser add-ons are a dime a dozen: some free, some low cost, and some part of a larger solution. Alternatively, you can build your own if you are (or know) a programming whiz. Each tool has its potential pros and cons, however, to keep in mind as you determine which type of tool would best fit your business need.

Program-your-own screen scraper

Pros:

    Using in-house resources doesn't require additional budget

Cons:

    Properly creating scripts to automate screen scraping can take a significant amount of time initially, and continues to take time in order to maintain the process. If, for instance, objects from which you're gathering data move on a web page, the entire process will either need to be re-automated, or someone with programming acumen will have to edit the script every time there is a change.

    It's questionable whether or not this method actually saves time and resources

Free or cheap scrapers

Pros:

    Here again, budget doesn't ever enter the picture, and you can drive the process yourself.

    Some tools take care of at least some of the programming heavy lifting required to screen scrape effectively

Cons:

    Many inexpensive screen scrapers require that you get up to speed on their programming language—a time-consuming process that negates the idea of efficiency that prompted the purchase.

Screen scraping as part of a full automation solution

Pros:

    In the amount of time it takes to perform one data extraction task, you have a completely composed script that the system writes for you

    It's the easiest to use out of all of the options

    Screen scraping is only part of the package; you can leverage automation software to automate nearly any task or process including tasks in Windows, Excel automation, IT processes like uploads, backups, and integrations, and business processes like invoice processing.

    You're likely to get buy-in for other automation projects (and visibility for the efficiency you're introducing to the organization) if you pick a solution with a clear and scalable business purpose, not simply a tool to accomplish a single task.

Cons:

    This option has the highest price tag because of its comprehensive capabilities.

Looking for more information?

Here are some options to dig deeper into screen scraping, and deciding on the right tool for you:

 Watch a couple demos of what screen scraping looks like with an automation solution driving the process.

 Read our web data extraction guide for a complete overview.

 Try screen scraping today by downloading a free trial.

Source: https://www.automationanywhere.com/screen-scrapers

Data Mining Outsourcing in a Better and Unique Approach

Data mining outsourcing services are ideal for clarity in various decision making processes.  It is the ultimate goal of any organization and business to increase on its profits as well as strengthen the bond with its customers. Equipping the business in such a way that it’s very easy to detect frauds and manage risks in a convenient manner is equally important. Volumes of data that are irrelevant or cannot be used when raw needs to be converted to a more useful form.  The data mining outsourcing services can greatly help you to analyze and interpret data in a more diligent way.

This service to reliable, experienced and qualified hands is very important. Your research project or engineering project can be easily and conveniently handled by experienced staff who guarantees you an accuracy level of about 98% and a massive reduction in operating costs. The quality of work is unsurpassed and the presentation is done in a format that is easy and simple for you. The project is done in a very short time alleviating you delays as well as ensuring on-time completion of your projects. To enjoy a successful outsourcing experience, you need to bank on a famous and reliable expertise.

The only time to rely with data mining outsourcing services is when you do not have a reliable, experienced expertise in your business.  Statistics indicate that it’s very easy to lose business intelligence or expose the privacy of the customers through this process. However companies which offer secure outsourcing process are on the increase as a result of massive competition. It’s an opportunity to develop your potential of sourced data and improve your business in all fields. 

Data mining potential applications are infinite. However major applications are in the marketing research and scientific projects. It’s done both on large and small quantities of data by experienced staff well known for their best analytical procedures to guarantee you accurate and easy to use information. Data mining outsourcing services are the only perfect way to profitability.

Source:http://www.e-edge.biz/Data_Mining_Outsourcing_in_a_Better_and_Unique_Approach.html

NHL ending dry scraping of ice before overtime

TORONTO (AP) — The NHL will no longer dry scrape the ice before overtime.

Instituted this season in an effort to reduce the number of shootouts, the dry scraping will stop after Friday's games.

The general managers decided at their meeting Tuesday to make the change after the league talked to the players' union the past few days.

Beginning Saturday, ice crews around the league will again shovel the ice after regulation as they did in previous years. The GMs said the dry scrape was causing too much of a delay. Director of hockey operations Colin Campbell said the delays were lasting from more than four minutes to almost seven.

The dry scrape initially had been approved in hopes of reducing shootouts by improving scoring chances without unduly slowing play by recoating the ice.

The GMs also discussed expanded video review, including goaltender interference, and the possibility of three-on-three overtime. The American Hockey League is experimenting with the three-on-three format this season.

This annual meeting the day after the Hockey Hall of Fame induction usually doesn't produce actual changes, with the dry scrape providing an exception.

The main purpose is to set up the March meeting in Boca Raton, Florida, where these items will be further addressed.

