Method for personalizing information retrieval in a communication network

• US 8,386,482 B2
• Filed: 07/05/2010
• Issued: 02/26/2013
• Est. Priority Date: 09/02/2009
• Status: Active Grant

• Current Assignee: Xurmo Technologies Private Limited
• Original Assignee: Xurmo Technologies Private Limited
• Inventors: Gopalakrishnan, Sridhar
• Primary Examiner(s): Wong, Leslie
• Application Number:
  US12/830,362
• Publication Number:
  US 20110055186A1
• Time in Patent Office: 967 Days
• Field of Search: 707/732, 707/733, 707/758, 707/765, 707/767, 707/999.01, 707/999.107, 707/E17.059, 707/E17.06
• US Class Current: 707/732
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering

Abstract

The various embodiments herein provide a method and system for personalizing information retrieval in a communication network. The method involves receiving one or more inputs from several sources and processing the information present in the received inputs. A contextual and personalized search is performed based on the received inputs and the results are displayed in support of the information in the communication network. The system has an application server for receiving the inputs from the sources and a processor for processing the information in the received inputs. One or more virtual documents are generated based on the processed information to perform a search and the results are displayed.

Claims

1. A method for personalizing information retrieval in a communication network, the method comprising:

receiving one or more inputs from a plurality of sources;

processing an information present in the one or more inputs received from the plurality of sources, wherein processing the information comprises refining the information present in the one or more inputs of the plurality of sources, enhancing the information present in the one or more inputs of the plurality of sources, boosting the information present in the one or more inputs of the plurality of sources and extracting the information present in the one or more inputs to build people and artifact profiles;

generating one or more virtual documents based on processed information present in the one or more inputs and wherein appropriate weights assigned to generate virtual profile documents for people and artifacts;

performing a search based on the virtual documents generated and wherein performing a search comprises performing a contextual and personalized search based on the inputs received;

generating an intelligent web of words from the one or more inputs of the plurality of sources;

performing an iterative comparison between the information present in a query document with the information of the profile document;

personalizing one or more results that to be displayed based on comparison of the results with the profile document of a user;

refining a search result based on user inputs; and

displaying one or more results in support of the information in the communication network.

2. The method according to claim 1, wherein the plurality of sources comprises an instant messenger, an electronic mail, a short message service, a browser, a customized client application, a voice command, and an incumbent information source.

3. The method according to claim 1, wherein the one or more inputs comprising at least one of a user interaction and a prevalent information content.

4. The method according to claim 1, wherein receiving one or more inputs from a plurality of sources comprises establishing a customized connection to the communication network and extracting information present in the plurality of sources.

5. The method according to claim 4, wherein establishing a customized connection to the communication network comprises at least one of establishing or connecting to an authenticated network and enabling privileged access to information in the communication network, and developing suitable interfaces for efficient extraction and search of information in the network.

6. The method according to claim 1, wherein processing the information further comprises synthesizing information present in the one or more inputs wherein synthesizing the information comprises building a social graph which is a web that defines the nature of interconnections between people in the network, creating profiles of people and artifacts in the communication network, and generating a virtual query document.

7. The method according to claim 6, wherein refining the information further comprises removing noise or unwanted information present in the one or more inputs of the plurality of sources through the use of a filters and wherein the filter includes a stop word filter and a histogram filter.

8. The method according to claim 6, wherein the enhancing the information comprises determining associations between words and determining a strength between associations of words.

9. The method according to claim 8, wherein determining association between words and the strength between associations of words is based on developing a web of words.

10. The method according to claim 9, wherein the web of words called Word-Web is built using statistical, mathematical and semantic data models.

11. The method according to claim 10, wherein developing the Word-Web uses at least one of a word net, one or more public websites, one or more domain specific websites, one or more predefined documents and user inputs.

12. The method according to claim 10, the Word Web is an ever growing corpus of words across various languages connected to each other in a non-hierarchical pattern.

13. The method according to claim 10, wherein the Word-Web circumvents the need for W3C prescribed semantic models.

14. The method according to claim 10, wherein the associations of words are remolded continuously to reflect the search engines increasing understanding of the environment.

15. The method according to claim 10, wherein the Word-Web is deployed over heterogeneous networks such as enterprise network, public networks and consumer groups.

