SYSTEMS AND METHODS FOR DETERMINING CONTEXTUALLY-RELEVANT KEYWORDS

• US 20190332717A1
• Filed: 09/28/2018
• Published: 10/31/2019
• Est. Priority Date: 04/30/2018
• Status: Active Grant

• Current Assignee: Innoplexus AG
• Original Assignee: Innoplexus AG
• Inventors: Keskar, Abhijit
• Granted Patent: US 11,182,410 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 21/6209 to a single file or object, e.g. in a secure envelope, encrypted and accessed using a key, or with access control rules appended to the object itself
• CPC Class Codes: G06F 40/247 Thesauruses; Synonyms
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition

Abstract

Disclosed is a system for determining contextually-relevant keywords related to text of an electronic document. The system comprises a memory, a database arrangement comprising an ontology and a synonym databank, a data processing arrangement and a sever arrangement. The data processing arrangement is operable to fetch the electronic document stored within the memory. Furthermore, the data processing arrangement determines the common words from the ontology and the synonyms corresponding to the common words from synonym databank. Moreover, the data processing arrangement determines the generated set of keywords comprising the common words and the synonyms. The server arrangement determines a preference score and an importance score for each keyword of the generated set of keywords. Furthermore, the server arrangement determines the cumulative rank of each keyword based on the preference score and the importance score, therefore determining contextually-relevant keywords.

Claims

1. A system for determining contextually-relevant keywords related to text of an electronic document, the system comprising:

a memory operable to store the electronic document therein;

a database arrangement comprising an ontology and a synonym databank;

a data processing arrangement communicatively coupled to the memory and the database arrangement, wherein the data processing arrangement is operable to

fetch the electronic document stored within the memory,

extract the text of the electronic document,

receive a plurality of words from the ontology,

compare each of the plurality of words with the text of the electronic document, to identify common words from the ontology that appear within the electronic document,

receive from the synonym databank, a plurality of synonymous words corresponding to the common words, and

generate a set of keywords comprising the common words and the plurality of synonymous words corresponding to the common words; and

a server arrangement communicatively coupled to the database arrangement and the data processing arrangement, wherein the server arrangement is operable to

receive the set of keywords from the data processing arrangement,

determine a preference score for each keyword of the set of keywords, wherein the preference score is based on one or more preference parameters,

identify a semantic relationship of each keyword with other keywords of the set of keywords,

determine an importance score of each keyword of the set of keywords, based on the semantic relationship,

determine a cumulative rank for each keyword of the set of keywords, based on the preference score and the importance score of the keyword, and

determine, based on the cumulative rank of each keyword of the set of keywords, the contextually-relevant keywords related to the text of electronic document.

2. The system of claim 1, wherein the electronic document is stored within the memory in an encrypted format.

3. The system of any one of the claims 1, wherein the memory is implemented within a client device and wherein the data processing arrangement is implemented in one of:

the client device, or the server arrangement.

4. The system of any one of the claims 1, wherein the ontology is a hierarchical ontology comprising a plurality of nodes, and wherein each of the plurality of nodes is associated with a plurality of words related to a specific subject-matter.

5. The system of claim 4, wherein the data processing arrangement is further operable to:

receive an input from a user associated with the electronic document, wherein the received input corresponds to a subject-matter associated with the text of the electronic document;

compare the subject-matter corresponding to the input with the subject-matter corresponding to the plurality of nodes;

identify, based on the comparison, the node of the plurality of nodes related to the subject-matter corresponding to the input; and

receive the plurality of words associated with the node of the plurality of nodes.

6. The system of claim 1, wherein the data processing arrangement is operable to identify the common words and receive the plurality of synonymous words corresponding to the identified common words in a:

sequential manner, or

simultaneous manner.

7. The system of claim 1, wherein the one or more preference parameters comprise:

frequency of occurrence of each keyword of the set of keywords within search strings previously used by a plurality of users of a search engine, frequency of occurrence of the common words within the text of the electronic document, location of occurrence of the common words within the text of the electronic document and publicly-available information corresponding to each keyword of the generated set of keywords of the electronic document.

8. The system of claim 1, wherein the server arrangement is operable to identify the semantic relationship of each keyword with other keywords of the generated set of keywords, based on at least one of:

significance of the text of the electronic document and publicly-available information corresponding to each keyword of the generated set of keywords of the electronic document.

9. The system of claim 5, wherein the server arrangement is further operable to:

generate, based on the cumulative rank of each keyword of the set of keywords, a list comprising the contextually-relevant keywords arranged in at least one of;

an increasing order, or a decreasing order of the cumulative ranks; and

recommend the list of the contextually-relevant keywords to the user associated with the electronic document.

10. The system of claim 1, wherein the server arrangement is operable to implement a machine-learning algorithm for determining the contextually-relevant keywords related to the text of electronic document.

11. A method for determining contextually-relevant keywords related to text of an electronic document, wherein the method comprises:

extracting the text of the electronic document;

receiving a plurality of words from an ontology;

comparing each of the plurality of words with the text of the electronic document, to identify common words from the ontology that appear within the electronic document;

receiving a plurality of synonymous words corresponding to the common words;

generating a set of keywords comprising the common words and the plurality of synonymous words corresponding to the common words;

determining a preference score for each keyword of the set of keywords, wherein the preference score is based on one or more preference parameters;

identifying a semantic relationship of each keyword with other keywords of the set of keywords;

determining an importance score of each keyword of the set of keywords, based on the semantic relationship;

determining a cumulative rank for each keyword of the set of keywords, based on the preference score and the importance score of the keyword; and

determining, based on the cumulative rank of each keyword of the set of keywords, the contextually-relevant keywords related to the text of electronic document.

12. The method of claim 11, wherein the method further comprises storing the electronic document in an encrypted format.

13. The method of any one of the claims 11, wherein the ontology is a hierarchical ontology comprising a plurality of nodes, and wherein each of the plurality of nodes is associated with a plurality of words related to a specific subject-matter.

14. The method of claim 13, wherein the method further comprises:

receiving an input from a user associated with the electronic document, wherein the input corresponds to a subject-matter associated with the text of the electronic document;

comparing the subject-matter corresponding to the input with the subject-matter corresponding to the plurality of nodes;

identifying, based on the comparison, the node of the plurality of nodes related to the subject-matter corresponding to the input; and

receiving the plurality of words associated with the node of the plurality of nodes.

15. The method of claim 11, wherein the method comprises identifying the common words and receiving the plurality of synonymous words corresponding to the common words at in a:

sequential manner, or

simultaneous manner.

16. The method of claim 11, wherein the one or more preference parameters comprise:

frequency of occurrence of each keyword of the set of keywords within search strings previously used by a plurality of users of a search engine, frequency of occurrence of the common words within the text of the electronic document, location of occurrence of the common words within the text of the electronic document and publicly-available information corresponding to each keyword of the generated set of keywords of the electronic document.

17. The method of claim 11, wherein the method comprises identify the semantic relationship of each keyword with other keywords of the generated set of keywords, based on at least one of:

significance of the text of the electronic document and publicly-available information corresponding to each keyword of the generated set of keywords of the electronic document.

18. The method claim 14, wherein the method further comprises:

generating, based on the cumulative rank of each keyword of the set of keywords, a list comprising the contextually-relevant keywords arranged in at least one of;

an increasing order, or a decreasing order of the cumulative ranks; and

recommending the list of the contextually-relevant keywords to the user associated with the electronic document.

19. A software product recorded on machine-readable non-transient data storage media, wherein the software product is executable upon computing hardware to implement a method of claim 11.