Extracting Consumer’s Needs through Web Mining

Data Mining the major technique to know the user’s needs
Extracting Consumer’s Needs through Web Mining
Source:  Unsplash

Introducing a Web Mining approach which will automatically identify new idea extracted from weblogs as known as a blog. A large number of blogs from every topic where users present their needs for new products. These ideas may be valuable for producers and developers because they can lead a different new product development process that is called marketing. This approach is implemented by a web-based application named Product Idea Web Log Miner where users from the marketing department provide descriptions of existing products.

Weblogs are web pages that consist of textual and nontextual information combined with links to further weblogs, web pages, etc. Most web pages offer a reader to leave a comment in an interactive format. To analyze this textual information, tools, and methods from text mining can be used.

Rational Approach

Searching in weblogs can be done by using a search engine and by limiting the query results to textual information from weblogs but a search query consists of several domain-specific terms. Every retrieved document from a query result consists of a title, description, and a link. A description contains search terms from the query in a link that leads directly to the full text of the retrieved weblog site.

Methodology

Product description -> Text preparation -> Creating text patterns -> Creating search queries -> Executing queries in web logs -> Filtering results -> New product Ideas.

This weblog mining approach has the aim to support users by finding new product ideas from weblogs.

Text Acquisition and Preparation

To this weblog mining approach, the user has to provide a product brief description. Sometimes, the description of future products that are in a product development process is not existent. To extract new ideas concerning their future products, the user first has to create these descriptions.



Text Pattern and Search Query Creating

Around each term in the provided product description, they build a text pattern if the selected term is not a stop word. By computing the length of text patterns by a provided length from the user term weighting schema. It differs between stop words and non-stop words because they are not equally important.

They build search queries from the created text pattern. As described in the rationale, a person uses several terms from the product description to build a search query.

Extracting Consumer’s Needs through Web Mining
This example shows how text patterns are extracted from a user-provided product description

Then, they build search queries that consist of four stemmed and stop word filtered terms, which occur together in a text pattern.

Search Query Executing and Result in Filtering

Web services help to execute created search queries. Web service is a software system designed to support interoperable machine to machine interaction over a network. It is just a web-based advanced Programming interface. Through the internet, it can be accessible.

Query results consist of a title, a brief description that contains terms from the search query.


Through these approaches, consumers write new product ideas in weblogs by use of Colloquial language. In contrast to the technical language, they see that
terms are not defined exactly and that many homonym and synonym problems occur by evaluating this textual information with text mining methods.

Leave a Reply

Your email address will not be published.