16. The method according to claim 15, wherein a unified Word-Web which is an amalgamation of the Word-Web deployed over different networks developed to form a semantic dictionary adapted to understand words across languages and create meaningful associations between them.

17. The method according to claim 10, wherein a user-specific version of the Word-Web is developed and wherein the user-specific version of the World-Web captures a user'"'"'s perspective of relationship between words.

18. The method according to claim 10, wherein the intelligent Word-Web is versatile and created for various levels of granularity to capture nuances at the user level and the network level.

19. The method according to claim 6, wherein profiles of people and artifacts in the network are represented as virtual profile documents.

20. The method according to claim 6, wherein each user'"'"'s profile consists of a user specific version of the Word-Web, the relative strength of the user'"'"'s association with the words and the user'"'"'s relationship with information assets including people and artifacts in the network.

21. The method according to claim 6, wherein artifact profiles includes the content in the artifacts, its metadata and the people associated with in the communication network.

22. The method according to claim 6, wherein the personalized information retrieval search engine algorithms constantly evaluate effectiveness of relationships between words, people and documents so that the accuracy of the user profiles do not diminish with time.

23. The method according to claim 6, wherein a query document comprises at least one of the words relevant to the query post processing.

24. The method according to claim 23, wherein the processing of query involves refining, enhancing and boosting the query by performing natural language processing, by using a noise filters and by using the Word-Web and wherein the noise filter includes a stop word filter, a histogram filter, and an enhancer, and wherein the enhancer includes a co-locater, a stemmer and a lemmatizer, and wherein boosting of information is performed by assigning weights to words in the information of the one or more inputs based on different parameters and wherein the parameters includes at least one of source of information, the frequency of occurrence of words and the frequency of updation of words.

25. The method according to claim 1, wherein performing the search comprises performing an iterative comparison between the information present in query documents with the information of the profile documents and personalizing one or more results that are to be displayed based on the profile document of a user and the social graph of the network.

26. The method according to claim 1, further comprises displaying the results based on the inputs from the plurality of sources;

and refining the one or more results displayed.

27. The method according to claim 26, wherein displaying the results comprises showcasing details of the relevant people, artifacts, archives, advertisements and derived information in the network.

28. The method according to claim 26, wherein displaying the results comprises providing an analysis on the data in the communication network based on the criteria defined by the user.

29. The method according to claim 26, wherein the result refining due to user or system inputs comprises modifying the information present in the query documents, updating user profiles based on the feedback from refining results and updating the intelligent Word-Web.

30. The method according to claim 26, wherein the result refining further comprises performing a search based on the updated query and profile documents and displaying a personalized and context relevant one or more results.

31. A system for personalizing information retrieval in a communication network, the system comprising:

an application server to receive one or more inputs from a plurality of sources;

a processor to process an information present in the one or more inputs received from the plurality of sources, to generate one or more virtual documents based on the processing of the information present in the one or more inputs and to perform a search based on the virtual documents generated; and

a display unit for displaying one or more results based on the document generated for the information;

wherein the processor comprises a Word-Web generator to generate an intelligent web of words from the one or more inputs of the plurality of sources, and wherein the processor comprises a content indexer and pro filer to extract information from the one or more inputs of the plurality of sources to build people and artifact profiles, and wherein the processor comprises a content refiner for refining the contents from the one or more inputs of the plurality of sources, and wherein the processor comprises a content modeler and profile document generator to assign appropriate weights to generate virtual profile documents for people and artifacts, and wherein the processor comprises a search engine for performing an iterative comparison between the information present in a query document with the information of the profile document, and wherein the processor comprises a personalize module for personalizing one or more results that are to be displayed based on comparison of the results with the profile document of a user, and wherein the processor comprises a search result refiner to refine search results based on user inputs.

32. A system according to claim 31, wherein the processor further comprises:

a query refiner for refining the information present in the one or more inputs;

a query enhancer for enhancing the information present in the one or more inputs of the plurality of sources;

a query booster for boosting the information present in the one or more inputs of the plurality of sources and the boosting of information is performed by assigning weights to words in the information of the one or more inputs based on different parameters and wherein the parameters includes at least one of source of information, the frequency of occurrence of words and the frequency of updation of words; and

a query document generator for generating a virtual document based on the processing of the information.