INTERACTIVE SEARCHING SYSTEM AND METHOD
Field of the Invention The invention relates to a method for efficiently searching within a database. Particularly, the invention is directed to uses related to large databases, such as those used to search over the Internet.
Background of the Invention Searching items over the Internet may be a difficult and frustrating process, because of the many different names by which an item can be found, and the relatively inefficient restriction methods employed by search engines to limit the number of results of a given search. In addition, a specific text can be found in a large number of irrelevant results. This is particularly true in the field of e-commerce. A specific product that is sold on the Internet (or outside it) may be known by many names. There is no unique name that relates a given product to its name. A search that is based only on the text of the product name, or even on the text of its specifications/categories, does not solve the above mentioned problems. On the one hand, the user may receive a large number of results in which the name of the product o its description contains a series of letters that fit the criteria used by the user, though these results are totally irrelevant. (For example - the user searches for a television, and obtains results dealing with a television table). On the other hand, the search results do not contain all the products that are relevant for the
user, but which are not identified by the exact text that the user has entered as a search criterion.
The art has so far failed to provide an efficient solution to the abovementioned problems, although many attempts have been made, and several different search engines exist.
It is therefore an object of this invention to provide a method and a system which overcome the aforementioned problems of the prior art.
It is another purpose of this invention to provide a simple interactive method by means of which the items actually desired by the user may be speedily located in a database that contains them, even if the keywords used as search criteria are not used to identify the item ion the database. The term "item", as used herein, is meant to indicate any searchable matter, including words, text paragraphs, documents, etc., and is not limited to any particular, physical or otherwise, type of item.
It is yet another purpose of this invention to provide a method which, at the same time, amplifies and reduces the results list, to enable the user to locate the desired item, if it exists in the searched database, even if it is not directly retrievable by the search criterion used by the user.
Other purposes and advantages of this invention will appear as the description proceeds.
Summary of the Invention
The invention is directed to an interactive method for searching a database comprising the steps of:
- receiving a search criterion from a user, or if no search criterion is provided, querying the user so as to generate a results space as hereinafter defined;
- if a search criterion is provided, analyzing said search criterion to determine additional search criteria related thereto or stemming therefrom; - generating a results space comprising results selected from an existing database, which results are retrieved using the user's search criterion and the additional search criteria;
- restricting the number of results in the results space by: i) asking the user questions having two or more possible answers, each possible answer dividing the results space into different segments; and ii) removing from the results space results which fall outside the segment of the results space that matches the user's answer.
In this Step i) above, the user can also insert additional text that will also be treated as information usable to divide the results space.
According to a preferred embodiment of the invention, the method further comprises repeating the step of restricting the number of results in the results space, until a predetermined number of total results is reached.
According to another preferred embodiment of the invention, the method further comprises repeating the step of restricting the number of results in the results space, until the user decides that the total number of results obtained is that from which he wishes to choose.
Of course, the user is not obliged to provide a positive answer to a given question, and in the case that he chooses not to answer a question, a further question is asked without previously segmenting the results space.
All the above characteristics and advantages of the invention will be better understood through the following illustrative and non-limitative
description of preferred embodiments, with reference to the appended
drawings.
Brief Description of the Drawings
- Figs. 1 through 5 schematically illustrate an example of a search,
according to a preferred embodiment of the invention; and
Fig. 6 (A and B) is a simplified, schematic flow-chart illustrating
the interaction of the user with the system.
Detailed Description of Preferred Embodiments
Looking at the figures, the following steps illustrate the invention:
The user inserts a search criterion. In the example of Fig. 1 the search
criterion (key word) is "shoe". Of course, this key-word may be related to
many different products that the user is searching.
2. The system searches in the database for matching products/results,
which contains the text that the user had inserted. The result of this
search will be referred to as "results space". The searched database might
be, for example, the catalog of one or more stores.
3. The system expands the results space by adding other results (from
the database), that although not containing the text the user had
inserted, contain a related concept. This related concept may be a
synonym, a similar context, etc. An algorithm that uses thesauruses,
adaptive historical knowledge, self-conclusion engine etc, performs this
expansion. Providing such algorithms, in order to expand the results
space, if of course within the scope of the skilled person, and therefore is not described herein in detail, for the sake of brevity.
4. The results space, generated in Steps 2 and 3 above, is shown in
Fig. 2.
5. The system analyzes this results space . (online, or beforehand),
using methods such as NLP, neural networks, pattern recognition, classification etc., which are well known to the skilled person. The result
of this analysis is a segmenting question, which divide the results space in the most efficient way. For example, in a multiple choice question, with X
possible answers, each answer divides the results space to as close as
possible to 1/x.
6. An example of question, with 2 possible answers (i.e., "Is it bigger
than a basketball?"), is shown in Fig. 2.
7. After the user answers the question, the results space is reduced, and he all the matching results/products are displayed to him (as shown
in Fig. 2). The user may choose not to answer a specific question (for
example, when he does not know the answer). In this case, he clicks the
button that says "I don't know", and the system chooses another question.
Alternativelt- it is also possible to present the user with several questions, and to ask him to choose which question he would like to answer.
8. Stages 5-7 are repeated (stage 5 being performed each time on the
new space result), until the user asks to stop, or until the system cannot
generate more questions, or until there is a sufficiently small number of
results. The whole process is illustrated in Figs. 2-5.
9. The user may choose a product, or may argue that the product that
he wants, does not appear on the list. In this case, the system will present
him other products that do not match all the answers of the user, but
match some of them. If he chooses a product from the new list, the system
will inform him about any contradiction between his answers and the knowledge of the system.
10. The system also acts as a learning system, and adapts its own
information used to carry out the process, according to the result (i.e.
product) that the user has chosen. This is done, for example, by correcting
the correlations between the chosen product and the questions that the
user has answered.
A process of the type described above, according to a preferred
embodiment of the invention, is schematically illustrated in Fig. 6 (A and
B), and is self-explanatory. In the figure "DB" means "database".
As will be apparent to the skilled person, the invention provides two novel
and substantial improvements over the prior art:
1. The processes of expansion and reduction of the database, that enable
the user to locate exactly what he wants, assuming it exists on the
database, even if he doesn't know its exact name; and
2. The interaction with the user, in which the "user is led to the product in
which he is interested, in the most efficient way.
While embodiments of the invention have been described by way of illustration, it will be understood that the invention can be carried out by
persons skilled in the art with many modifications, variations and adaptations, without departing from its spirit or exceeding the scope of the claims.