Techniques of Data Extraction

Extract data from a website probably the most common technique traditionally used copies you want (for example, the URL and link titles); the process is to come up with some regular expressions. In fact, this is the reason for our screen scraper software application written for the application started. Exactly Perl You are already familiar with regular expressions, and scrape the project is relatively small, at the same time; they can be a good solution.

Some of the programs to analyze the semantic content of an HTML page then drag that piece of intelligent interest. Still other approaches, or materials, which are intended to represent the domain names to go to the development of a hierarchical vocabularies.

Screen scraping specifically to the fact that a number of commercial applications (including your own) are. Applications vary widely, but in the medium-sized and large projects, they are often a good solution. Everyone has their own learning curve, a new application you are going to learn the ins and outs should take the time.

What is the best way to retrieve the data? It depends on what your needs are and what resources are available. There are a number of approaches, as well as suggestions on how you can use each one, there are some pros and cons:

RAW regular expressions and code


– If you are already familiar with regular expressions and at least one programming language, it can be a quick fix.

– Regular Expression content of such small changes that do not break the “obscurity” to provide a reasonable amount.

– Probably (a regular expression that you are already familiar with the program, starting again) does not need to learn new languages or tools.

– Regular expressions are supported in nearly all modern programming languages. Heck, even if the regular expression engine VBScript. Regular expression syntax is different in its implementation, as it is not too much different.


– They do not have a lot of experience of those who may be complex. Learning Perl regular expressions in Java is not the way. Pearl to see the problem in a very different way wrap in XSLT, the mind is like.

– They are often mistaken for analysis.

– The process of data discovery portion (if you want to get information from different web-crossing) is yet to be addressed, and if you want to handle cookies, and the like can be quite complex.

Artificial Intelligence


– You build it once and it more or less material, which can extract data from each page of a domain.

– Data models usually you can pick up information on the web all the cars extraction engine, model, and price are what you know is the built-in example, if the existing data structures they can map (for example) a database insert data in the right places.

– If the required relatively little maintenance in the long term.


– It is the kind of the engine is relatively complicated to build and operate.

– These types of engines are expensive to produce.

– Find the section is taken.

About Author

Leave A Reply