Source:http://missoulian.com/sports/hockey/nhl-ending-dry-scraping-of-ice-before-overtime/article_3dd5473c-6102-5800-99f7-2c98be0f99ad.html

Scraping websites using the Scraper extension for Chrome

If you are using Google Chrome there is a browser extension for scraping web pages. It’s called “Scraper” and it is easy to use. It will help you scrape a website’s content and upload the results to google docs.

Walkthrough: Scraping a website with the Scraper extension

•     Open Google Chrome and click on Chrome Web Store
•     Search for “Scraper” in extensions
•     The first search result is the “Scraper” extension
•     Click the add to chrome button.
•     Now let’s go back to the listing of UK MPs
•     Open http://www.parliament.uk/mps-lords-and-offices/mps/
•     Now mark the entry for one MP
•     http://farm9.staticflickr.com/8490/8264509932_6cc8802992_o_d.png
•     Right click and select “scrape similar…”
•     http://farm9.staticflickr.com/8200/8264509972_f3a9e5d8e8_o_d.png
•     A new window will appear – the scraper console
•     http://farm9.staticflickr.com/8073/8263440961_9b94e63d56_b_d.jpg
•     In the scraper console you will see the scraped content
•     Click on “Save to Google Docs…” to save the scraped content as a Google Spreadsheet.

Walkthrough: extended scraping with the Scraper extension

Note: Before beginning this recipe – you may find it useful to understand a bit about HTML. Read our HTML primer.

Easy wasn’t it? Now let’s do something a little more complicated. Let’s say we’re interested in the roles a specific actress played. The source for all kinds of data on this is the IMDB (You can also search on sites like DBpedia or Freebase for this kinds of information; however, we’ll stick to IMDB to show the principle)

    Let’s say we’re interested in creating a timeline with all the movies the Italian actress Asia Argento ever starred; where do we start?

    The IMDB has a quite comprehensive archive of actors. Asia Argento’s site is: http://www.imdb.com/name/nm0000782/

    If you open the page you’ll see all the roles she ever played, together with a title and the year – let’s scrape this information

    Try to scrape it like we did above

    You’ll see the list comes out garbled – this is because the list here is structured quite differently.

    Go to the scraper console. Notice the small box on the upper left, saying XPath?

    XPath is a query language for HTML and XML.

    XPath can help you find the elements in the page you’re interested in – all you need to do is find the right element and then write the xpath for it.

    Now let’s assemble our table.

    You’ll see that our current Xpath – the one including the whole information is “//div[3]/div[3]/div[2]/div”

    http://farm9.staticflickr.com/8344/8264510130_ae31697fde_o_d.png

    Xpath is very simple it tells the computer to look at the HTML document and select <div> element number 3, then in this the third one, the second one and then all <div> elements (which if you count down our list, results in exactly where you are right now.

•     However, we’d like to have the data separated out.
•     To do this use the columns part of the scraper console…
•     Let’s find our title first – look at the title using Inspect Element
•     http://farm9.staticflickr.com/8355/8263441157_b4672d01b2_o_d.png
•     See how the title is within a <b> tag? Let’s add the tag to our xpath.
•     The expression seems to work well: let’s make this our first column
•     In the “Columns” section, change the name of the first column to “title”
•     Now let’s add the XPATH for the title to it
•     The xpaths in the columns section are relative, that means “./b” will select the <b> element
•     add “./b” to the xpath for the title column and click “scrape”
•     http://farm9.staticflickr.com/8357/8263441315_42d6a8745d_o_d.png
•     See how you only get titles?
•     Now let’s continue for year? Years are within one <span>
•     Create a new column by clicking on the small plus next to your “title” column
•     Now create the “year” column with xpath “./span”
•     http://farm9.staticflickr.com/8347/8263441355_89f4315a78_o_d.png
•     Click on scrape and see how the year is added
•     See how easily we got information out of a less structured webpage?

Source: http://schoolofdata.org/handbook/recipes/scraper-extension-for-chrome/

Building Java Object Graph with Tour de France results – using screen scraping, java.util.Parser and assorted facilities

Last Saturday, the Tour de France 2011 departed. For people like myself, enjoying sports and working on Data Visualizations on the one hand and far fetched uses of SQL on the other, the Tour de France offers a wealth of data to work with: rankings for each stage in various categories, nationalities and teams to group by, distances and velocity, years to compare with one another and the like. So it has been my intention for some time to get hold of that data in a format I could work with.

Today I finally found some time to get it done. To locate the statistics for the Tour de France editions for the last few years and get them onto my laptop and into my database. This article describes the first part of that journey: how to get the stage results from some source on the internet into my locally running Java program in an appropriate object structure.

My starting point is the official Tour de France website:

Image

This website goes back to 2007 and also has the latest (2011) results. It presents the result in a format pleasing to the human eye – based on an HTML structure that is fairly pleasing to my groping Java code as well.

