To suggest additional items which may be of interest to the searcher based on the items already retrieved, Data Harmony® has built the Recommender. Once the user has selected items of interest, the Recommender comes into play.
There are many theories about how to create a recommendation engine. Most depend on the user profile; however, Data Harmony already tags information objects with terms from the taxonomy. Therefore, the easiest way to make a selection for recommendation is to choose another item, one with exactly the same index terms as that which the user has selected. If the data has been tagged with taxonomy terms, the more closely the indexing matches, thus the more likely the recommended articles will satisfy the user.
Say, for example, articles receive 5 index terms each from a collection’s attached taxonomy. Should a search select one of these articles, another article with the same five index terms will most likely also answer the needs of the search. When a 100% match between articles does not exist in the collection, articles containing the next highest level of matches are chosen. Applying relevance ranking in a search renders results with more absolute confidence than computer system guesses based on recent search profiles.
As with other elements of the Data Harmony suite, Recommender is repository and compatible with multiple search software formats. Recommender is dependent on access to the taxonomy terms applied to individual documents within a collection. Using not co-occurrence, but rather keywords associated with the individual document, Recommender is less system-intensive and a more precise way to produce relevant results to the searcher.
To see the Data Harmony Recommender in action, try the link below. Enter a search term into the box, then click on one of the results to see others with the same indexing hits.
For e-commence implementations, suggestions are added to the selection page.