This page is intended to help you use the search functionality on the Guardian website. If it does not provide the information you were looking for, or you have any further questions about using the search engine on the Guardian website, please email userhelp@theguardian.com
Viewing your results | Restricting your results to very recent articles | Refine by date | Refine by section and publication | Refine by contributor | Refine by tone | Keyword suggestions | Searching for multimedia content | Searching a specific section | Using quotation marks to search for phrases | Searching for articles published before 1999 | Why can't I find an article that I know was published? | How search engines work
Viewing your results
When your search page loads, you will see up to five results sorted by chronological order, with the latest articles first, and up to five articles sorted by relevance. A search for 'titanic' sorted by date order may bring up as the top result an article talking about 'a titanic struggle'. Sorted by relevance however, your first results will be more likely to feature articles about the ship or the film. We think that by showing a selection of recent and relevant results, you should get some satisfactory results, regardless of whether your query is about something that is happening in the news right now, or you are looking for something from our archive. In order to see more results, you can either click 'Show more by date' or 'Show more by relevance'.
Restricting results to very recent articles
As part of the filtering mechanism displayed to the right-hand side of the results, the Guardian website search allows you to only view articles from today, yesterday, the previous 7 days and the previous 30 days. This is a very quick way of narrowing down your search results to the most recent content, and especially useful if you are looking for something that you know you have seen in the newspaper or on the website in the last couple of days.
Refine by date
The search filtering mechanism allows you to 'drill-down' to results from a specific year. On the right-hand side of the screen, next to your results, you will see a set of filters labelled 'Refine by date'. The number in brackets next to the year represents the number of results that matched your search from that year. Selecting a year by clicking it will reduce your results to showing only articles published in that year. If you are looking for content published prior to 1999, you will need to search using our Digital Archive instead.
Refine by section and publication
As well as being able to refine results according to when they were published, you are also able to restrict results according to where an article was published. Sections refer to areas of the website, e.g. Life & Style or UK News. Publications refer to printed newspaper, e.g. The Guardian, The Observer or the Society Guardian supplement. Content that has appeared on the web, but not in print, for example some blog posts, is labelled as being published by the Guardian website. To access the sections and publication filters, click 'More options' underneath the 'Refine by date' filters.
Refine by contributor
We know from research and from observing the search log files that people often search for articles by a particular writer or contributor. The contributor filter on the right-hand side can help you locate articles by a specific person. The names that initially appear will be those who have written the most articles matching your search query. To access the contributor filters, click 'More options' underneath the 'Refine by date' filters.
If you do follow a particular writer, you may find it useful to bookmark or note down their profile page. This always displays the latest content they have published on the site.
Refine by tone
All content on the Guardian website is tagged with a 'tone' which describes the type of content, for example a news articles, an obituary, a review or a match report. You can use the 'Refine by tone' controls to restrict your results to just one of these types. To access the tone filters, click 'More options' underneath the 'Refine by date' filters.
Keyword suggestions
All of the articles on the Guardian website are 'tagged' with keywords that describe the main topics of the article. These 'tags' each have their own page, aggregating all the stories that have been written on a topic. If your search query matches one of our keywords, then you'll see the 'tag' page listed near the top of your search results. If you are researching a particular topic, or looking for the latest articles and blog posts on a topic, the 'tag' page is a great starting point.
Searching for multimedia content
The search engine filters give you the opportunity to restrict your results to particular types of media published on the site. As well as articles, the search engine index includes Audio, Cartoons, Competitions, Interactives, Polls and Video, and a filter can be applied to only return results of that type. To narrow to one of these types of content, use the 'Refine by type' filter. To reveal this filter, click 'More options' underneath the 'Refine by date' filters.
Searching a specific section
On most pages on the Guardian website, search is accessed by using the input box in the top right-hand corner of the page. The box also includes a drop-down menu control, which tells the search engine what type of content to search across. If you browse to the 'Sport' section, for example, you will see that the selected default of this drop-down changes, and that carrying out a search from the 'Sport' front page will produce results already filtered to only contain articles from the 'Sport' section. You can change this by selecting 'guardian.co.uk' from the drop down when you enter your search.
Using quotation marks to search for phrases
Putting quotation marks around your search terms will force the search engine to look for the words appearing as a phrase. For example, a search for leeds united will match any page page that mentions Leeds and uses the word united. A search for "leeds united" will only return content that mentions the exact phrase "Leeds United". This will mostly be about football, but it would still match an article that said "This saw the whole city of Leeds united".
Searching for articles published before 1999
The Guardian website currently contains articles going back to 1999, and some selected highlights from the newspaper archive prior to that date. If you are looking for material older than 1999, you may have more success by searching through the digital archive of Guardian and Observer content. The archive covers the period of 1821-2000 for the Guardian, and 1791-2000 for the Observer.
Why can't I find an article that I know was published?
Sometimes an article is missing from our search index. This could be for the following reasons:
The article came from a third-party feed: the Guardian website takes some content from third parties in the form of wire stories. These third parties include Reuters, the Press Association and the Associated Press. We don't include these articles in our search index.
The article appeared in the newspaper, but we don't have the right to reproduce the article online: Sometimes, articles which appeared in our newspapers don't appear on the website because we don't have the right to reproduce them. Examples include extracts from published works, such as book extracts, or articles from other publications. These articles won't appear in the search results.
The article appeared in the newspaper, but its format meant that it was not reproducible online: Some articles don't appear online, because their format is not suitable for the website. These include some articles which are primarily tables (e.g. shopping comparison guides), or composites of images (e.g. collages). A digital facsimile of recent editions of the daily newspaper is available as a paid-for service at http://digital.guardian.co.uk/.
The article has been deleted from our database: Very rarely, we do delete articles from our database because of legal reasons. If you know an article did appear on the Guardian website and it wasn't a wire story, and it doesn't appear in the search, this is the most likely explanation. It's normally impossible for us to go into the reasons for deleting an individual story, for the same legal reasons which led to its deletion.
How search engines work
Search engine work on a relatively simple principle, but using some very complex mathematics to derive their results. In essence, all of the words on the Guardian website are held in an index, and the words that the user types in as their search query are matched against that index. The search engine 'recalls' all of the pages that match all or some of the words, excluding very common 'stop words' like 'to', 'the', 'and' etc. If you have chosen to sort by date, these articles are 'recalled' in reverse chronological order, with the newest item first. If you have chosen to rank by relevance then the relative positions of a page in the search engine results list can be determined by factors such as how many times the query words appear on the page, whether they are grouped in a phrase, or whether they appeared in the headline. Their relative scarcity in the collection can also be a factor. For example, if a user searched for 'white peacocks', it is likely that in the Guardian website database the word 'white' has been used a lot more than the word 'peacocks'. Therefore the search engine will favour returning pages that mention the rarer word more frequently. Many of the founding principles of modern search engine technology were developed in the UK in the 1960s and 1970s, with Professor Karen Spärck Jones introducing the key information retrieval concept of inverse document frequency. Search Engine History has a timeline of how the technology has evolved from 1945 to the present day, starting with Vannevar Bush's concept of the Memex.
No comments:
Post a Comment