Analyzing the source of the Tour de France data

I start my explorations in Firefox, using the Firebug plugin. When I select the tab with the results for a particular stage, I inspect the (AJAX) call that is made to retrieve the stage results into the browser:

Image

The URL that was accessed is www.letour.fr/2010/TDF/LIVE/us/700/classement/ITE.html . When I access that URL directly, I see an HTML fragment with the individual ranking for the 7th stage in 2010. It turns out that with ITG instead of ITE in this URL, I get the overall ranking after the 7th Stage. Using IME in stead of ITE, I get the 7th stage’s climbers’ standing. And so on.

The HTML associated with the stage standing looks like this:

Image

Which is not as user friendly as the corresponding display in the browser:

Image

but still fairly well structured and programmatically interpretable.

Retrieving HTML fragments and parsing in Java

Consuming these HTML fragments with stage standings into my own Java code is very easy. Parsing the data and turning it into sensible Java Objects is slightly more work, but still quite feasible. From the Java Objects I next need to create a persistent storage for the data – that is the subject for another article.

Using the Java URL class and its openStream method to open an InputStream on whatever content can be found at the URL, it is dead easy to start reading the HTML from the Tour de France website into my Java program. I make use of the java.util.Scanner class to work my way through the HTML by Table Row (TR element). When you inspect the HTML fragments, it is clear early on that every individual rider’s entry corresponds with a TR element, so it seems only logical to have the Scanner break up the data by TR.

private static Stage processStage(int year, int stageSequence, Map<Integer, Rider> riders) throws java.io.IOException, java.net.MalformedURLException {

    String typeOfStanding = "ITE";
     URL stageStanding = new URL("http://www.letour.fr/"+year+"/TDF/LIVE/us/"
                                +(stageSequence==0?"0":stageSequence+"00") +
                                "/classement/"+typeOfStanding+".html");
    InputStream stream = stageStanding.openStream();
    Scanner scanner = new Scanner(stream);
    scanner.useDelimiter("</tr>");
    Stage stage = new Stage();
    stage.setSequence(stageSequence);
    boolean first = true;
    boolean firstStanding = true;
    while (scanner.hasNext()) {
        String entry = scanner.next();
        if (first) {
            first = false;
            Matcher regexMatcher = regexDistance.matcher(entry);
            if (regexMatcher.find()) {
                String distanceString = regexMatcher.group();
                stage.setTotalDistance(Float.parseFloat(distanceString.substring(0, distanceString.length() - 3)));
            }
        }
        if (!first) {
            String[] els = entry.split("/td>");
            if (els.length > 1) { // only the standing-entries have more than one td element
                Integer riderNumber = Integer.parseInt(extractValue(els[2]));

                Rider rider=null;
                if (riders.containsKey(riderNumber)) {
                    rider = riders.get(riderNumber);
                }
                else {
                    rider = new Rider(extractValue(els[1]),riderNumber, extractValue(els[3]));
                    riders.put(riderNumber,rider);
                }
                Standing standing =
                    new Standing(firstStanding ? 1 : (Integer.parseInt(extractValue(els[0]).replace(".", ""))),
                                  rider,extractValue(els[4]),
                                  extractValue(els[5]));
                firstStanding = false;
                stage.getStandings().add(standing);                }
        }
    } //while
    scanner.close();
    return stage;
}

Subsequently, the TR elements need to be broken up in the TD cell elements that contain the rank, rider’s name, their number, the team they ride for and the time for the stage as well as their lag with regard to the winner. I have used a simple split (on /td>) to extract the cells. The final logic for pulling the correct value from the cell is in the method extractValue. Note: this code is not very pretty, and I am not necessarily overly proud of it. On the other hand: it is one-time-use-only code and it is still fairly compact and easy to write and read.

private static String extractValue(String el) {
    String r = el.split("</")[0];
    if (r.lastIndexOf(">") > 0) {
        r = r.substring(r.lastIndexOf(">") + 1);
    }
    return r.split("<")[0];
}

I have created a few domain classes: Rider, Stage, Standing (as well as Tour) that are a business domain like representation of the Tour de France result data. Objects based on these classes are instantiated in the processStage method that is being invoked from the processTour method.

public static void processTour(Tour tour) throws IOException, MalformedURLException {
    if (tour.isPrologue())
      tour.getStages().add(processStage(tour.getYear(),0, tour.getRiders()));

    for (int i=1;i<= tour.getNumberOfStages();i++)  {
        tour.getStages().add(processStage(tour.getYear(),i, tour.getRiders()));
    }
}

