Member Avatar for tristan17

hi,

is there a way to find out what are the keyword tags to a web site that are used when we want to search the web for some information.

I mean, as we enter the words in the search bar, it will be match to the tags of the websites so as to return the relevant results.

Thanks in advance.

tris

  • 3 Contributors
  • 3 Replies
  • 114 Views
  • 3 Days Discussion Span
  • Latest Post Latest Post by zippee

Dani AI

Generated

For clarity: "keyword tags" can mean different things. If you mean the old HTML meta name="keywords" tag — that has been essentially ignored by major search engines for many years. If you mean the textual signals search engines use to match queries to pages (and images), those are broader: page title, headings, body text, anchor text, alt attributes, file names and paths, captions/figcaptions, structured data (schema.org/ImageObject), and embedded image metadata (EXIF/IPTC). Search engines and image search rely on a combination of these signals, not a single "keywords" field.

As hinted, crawling a site will let you harvest whatever metadata the site publishes. For CBIR work it's better to collect a defined set of contextual fields per image: alt, filename, immediate parent link text, nearest heading, nearby paragraph sentences, figcaption, page title and meta description, plus any EXIF/IPTC tags in the image file. Store those alongside the image feature vectors so experiments can compare textual vs. visual signal performance. As observed, SEO practices complicate the picture; many pages contain SEO-driven text that doesn't reflect image content, so treat such text as noisy labels.

Practical notes: test manually first (view source, inspect alt and captions). For scale use a polite crawler that respects robots.txt, rate limits, and site policies; for JavaScript-heavy sites render with a headless browser (Puppeteer/Selenium) to capture dynamically inserted captions or lazy-loaded images. To know what queries actually lead users to a verified site, use Google Search Console / Bing Webmaster Tools (site verification required). Third-party SEO tools can provide estimated ranking keywords for a domain, but they are approximations.

Cautions: obey robots.txt and copyright rules; do not assume meta keywords indicate relevance; and expect a lot of noisy or missing text in the wild. For robust CBIR, combine textual context with visual descriptors rather than relying on any single tag.

Recommended Answers

All 3 Replies

Member Avatar for tgreer

Not really. Reading between the lines, I suspect that you're trying to "optimize" your website so that it's listed "high" in a search engine. There's an entire industry that's grown up around this, called SEO. Daniweb has some SEO-related sections, you might post there.

Websites artificially manipulate search engines, search engines tune their algorithms, then we commercialize the whole thing with ads... what a mess. Sorry, I'm a cynic.

Member Avatar for tristan17

hi thomas,

thanks for the reply,well, i am not trying to optimize my website or watsoever, just that i am doing a research on CBIR (content based image retrieval) so i am also trying to understand the textual part of the search engine better.

anyway, hope to share more with you again!

tris

Member Avatar for zippee

You can use server side script like asp and php to crawl internet sites to a database. You can retrieve the meta tags including keywords and description during the crawling process. You can then create a search script to list down all sites (appeared in your database) that match the keyword your entered.

This is how a simple search engine technology is based on.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.