Systems and methods for semantic search engine analysis
Systems and methods for semantic search engine analysis are disclosed. Generally, the system receives user provided text and/or speech search input, and further processes the search input to determine a semantic meaning. If the search input is speech, the system may first convert the speech into text. The system lexically processes the search input for ensuring the search input is valid, tags portions of the search input with metadata for applying a meaning of the tagged portions, and generally analyzes the relative locations of individual words and phrases to determine grammatical or linguistic relationships within the search input. In some embodiments, the system may disambiguate words or search terms, as well as provide input suggestions to the user. The system may use the manipulated search input to generate a search query, such as a query for searching apartment listing databases, and further display the search query results to the user.
1. A system comprising:
a web-based platform operatively connected to an electronic computing device running an application, wherein the web-based platform is configured to receive real-time user input from the electronic computing device in response to a user providing a plurality of search terms via the application, and wherein the plurality of search terms comprise location, price, and/or size information relating to available dwellings for rent or purchase; and
a semantic search engine operatively connected to the web-based platform, wherein the semantic search engine is configured to receive the real-time user input from the web-based platform, and wherein in response to receiving the real-time user input, the sematic search engine is operable to:
perform text processing on the real-time user input for determining a semantic context of the real-time user input via natural language processing algorithms, wherein performing text processing comprises parsing the real-time user input to detect phonemes representative of spoken words in the real-time user input and analyzing the detected phonemes to identify the plurality of search terms, and wherein the text processing generates a search query comprising a plurality of search terms based on the detected phonemes;
transmit the search query to a search database for matching the plurality of search terms with relevant data, wherein relevant data comprises dwelling listings and corresponding dwelling listing information;
receive search results from the search database, wherein the search results comprise matched relevant data; and
transmit the search results to the user for presenting the search results on a display at the electronic computing device, wherein the search results comprise one or more listings for available dwellings for rent or purchase, wherein each dwelling comprises characteristics matching a semantic meaning of the location, price, and/or size information provided in the real-time user input.
2. The system of claim 1 , wherein text processing further comprises:
performing lexical processing on the real-time user input, wherein lexical processing comprises parsing the real-time user input to remove typos, and further comprises generating lexically processed text objects from parsed real-time user input;
tagging the lexically processed text objects with metadata to generate tagged text objects;
performing syntax analysis on the tagged text objects, wherein syntax analysis comprises analyzing individual terms in the tagged text objects for identifying a relational text structure based on predetermined grammar rules; and
generating the search query, wherein the search query comprises a plurality of search terms including the lexically processed text objects and/or the tagged text objects, to be transmitted to the semantic search engine.
3. The system of claim 1 , wherein the semantic search engine is further operable to, in response to performing syntax analysis on the tagged text objects, perform a disambiguation process on the tagged text objects, wherein the disambiguation process comprises evaluating a tagged text object as multiple tagged text objects for independently processing each of the multiple tagged text objects for generating multiple search queries, wherein each of the multiple search queries comprises a separate search query with a different semantic context, and wherein a particular search query of the multiple search queries is determined to be an optimal search query based on its search results from matching with the relevant data in the search database.
4. The system of claim 1 , wherein the semantic search engine is further operable to generate input suggestions based on the tagged text.
5. The system of claim 4 , wherein input suggestions are generated based on models trained via machine learning algorithms.
6. The system of claim 1 , wherein the real-time user input comprises text and/or voice data.
7. The system of claim 6 , wherein prior to performing text processing, the system determines if the real-time user input comprises voice data.
8. The system of claim 7 , wherein in response to determining that the real-time user input comprises voice data, converting the voice data to text via speech-to-text algorithms.
9. The system of claim 2 , wherein tagging the lexically processed text objects comprises inserting metadata at locations in the lexically processed text objects identified to include a search keyword.
10. The system of claim 9 , wherein a search keyword comprises a particular state, city, and/or zip code, and the metadata includes a geographic location tag for classifying the search keyword.
11. A method, wherein a semantic search engine operatively connected to a web-based platform is configured to receive real-time user input from the web-based platform, wherein the web-based platform is operatively connected to an electronic computing device running an application and is configured to receive the real-time user input from the electronic computing device in response to a user providing a plurality of search terms via the application, and wherein the plurality of search terms comprise location, price, and/or size information relating to available dwellings for rent or purchase, and in response to receiving the real-time user input, the sematic search engine is operable to perform the method steps of:
performing text processing on the real-time user input for determining a semantic context of the real-time user input via natural language processing algorithms, wherein performing text processing comprises parsing the real-time user input to detect phonemes representative of spoken words in the real-time user input and analyzing the detected phonemes to identify the plurality of search terms, and wherein the text processing generates a search query comprising a plurality of search terms based on the detected phonemes;
transmitting the search query to a search database for matching the plurality of search terms with relevant data, wherein relevant data comprises dwelling listings and corresponding dwelling listing information;
receiving search results from the search database, wherein the search results comprise matched relevant data; and
transmitting the search results to the user for presenting the search results on a display at the electronic computing device, wherein the search results comprise one or more listings for available dwellings for rent or purchase, wherein each dwelling comprises characteristics matching a semantic meaning of the location, price, and/or size information provided in the real-time user input.
12. The method of claim 11 , wherein text processing further comprises:
performing lexical processing on the real-time user input, wherein lexical processing comprises parsing the real-time user input to remove typos, and further comprises generating lexically processed text objects from parsed real-time user input;
tagging the lexically processed text objects with metadata to generate tagged text objects;
performing syntax analysis on the tagged text objects, wherein syntax analysis comprises analyzing individual terms in the tagged text objects for identifying a relational text structure based on predetermined grammar rules; and
generating the search query, wherein the search query comprises a plurality of search terms including the lexically processed text objects and/or the tagged text objects, to be transmitted to the semantic search engine.
13. The method of claim 11 , wherein the semantic search engine is further operable to perform the step of, in response to performing syntax analysis on the tagged text objects, performing a disambiguation process on the tagged text objects, wherein the disambiguation process comprises evaluating a tagged text object as multiple tagged text objects for independently processing each of the multiple tagged text objects for generating multiple search queries, wherein each of the multiple search queries comprises a separate search query with a different semantic context, and wherein a particular search query of the multiple search queries is determined to be an optimal search query based on its search results from matching with the relevant data in the search database.
14. The method of claim 11 , wherein the semantic search engine is further operable to perform the step of generating input suggestions based on the tagged text.
15. The method of claim 14 , wherein input suggestions are generated based on models trained via machine learning algorithms.
16. The method of claim 11 , wherein the real-time user input comprises text and/or voice data.
17. The method of claim 16 , wherein prior to performing text processing, the method further comprises the step of determining if the real-time user input comprises voice data.
18. The method of claim 17 , wherein in response to determining that the real-time user input comprises voice data, converting the voice data to text via speech-to-text algorithms.
19. The method of claim 12 , wherein tagging the lexically processed text objects comprises inserting metadata at locations in the lexically processed text objects identified to include a search keyword.
20. The method of claim 19 , wherein a search keyword comprises a particular state, city, and/or zip code, and the metadata includes a geographic location tag for classifying the search keyword.