When I run the TourManager class – a class that create a single Tour object for the Tour de France in 2010 –

public class TourManager {
     List<Tour> tours = new ArrayList<Tour>();
     public TourManager() {
        tours.add(new Tour(2010, 20, true));
        try {
            ProcessTourStandings.processTour(tours.get(0));
        } catch (MalformedURLException e) {
            System.out.println(e.getMessage());
        } catch (IOException e) {
            e.printStackTrace();
        }
    }
     public static void main(String[] args) {
        TourManager tm = new TourManager();
        for (Tour tour : tm.getTours()) {
            for (Stage stage : tour.getStages()) {
                System.out.println("================ Stage " + stage.getSequence() + "(" + stage.getTotalDistance() +
                                   " km)");
                for (Standing standing : stage.getStandings()) {
                    if (standing.getRank() < 4) {
                        System.out.println(standing.getRank() + "." + standing.getRider().getName());
                    }
                }
            }
        }
    }

it will print the top 3 in every stage:

Image

Source:http://technology.amis.nl/2011/07/04/building-java-object-graph-with-tour-de-france-results-using-screen-scraping-java-util-parser-and-assorted-facilities/

Scraping Data: Site-specific Extractors vs. Generic Extractors

Scraping is becoming a rather mundane job with every other organization getting its feet wet with it for their own data gathering needs. There have been enough number of crawlers built – some open-sourced and others internal to organizations for in-house utilities. Although crawling might seem like a simple technique at the onset, doing this at a large-scale is the real deal. You need to have a distributed stack set up to take care of handling huge volumes of data, to provide data in a low-latency model and also to deal with fail-overs. This still is achievable after crossing the initial tech barrier and via continuous optimizations. (P.S. Not under-estimating this part because it still needs a team of Engineers monitoring the stats and scratching their heads at times).

Social Media Scraping

Focused crawls on a predefined list of sites

However, you bump into a completely new land if your goal is to generate clean and usable data sets from these crawls i.e. “extract” data in a format that your DB can process and aid in generating insights. There are 2 ways of tackling this:

a. site-specific extractors which give desired results

b. generic extractors that result in few surprises

Assuming you still do focused crawls on a predefined list of sites, let’s go over specific scenarios when you have to pick between the two-

1. Mass-scale crawls; high-level meta data - Use generic extractors when you have a large-scale crawling requirement on a continuous basis. Large-scale would mean having to crawl sites in the range of hundreds of thousands. Since the web is a jungle and no two sites share the same template, it would be impossible to write an extractor for each. However, you have to settle in with just the document-level information from such crawls like the URL, meta keywords, blog or news titles, author, date and article content which is still enough information to be happy with if your requirement is analyzing sentiment of the data.

cb1c0_one-size

A generic extractor case

Generic extractors don’t yield accurate results and often mess up the datasets deeming it unusable. Reason being

programatically distinguishing relevant data from irrelevant datasets is a challenge. For example, how would the extractor know to skip pages that have a list of blogs and only extract the ones with the complete article. Or delineating article content from the title on a blog page is not easy either.

To summarize, below is what to expect of a generic extractor.

Pros-

minimal manual intervention

low on effort and time

can work on any scale

Cons-

Data quality compromised

inaccurate and incomplete datasets

lesser details suited only for high-level analyses

Suited for gathering- blogs, forums, news

Uses- Sentiment Analysis, Brand Monitoring, Competitor Analysis, Social Media Monitoring.

2. Low/Mid scale crawls; detailed datasets - If precise extraction is the mandate, there’s no going away from site-specific extractors. But realistically this is do-able only if your scope of work is limited i.e. few hundred sites or less. Using site-specific extractors, you could extract as many number of fields from any nook or corner of the web pages. Most of the times, most pages on a website share similar templates. If not, they can still be accommodated for using site-specific extractors.

cutlery

Designing extractor for each website

Pros-

High data quality

Better data coverage on the site

Cons-

High on effort and time

Site structures keep changing from time to time and maintaining these requires a lot of monitoring and manual intervention

Only for limited scale

Suited for gathering - any data from any domain on any site be it product specifications and price details, reviews, blogs, forums, directories, ticket inventories, etc.

Uses- Data Analytics for E-commerce, Business Intelligence, Market Research, Sentiment Analysis

Conclusion

Quite obviously you need both such extractors handy to take care of various use cases. The only way generic extractors can work for detailed datasets is if everyone employs standard data formats on the web (Read our post on standard data formats here). However, given the internet penetration to the masses and the variety of things folks like to do on the web, this is being overly futuristic.

So while site-specific extractors are going to be around for quite some time, the challenge now is to tweak the generic ones to work better. At PromptCloud, we have added ML components to make them smarter and they have been working well for us so far.

What have your challenges been? Do drop in your comments.

Source: https://www.promptcloud.com/blog/scraping-data-site-specific-extractors-vs-generic-extractors/