Method and system for providing alternative result for an online search previously with no result
A computing device hosting a website of a business may be operable to receive a first search input comprising a term submitted via a search bar on the website. Upon performing an internal search for the term with no result, the computing device may output information on the no-result. The computing device may then perform, using one or more external search engines, a search for the term. Search results of the search, performed using the external search engine(s), may be analyzed. Based on a result of the analysis, one or more particular character strings related to the term may be identified. The computing device may generate and store, based on the identifying of the particular character string(s), one or more alternative search suggestions. Upon subsequently receiving an input comprising at least a portion of the term entered in the search bar, the computing device may output the alternative search suggestion(s).
1 . A method comprising:
in a computing device hosting a website:
performing an internal search, on the website, according to a term;
modifying the term utilizing natural language processing (NLP), if the internal search receives no result, wherein the NLP modification comprises:
parsing the term into constituent tokens,
classifying at least one token according to a semantic type selected from: brand name, model number, or part number, and
normalizing the classified token to a standardized representation according to a predetermined ruleset for model/part number formatting;
performing an external search, on one or more different websites, according to the modified term; and
associating an item with the term according to a result of the external search.
2 . The method according to claim 1 , wherein the method comprises outputting one or more alternative search suggestions.
3 . The method according to claim 1 , wherein the term is associated with a string of alphanumeric characters.
4 . The method according to claim 3 , wherein the method comprises determining that the term is a valid term for search if the string of alphanumeric characters comprises at least one digit.
5 . The method according to claim 4 , wherein the method comprises normalizing the determined valid term in compliance with a model number or a part number for searching for a model or a part of a product.
6 . The method according to claim 1 , wherein the method comprises modifying the term to comprise a model number or a part number.
7 . The method according to claim 1 , wherein the method comprises:
assigning a score to each of a plurality of character strings identified while analyzing search results; and
identifying the one or more particular character strings based on the assigned scores.
8 . The method according to claim 1 , wherein the method comprises: generating one or more alternative search suggestions, and
selecting each of the one or more alternative search suggestions according to whether a characteristic of each of the one or more alternative search suggestions matches a predetermined characteristic condition.
9 . The method according to claim 8 , wherein the predetermined characteristic condition comprises:
the number of a particular character strings in each of the one or more alternative search suggestions being greater than a certain threshold; and
each of the one or more alternative search suggestions comprising a brand name.
10 . The method according to claim 1 , comprising storing the one or more alternative search suggestions in a form of autofill index.
11 . A system comprising:
one or more processors operable to host a search engine, wherein the one or more processors are configured to:
perform an internal search, on a website, according to a term;
modify the term utilizing natural language processing (NLP), if the internal search receives no result, wherein the NLP modification comprises:
parsing the term into constituent tokens,
classifying at least one token according to a semantic type selected from: brand name, model number, or part number, and
normalizing the classified token to a standardized representation according to a predetermined ruleset for model/part number formatting;
perform an external search, on one or more different websites, according to the modified term; and
associate an item with the term according to a result of the external search.
12 . The system according to claim 11 , wherein the one or more processors are configured to output one or more alternative search suggestions.
13 . The system according to claim 11 , wherein the term comprises a string of alphanumeric characters.
14 . The system according to claim 13 , wherein the one or more processors are configured to determine that the term is a valid term for search if the string of alphanumeric characters comprises at least one digit.
15 . The system according to claim 14 , wherein the one or more processors are configured to normalize the determined valid term in compliance with a model number or a part number for searching for a model or a part of a product.
16 . The system according to claim 11 , wherein the one or more processors are operable to modify the term to comprise a model number or a part number.
17 . The system according to claim 11 , wherein the one or more processors are configured to:
assign a score to each of a plurality of character strings identified while analyzing search results; and
identify the one or more particular character strings according to the assigned scores.
18 . The system according to claim 11 , wherein the one or more processors are configured to:
generate one or more alternative search suggestions, and
select each of the one or more alternative search suggestions according to whether a characteristic of each of the one or more alternative search suggestions matches a predetermined characteristic condition.
19 . The system according to claim 18 , wherein the predetermined characteristic condition comprises:
the number of the particular character strings in each of the one or more alternative search suggestions being greater than a certain threshold; and
each of the one or more alternative search suggestions comprising a brand name.
20 . The system according to claim 11 , wherein the one or more processors are operable to store one or more alternative search suggestions in a form of autofill index.